Online Superbowl betting 2005

How to Pick a Price Per Head Provider

2005-06-21T13:31:00.000-07:00

Featuring Dalton Wagner, Founder V.O. Group, S.A.

The number of Price Per Head providers has grown exponentially over the last twelve months with entrants to the marketplace almost daily. With so many choices, the big question is: “How do I know which Price Per Head provider or sports book to choose?”

When evaluating an offshore operation, it is often difficult to sort through the facts, for this reason, you must rely on the following to make a sound choice:

Years In Business: Make sure you choose a Price Per Head operator that has been in business for at least 6 years. A proven track record is a must. You can not afford to risk your clients or your sports season to a fly by night operator or ‘wanna be’ operator. Avoid new sportsbooks. Avoid anyone with an unreasonably low price (<$25/head/week) who may be attempting to gain entry into the market (sign of a new company).

Dalton Wagner Commentary: I invented the concept of PPH in 1999. I have the largest economies of scale and I write more PPH clients than anyone offshore. And, this is in addition to my sportsbook business, my casino business, my racebook business and my poker business. The fact is I make very little money at $25/head/week. It scares me to see new operators entering the market with prices as low as $25. I can tell you that unless they are writing over 15,000 head, they can not make it. Let me re-phrase that…if they are quoting prices as low as $25/head/week there is no way that they have enough employees to service an agent, there is no way they have tri-dundant phone service, there is no way they have tri-dundant internet service – in short, there is no way they can service an agent’s business well enough to help the agent grow. Secondly, PPH companies will lie to you about their company’s age. You must be able to verify the companies age. This is a must. Use Internic, use OSGA.com, and use ThePrescription.com. You must verify the facts.

Location: Check to make sure that the PPH provider you will outsource your clients to is licensed and legally able to service clients world-wide.

Dalton Wagner Commentary: There are only two locations that we know of that allow the outsourcing of bookmaking. The two locations are Costa Rica and Panama. The UK and Dutch Antilles to NOT allow this activity in their licenses. The only one you can trust at the current time is Costa Rica (as evidenced by the recent blow up of www.BetPanAm.com in Panama).

Reputation/Integrity: One of the best ways to make sure that your clients and your business will be safe with a provider is by checking their reputation. Use verifiable and known sources of information like the Off Shore Gaming Association, The Prescription, Bettors World, etc.

Dalton Wagner Commentary: If your outsourcing your business to someone, you expect them to treat it as their own. The person accepting the outsourcing better have a reputation as strong as or stronger than yours. This is the only way you should trust your business to them. One additional note, if the PPH Company also has a post-up division, check the reputation of the post-up side of the company. This will tell you volumes about the company. Rather, if the PPH company is tied to a loser post-up sportsbook or one that has a negative reputation, you know what to expect for your clients.

Service and Offerings: Reputable PPH companies are open 24-hours a day, 7-days a week, 365-days a year. They offer quarter lines, half time lines, game lines, money lines, propositions, multiple casinos, horse betting, and more. You are outsourcing to better service your clients – therefore service is one of the key deciding factors on who to go with.

Dalton Wagner Commentary: Simply put, the more accessible the service is and the more wagering options that exist, the more money the agent will make. You should consider the service mentioned above to be the minimum acceptable. Rather, if they have one casino – you should NOT work with them. If they do not have horses – you should NOT work with them. If they are not open 24-hours a day – you should NOT work with them, PERIOD!

Security: A PPH company must have safeguards and must keep your information and clients secure and anonymous If the company you are considering is publicly held, then your security is compromised (it CAN BE accessed via the government). If the company you are considering has integrity issues, you can assume your data will be passed on to the wrong people.

Dalton Wagner Commentary: Security is the ultimate issue. Our company has had the opportunity to be purchased or taken public on multiple occasions. We decided not to do so because it would put our clients/agents/bookmakers in a compromised situation. Partner agents and bookmakers use our service because our integrity and the security of their data is complete.

Additional Information on a Price Per Head Agency Relationship (also known as outsourcing, outsourced bookmaking, sportsbook agent, sportsbook service, call center outsourcing, etc.):

The concept of Price Per Head Agency or Outsourcing Bookmaking to an Offshore Sports book has become popular recently. By using software written by offshore sports books, agents or illegal credit bookmakers have the ability to outsource the writing of their business to companies like www.BettorsNet.com (owned by V.O. Group, S.A. one of the largest and most respected offshore operators in the world – contact is 1-877-512-1001). Therefore, the agent or bookmaker simply points their clients to a phone number and/or website and allows them to bet with a simple Pin/Password system. The agent or bookmaker then pulls reports at the end of the week, and pays or collects what the client won or lost. The only money that trades hands between the agent or bookmaker and the service provider is a simple Price Per Head service fee.

Obvious benefits of the Price Per Head Agency relationship are (i) improved earning potential for the agent, (ii) decreased legal exposure for the agent, (iii) increased customer service for the agent’s clients, and (iv) ability for the agent to do something other than answer wagering calls (they actually get time to enjoy the money they are earning).

Dalton Wagner’s commentary is found throughout the above. Mr. Wagner has the reputation of being the most innovative operator in offshore gaming and the largest Price Per Head operator offshore. His most recent innovation has allowed small sportsbooks to outsource their complete operation to V.O. Group, S.A. while focusing completely on marketing. With formidable competition, Mr. Wagner founded V.O. Group, S.A. in 1998. While most companies formed that late failed, V.O. Group, S.A. has become one of the top 10 operators offshore (as measured by AnteUp; Gambling Online Magazine; Poker Player Magazine; and Many other Publications).

BetCris.com Opinion on Price Per Head Explored

2005-06-21T12:33:00.000-07:00

Featuring Dalton Wagner, Founder V.O. Group, S.A.

The concept of Price Per Head Agency or Outsourcing Bookmaking to an Offshore Sports book has become popular recently. By using software written by offshore sports books, agents or illegal credit bookmakers have the ability to outsource the writing of their business to companies like www.BettorsNet.com. Therefore, the agent or bookmaker simply points their clients to a phone number and/or website and allows them to bet with a simple Pin/Password system. The agent or bookmaker then pulls reports at the end of the week, and pays or collects what the client won or lost. The only money that trades hands between the agent or bookmaker and the service provider is a simple Price Per Head service fee.

This new system is a departure from the way it used to work in which the agent would split the winnings or losings with the offshore sportsbook.

Obvious benefits of the Price Per Head Agency relationship are (i) improved earning potential for the agent, (ii) decreased legal exposure for the agent, (iii) increased customer service for the agent’s clients, and (iv) ability for the agent to do something other than answer wagering calls (they actually get time to enjoy the money they are earning).

There has been much debate over which of the two methods is better for the partner agent or bookmaker. In an effort to get to the bottom of this, we attempted to contact the dominant player in each of the models:

Revenue Split Model: The owners of www.BetCris.com (Costa Rica International Sportsbook is a well known offshore credit bookmaker who operates only on a revenue split model) were contacted but refused to return our call or comment;

Price Per Head Agency Model: Dalton Wagner, proprietor of V.O. Group, S.A. and BettorsNet.com (one of the largest offshore gambling operators who only operates on a Price Per Head basis when dealing with credit agents) was contacted and agreed to be interviewed.

Unfortunately, since only one of the persons contacted agreed to participate, our interview will show mostly his view. However, we did our best to probe the positives and negatives of each model.

Post-Up Newsletter: The Price Per Head model appears to be the most appropriate model for an agent or bookmaker to maximize his revenue. Why then are companies like www.BetCris.com so against them?
D. Wagner: The answer is simple dollars and cents. If you have a client that loses $10,000 in a week and you have a Price Per Head Agency relationship with www.BettorsNet.com, you will simply need to pay your service fee of $25-$30 for that client using our service. However, if you are using a revenue split relationship, you will have to fork over $5,000. So, in the example you provided, companies like the one you mentioned are in danger of losing $4,970 in revenue if their agents find out about PPH.

Post-Up Newsletter: Yes, I understand the math, but when speaking to companies like www.BetCris.com they act as if the PPH industry is putting the offshore industry in danger. They speak of people like you as if you had the plague. Why?
D. Wagner: I know the gentleman that owns Bet Cris. I know the guys who run Bet Cris. They are the best in the world at what they do. There is no way I would dispute that and no way anyone would dispute that. But, BetCris.com can not exist by taking all of their revenue split relationships and having them converted to PPH. So, if I had to guess, they act that way because I am endangering their way of life. I have actually had conversations with the owner of BetCris.com and he firmly feels that he provides his agents a better opportunity to earn more, therefore he feels his much higher price is warranted. I don’t know if I agree, but I think it best for my reputation to speak for itself. The fact is that I have NEVER had an agent or bookmaker come to me who switched to PPH from revenue split EVER GO BACK. 100% of the people who have tried my service are still with me from day 1. That shows me that agents do not agree with BetCris.com. They think the service is comparable and they think the earning potential is greater with www.BettorsNet.com.

Post-Up Newsletter: I am lost. Why does the owner of BetCris.com feel he provides his agents a better opportunity to earn more? The math seems obvious.
D. Wagner: Actually, I agree with him in one way and one way only. If an agent is weak and is booking ‘over his head’ (an industry term that implies the bookmaker or agent is taking more risk than they can afford in the long term), then a partnership with www.BetCris.com can be a positive thing. Not only do you share 50% of your revenue with them on a revenue split relationship, but you also share 50% of the risk. So, if you are an agent that has a bettor that wants to bet $5,000 a game, and you can only afford to book $2,500, then a partnership with Bet Cris is a good thing and will allow you to stay in business longer. You should never ever book over your head.

Post-Up Newsletter: Oh, I see. So that is an example of an agent you would tell to use a revenue split model, right?
D. Wagner: Yes. And, No. Yes, if the agent plans on booking over his head for the long term. No, if he can cut his client down to an acceptable level and keep all the profit for himself. Then, when the agent or partner bookmaker has a large enough bank roll, he can open the client up to the larger wagers. I still believe, and I believe the math proves that PPH is always bettor for the agent.

Post-Up Newsletter: Are there any other Price Per Head agency companies that you would recommend?
D. Wagner: Absolutely NOT. Please print that in big bold letters. It is difficult for any company in our industry to survive. The overhead is huge. Most Price Per Head companies have come to Costa Rica or Panama or Antigua to try and compete with just Price Per Head revenue. Trust me, I have been doing this for 7 years, and PPH does NOT generate enough for survival. You have to build a relationship with a company that has a huge staff, technical know-how, tri-dundant phones, tri-dundant internet service, millions of dollars in investment in hardware and software – and, we are the only ones that have it at the current time. Why? Because we have been operating major post-up brands for years. We are one of the most successful operators offshore and we are able to write PPH as a sideline business. We could NEVER survive on PPH alone. And, anyone who says they can, is lying to you. The guys who say they can write business at $17or $20/head are on their way out – they just don’t know it yet.

Post-Up Newsletter: So how does a bookmaker or agent learn more about this?
D. Wagner: Simply go to any of the V.O. Group, S.A. sites to learn more. I suggest starting at www.BettorsNet.com or calling 1-877-512-1001 or emailing agents@bettorsnet.com to get more information.

Additional Information on the Companies Mentioned Above:

As Founder and Proprietor of V.O. Group, S.A., Dalton Wagner has survived multiple buyout attempts and maintains one of the largest privately held offshore gaming companies in the industry. While the rest of the industry is going public and being required to provide more and more information in a public format, Mr. Wagner’s privately held company is a favorite associate and partner for bookmakers and agents onshore. Privacy is one thing you never have to worry about when dealing with Dalton and V.O. Group, S.A. With formidable competition, Mr. Wagner founded V.O. Group, S.A. in 1998. While most companies formed that late failed, V.O. Group, S.A. has become one of the top 10 most admired offshore companies (as measured by Online Gambling Magazine; Post-Up Newsletter; Offshore Operator Industry Magazine; and Many other Publications). If you are looking for a Price Per Head Agency relationship, there is no place better online or offshore.

BetCris.com is considered the undisputed leader in offshore credit bookmaking. The owner of the company has been called the Godfather of the Offshore Bookmaking Industry in many publications. It is estimated that 90% of Costa Rica International Sportsbooks revenue is generated via revenue splits with Partner Agents and Bookmakers. If you are looking for a Revenue Split relationship, there is no place better online or offshore.

Benefits of Price Per Head (PPH) Exploration & Quantification

2005-06-21T09:56:00.000-07:00

Q & A with
Dalton Wagner, Founder V.O. Group, S.A.

Dalton Wagner, Proprietor of V.O. Group, S.A. has found a great deal of success by tripling his bottom line profit numbers with the addition of what has been called the Greatest Price Per Head Offering online. While his Board of Directors and Investors are thrilled, we wanted more than vague statements – we wanted facts. We recently caught up with Dalton in San Juan, Puerto Rico while he was on vacation at El Conquistador Hotel and Resort. Our goal was to Explore and Quantify the actual benefits that a Price Per Head (PPH) Agency Relationship offers the bookmaker, agent or entrepreneur.

As Founder and Proprietor of V.O. Group, S.A., Dalton Wagner has survived multiple buyout attempts and maintains one of the largest privately held offshore gaming companies in the industry. While the rest of the industry is going public and being required to provide more and more information in a public format, Mr. Wagner’s privately held company is a favorite associate and partner for bookmakers and agents onshore. Privacy is one thing you never have to worry about when dealing with Dalton and V.O. Group, S.A. With formidable competition, Mr. Wagner founded V.O. Group, S.A. in 1998. While most companies formed that late failed, V.O. Group, S.A. has become one of the top 10 most admired offshore companies (as measured by Online Gambling Magazine; Post-Up Newsletter; Offshore Operator Industry Magazine; and Many other Publications).

The Offshore Operator: We have heard about all of the success you have found with the best Price Per Head (PPH) offering in the offshore gambling industry. However, we are not hearing actual numbers. Can you tell us how many clients you are now writing business for?
Dalton Wagner: As you know, the answer really depends on the season you are speaking about. In football season, we are now writing about 7,500 PPH clients a week (7,500 clients of our partner agents which outsourced the writing of their bets via our service and software). And, in baseball season we have historically written only half of that number. But, there is a very favorable trend occurring for our Agent and Bookmaker Partners. Since the addition of our Horse Wagering at over 300+ tracks and Two Casino Platforms with over 35 games, we are now seeing an increasing number of clients playing year-round. This, as you know, makes our Agent and Bookmaker Partners very, very happy.

The Offshore Operator: 7,500 PPH clients clearly makes you the largest PPH provider offshore. What do you attribute your success to?
Dalton Wagner: I am not sure we can find one distinct reason for our success. But, I can tell you what our Agent and Bookmaker Partners tell us. On average an Agent or Bookmaker partner doubles his revenue when he switches his clients to our service. The reason for this is access to 24-hour wagering online or via our call center, about 700% more wagering options when you look at our period wagering and propositions, and the addition of two casinos and horses has not hurt either. So, if you are a bookmaker or agent making $200k a year, you can reasonably expect to make $400k a year using our service. It is virtually guaranteed.

The Offshore Operator: Dalton, you are famous for making that kind of statement. But, can you back it up with quantifiable numbers?
Dalton Wagner: Absolutely I can. Ask me what you would like.

The Offshore Operator: OK. Let’s start with examples of how you can guarantee a 200% increase in earnings for Partner bookmakers or agents?
Dalton Wagner: Easy enough. In the United States, Mexico and Canada the typical credit bookmaker offers lines on games and totals. Sometimes they offer money lines too – but, rarely. And, these same bookmakers do not offer period betting or propositions. Furthermore, they take wagers only over the phone and only 2-4 hours a day. If you use our Price Per Head Software, you offer spreads, money lines, totals, half-time spreads, half-time money lines, half-time totals, quarter lines, propositions, and we allow access to betting 24-hours a day; 7-days a week; 365 days a year. In summary, I can safely say we offer 600% of the access time a normal credit bookmaker would provide as we are opened 24 hours compared to 4 hours. And, I can safely say we offer 700% more wagering options. These increases of 600% in access and 700% in wagering options easily equate to 200% more revenue for the partner bookmaker or agent.

The Offshore Operator: I understand, but I don’t necessarily agree. For example, if I am a credit bettor and I have $200 to lose, does it matter that I can access your site 24-hours a day?
Dalton Wagner: That question shows that you are not a bookmaker or agent. The reason is, the answer is ABSOLUTELY YES. If you compare a post-up client to a credit client, the credit client will typically lose 3-5x what the post-up client will. The reason is access. If you give a client access to credit, and access to use the credit, he will use it. Heck, forget about that example…if you are married and your wife has a credit card, you know that. The fact is as you allow more access to betting options and credit, the client will lose more. And, my current partner bookmakers and agents report to me that they typically double their earnings within weeks after joining my service.

The Offshore Operator: OK. That makes more sense to me. How about the casino, how does that affect the partner bookmaker or agents earnings?
Dalton Wagner: The online casino is relatively new to partner bookmakers and agents. The fact is that in the past, agents were scared to open up their clients to online casinos. The reason is that when clients lose in sports – they know they were treated fairly. However, when they lose in an online casino, for some reason their first reaction is “I was cheated”. However, as integrity in online gaming has risen over the years, so has trust in online casinos. For this reason, you are starting to see partner agents and bookmakers open their clients up to our two online casinos. And, boy, when they do – they are very happily surprised. An online casino allows wagering 24-hours a day. This means revenue potential for the agent or bookmaker 24-hours a day. It’s nice to run your reports in the morning and realize you made an extra $20,000 while you were sleeping! We estimate that when a partner agent or bookmaker opens up his clients to our casino, they will make an extra 15%-20% a year. This is purely earned off casino winnings.

The Offshore Operator: Fifteen to Twenty percent a year is nothing to laugh at. What about the horses?
Dalton Wagner: The horse product we have offers 300+ tracks including harness, thoroughbred and dogs. It is not just horses. Typically when the partner bookmaker or agent opens it up, adoption is slow. But, people who bet horses and dogs, love to bet horses and dogs. It is not rare to see a client who likes the ponies to double his daily action. And, as any bookmaker knows, in sports you hope to hold 6% on turnover. Well, in horses you hope to hold 20%. The difference in expected hold is where the value is in horses.

The Offshore Operator: Wow! So, having all the offerings really does matter doesn’t it?
Dalton Wagner: What I am selling is access – pure and simple. My goal is to take a partner agent or bookmakers clients and put them in the candy store 24-hours a day surrounded by all of their favorite treats. And, as you know, when any of us are put in that situation, we will indulge. And, they do.

The Offshore Operator: Have you ever had anyone become a partner Agent or Bookmaker and leave because they did not like what you had to offer?
Dalton Wagner: That is a very interesting question. The fact is, and I should have used this in the past know that I think about it, the answer is NO. As a matter of fact, every single person to use our service or try our service has been with us since initiating trial. We really are, without exaggeration, at 100%. Thank you for pointing that out to me.

The Offshore Operator: How could you not have known that?
Dalton Wagner: Embarrassingly, I never thought about it. I have always said “try it, you’ll love it”. But, I guess I never looked at the quantifiable fact – NOBODY HAS EVER TRIED OUR SERVICE AND LEFT OUR SERVICE they’re clients love it, the agents love it, and the agents make more than ever before – so they stay around!

The Offshore Operator: So how does a bookmaker or agent learn more about this profitable product?
Dalton Wagner: Simply go to any of the V.O. Group, S.A. sites to learn more. I suggest starting at www.BettorsNet.com or calling 1-877-512-1001 or emailing agents@bettorsnet.com to get more information. V.O. Group, S.A. sites include www.BettorsNet.com; www.MVPsportsbook.com; www.V-Wager.com; www.Players-SB.com; etc.

Success With Price Per Head (Small Sportsbook Success)

2005-06-21T07:37:00.000-07:00

Case Study in Small Sportsbook Success via Use of Price Per Head (PPH) Software Written and Owned by V.O. Group, S.A., the most successful offshore sportsbook in Costa Rica.

Price Per Head vs. Small Sportsbook Operation

Featuring Dalton Wagner, Founder V.O. Group, S.A.

As Founder and Proprietor of V.O. Group, S.A., Dalton Wagner has the reputation of being the most innovative operator in offshore gaming. His most recent innovation has allowed small sportsbooks to outsource their complete operation to V.O. Group, S.A. while focusing completely on marketing. With formidable competition, Mr. Wagner founded V.O. Group, S.A. in 1998. While most companies formed that late failed, V.O. Group, S.A. has become one of the top 10 operators offshore (as measured by AnteUp; Gambling Online Magazine; Poker Player Magazine; and Many other Publications).

Playing Poker Online Magazine: Recently we have been told that V.O. Group, S.A. is now allowing small sportsbook operations to outsource the complete writing of their business to your organization. How is this possible?
Wagner: The situation you speak about is 100% true. However, has only happened one time in our history.

Playing Poker Online Magazine: Can you be more specific?
Wagner: When an offshore sportsbook starts, in 99% of all cases it will fail in the first 24 months of operations given the current market conditions. Examples are everywhere from The Dunes, Platinum, Super Sports Book, etc. It happens every single year. In the past, large sportsbooks like MVP Sportsbook or V-Wager.com would swoop in and try to save the books via financial help or simply buy them at a low price. Well, V.O. Group, S.A. now offers another more innovative solution.

Playing Poker Online Magazine: OK. You’ve got me, explain!
Wagner: We write business for onshore bookmakers and agents charging them a simple price per head per week. This price takes into consideration our considerable economies of scale and a small profit margin. In the case you are speaking of, instead of buying the sportsbook for a low cost that basically sent the owners home broken hearted and with empty pockets, we cut them a deal where they could focus on marketing and leverage off of our economies of scale. Simply put, we took over writing their business for them, and they now focus their limited resources on marketing 100% of the time.

Playing Poker Online Magazine: That sounds like snake oil. How exactly would that work?
Wagner: We don’t sell snake oil. The facts speak for themselves. When V.O. Group, S.A. opened it’s doors in 1998 the market was different. You could recruit players, write their business, and make a profit in years one, two and three. With the current market conditions, there is NO WAY a start up credit or post-up book can be profitable in year one, two or three without 2,000 head a week in business. That is the theoretical breakeven point. If you are writing less than 2,000 head a week, you can double or triple your net earn by letting V.O. Group, S.A. write your business.

Playing Poker Online Magazine: OK. That makes sense because in 1998 and 1999, it was cheaper to find a new client. Is that where the difference in earnings comes from?
Wagner: The fact is yes, that is where the difference is. But, that is not our selling point. Our selling point is that if you stop worrying about writing business and let us do it, you can focus on marketing – which is the key to the success of an online sportsbook. And, that is the simple fact for credit bookmaking, post-up bookmaking and future success in this industry. As long as you think ‘beating the clients’ is the key, you are limiting your business. Let our company worry about that – and, you find clients. Spend your time finding clients and you will make money.

Playing Poker Online Magazine: The theory is sound, but what about the facts?
Wagner: The owners of the sports operation in question do not want me to be specific about whom they are. But, let me simply say that when we took over their operations, they could not even make the post-up money good. Now they have $1.4 million in the bank. So, yes, the facts do follow the theory.

Playing Poker Online Magazine: Wow! That is incredible. Can any poker room, casino, racebook or sports book do this with V.O. Group, S.A.
Wagner: Yes. We have programs for all online products except for bingo. And, by 2006 we should be able to offer options for even bingo operators..

Playing Poker Online Magazine: So how does a bookmaker or current operation learn more about outsourcing their operations to V.O. Group, S.A.?
Wagner: Simply go to any of the V.O. Group, S.A. sites to learn more. I suggest starting at www.BettorsNet.com or calling 1-877-512-1001 or emailing agents@bettorsnet.com to get more information.

Article Note: The situation in this Case Study is a true case, however may not be representative of your situation. All situations are different and therefore may result differently.

Success With Price Per Head (One Agents Story)

2005-06-20T08:22:00.000-07:00

Case Study demonstrating one agents success via Use of Price Per Head (PPH) Software Written and Owned by V.O. Group, S.A., the most successful offshore sportsbook in Costa Rica.

Price Per Head vs. One Agents Old Operation

Featuring Dalton Wagner, Founder V.O. Group, S.A.

As Founder and Proprietor of V.O. Group, S.A., Dalton Wagner has the reputation of being the most innovative operator in offshore gaming. His most recent innovation involves the direct attack upon the Onshore Bookmaking market via the unique value proposition of actually partnering with Bookmakers and Agents onshore (as opposed to attempting to steal their business). With formidable competition, Mr. Wagner founded V.O. Group, S.A. in 1998. While most companies formed that late failed, V.O. Group, S.A. has become one of the top 10 operators offshore (as measured by AnteUp; Gambling Online Magazine; Poker Player Magazine; and Many other Publications).

Sports Betting Business: The ‘buzz’ in offshore gambling appears to be focusing on only two subjects as of late, including: Poker and Your Price Per Head Offering. What exactly is Price Per Head and Price Per Head Agency?
Wagner: As you know, one market that a post-up sportsbook has historically been unable to attack is the huge market of bettors that have been betting with their local guy for years (local guy = agent or corner bookie). The reason this market exists is the personal relationship that an onshore bookie or agent has with his clients. Sportsbooks have attempted to educate the player and lure him away from the local bookie, however V.O. Group, S.A. has decided to take another tact. Via our Price Per Head software written by Matthew Wilson, we have been able to PARTNER with onshore bookies or agents with our Price Per Head Agency Relationship.

Sports Betting Business: You called us with a specific Success Story that you said Had To Be Told Immediately! What in the heck could be so important?
Wagner: As the industry knows, we have been pushing Price Per Head Agency for a while now. We have a story that has to be told…it is that simple!

Sports Betting Business: OK. Go ahead and tell the story!
Wagner: The story starts with a gentleman named Paul Harris in Austin, Texas who is was a small town bookmaker making $250,000 a year in income.

Sports Betting Business: I already see where this is going. Now you are going to tell me of how V.O. Group, S.A. leveraged your PPH software and relationship with Paul to make him 30% or 50% more. Right?
Wagner: No, that is not the story at all. While 50% would truly be something to celebrate, it does not compare to this story!

Sports Betting Business: Stop already. Tell the story.
Wagner: Mr. Harris contacted V.O. Group, S.A. through our BettorsNet.com brand in August 2004, right before football season. After three phone calls we switched his 75 players from playing through his cell phone to playing via our website and call center. And, with the increased availability to the product, more offerings, and all the other items BettorsNet.com brings, guess how much Paul made?

Sports Betting Business: Well, since you did not bite on 50% more, I will say 100% more. But I doubt it!
Wagner: Yea, I doubt it too. How about 350% more. That’s right! I just ran the reports for Paul Harris and up to February 1, 2005, Paul has made $750,000.00.

Sports Betting Business: Wait, $750,000 is a hell of a lot more than 350% more. Isn’t it?
Wagner: Actually, if you take into consideration the decline we will see after the end of March Madness, he should easily make $1,200,000 this year. But, I didn’t want to overstate the facts!

Sports Betting Business: Are you telling me that by having you take the phone calls and do the accounting for Mr. Harris you made this difference in his life?
Wagner: That is exactly what I am saying!

Sports Betting Business: Was he that bad of a bookmaker?
Wagner: Well, I think we are better – but, he was fine. The increase doesn’t come from us beating his players more than he did. The increase comes from the product being available 24 hours a day and our company having more offerings. And, the secret ingredient helps too!

Sports Betting Business: Oh, I see. Instead of having his cell on for a few hours a day, you put the product in front of his clients all day and night for the buying. I see. But, what is the ‘secret ingredient’?
Wagner: The availability will make a person 100% more over night. But, the secret ingredient made the difference for Mr. Harris. And, the secret ingredient is TIME.

Sports Betting Business: OK, I am lost again. What does that mean?
Wagner: Paul Harris used to set lines, answer phone calls, watch games, grade wagers and keep figures. He did it so well, and so often he already lost his girlfriend. He simply spent all of his time operating as a bookie. When we took over, he became a RECRUITER. All Paul does now is sit in bars and recruit business. We write the business, and Paul grows his business daily. That is the secret ingredient.

Sports Betting Business: Ah, I see. He no longer spends time on the items that cost time and money…he spends his time on what makes him money – recruiting players!
Wagner: Christ! Your brilliant! It typically takes a bookmaker or agent about 3 months to figure that out. But, you, like Paul Harris already figured it out. You recruit, and I will beat them. Therefore the only thing that matters is getting more clients. The more you get, the more you will make.

Sports Betting Business: Of course I get it. That’s the Las Vegas model. Build in an edge and then find as many people to bet against the edge as possible. It’s simple!
Wagner: You sure make it sound simple. But, most bookies are stuck in the mode of trying to ‘beat the players’. The fact is that all you have to do is get a new client or two every day and you will be a millionaire over night. Simply ask Paul Harris who is on campus at University Of Texas recruiting players on 6th Street right now.

Sports Betting Business: So how does a bookmaker or agent learn more about this?
Wagner: Simply go to any of the V.O. Group, S.A. sites to learn more. I suggest starting at www.BettorsNet.com or calling 1-877-512-1001 or emailing agents@bettorsnet.com to get more information.

How Much Should a Bookmaker Expect to Make with PPH

2005-06-16T13:29:00.000-07:00

Q & A with
Dalton Wagner, Founder V.O. Group, S.A.

Dalton Wagner, Proprietor of V.O. Group, S.A. has recently found himself one of the most popular interviews in the offshore sportsbook industry. The reason is his new Price Per Head Agency offering for Onshore Credit Bookmakers. We had an opportunity to catch up with Dalton Wagner and discuss the issue on his last visit to St. Kitts.

Vegas Offshore: Last time we talked you mentioned that your Price Per Head offering was growing beyond your wildest dreams. How is it doing now?
Wagner: Our Price Per Head software allows onshore bookmakers to make more than they have ever dreamed. For this reason, the product is growing faster than we could have ever imagined.

Vegas Offshore: Can you be a little more explicit?
Wagner: We wrote the software offering with input from two bookmakers 24 months ago. Today, we have over 500 bookmakers using the software writing business for over 20,000 clients a week, generating over $500,000 in revenue on a weekly basis in football season.

Vegas Offshore: That is incredible. For you to grow that fast, bookmakers must simply love the product.
Wagner: Yes, bookmakers love the product. Not only does it take the burden of answering phone calls and doing accounting away, but it also allows bookmakers to insulate themselves from legal issues and market their product more aggressively. We have literally 50+ examples bookmakers that are writing hundreds of clients with our software. Can you imagine how they could have done that without us?

Vegas Offshore: The largest onshore bookmaker I have ever met wrote 300 clients and had a staff of 10 clerks on football Sunday. Who is your largest onshore bookmaker?
Wagner: I obviously can’t reveal identities, but I can tell you that she is writing 750 head a week on average and employees nobody.

Vegas Offshore: You have a bookmaker working with you that writes over 750 head a week, what do you think she makes annually?
Wagner: I know exactly what she makes. Let’s just say it is well over a million a year.

Vegas Offshore: How many bookmakers do you currently work with?
Wagner: We are writing business for over 250 bookmakers now. Some represent groups as small as 2 or 3; and others represent groups as large as 750.

Vegas Offshore: Anything else you would like to add?
Wagner: Sure. If you’re a bookmaker onshore, call us or visit us online to learn how to save money and/or grow your business beyond your wildest dreams. The service is modestly priced, and we have never had a disgruntled user. The value proposition is simply huge for any bookmaker.

Vegas Offshore: So how does a bookmaker learn more about this?
Wagner: Simply go to any of the V.O. Group, S.A. sites to learn more. I suggest starting at www.BettorsNet.com or calling 1-877-512-1001 or emailing agents@bettorsnet.com to get more information.

Innovation In Bookmaking

2005-06-14T08:43:00.000-07:00

Q & A with
Dalton Wagner, Founder V.O. Group, S.A.

As Founder and Proprietor of V.O. Group, S.A., Dalton Wagner has the reputation of ferreting out new markets and attacking while other companies sit on their hindquarters and watch. One market that has always eluded offshore bookmakers is the ‘guy who likes dealing with his credit bookmaker’. Typically this market has been untouchable for the offshore bookmaker, but NOT for Dalton and V.O. Group, S.A. We had an opportunity to catch up with Dalton Wagner on his last visit to Panama.

Vegas Offshore: Tell us about this new product that you are calling your Price Per Head Agency Relationship.
Wagner: As you know, one market that a post-up sportsbook has historically been unable to attack is the huge market of bettors that have been betting with their local guy for years. The reason this market exists is the personal relationship that an onshore bookie has with his clients. Sportsbooks have attempted to educate the player and lure him away from the local bookie, however we have taken another tact. We have partnered with local bookies around the world to make their offering better and more valuable.

Vegas Offshore: So instead of attacking the local guy, you actually approach him with your offer?
Wagner: Exactly. The fact is that the relationship is the last thread holding clients to a local guy. Yes, there is the fact that the local guy extends credit, but with bonuses and promotions we can overcome that offering. The key is the relationship.

Vegas Offshore: I’m confused, please explain.
Wagner: As you know, the onshore bookmaking market has lost steam to the offshore providers like V.O. Group, S.A. The reason is that the offerings that I can provide are far superior to those that any onshore bookmaker can offer. The PPH concept is based on allowing the onshore bookmaker to provide his clients my offerings on a purely anonymous basis via an online automated tool. Rather, by signing up with my PPH service, an onshore bookmaker can compete with all offshore sportsbooks on equal footing offering 24-hour wagering, 365 days a year. All the bookmaker does is use an online system to assign PINs and PASSWORDs and pay me a small fee per week for the service.

Vegas Offshore: Wait a minute. So, this means that a small time operator in Cleveland, Ohio can offer all the same things that V.O. Group, S.A. does without making the multi-million dollar investment?
Wagner: Now your catching on. For as little as $25 per head per week, a guy in Ohio can compete with the biggest sportsbooks in the world. And, better yet, never answer another wagering phone call again.

Vegas Offshore: OK, I get it. They use your automated tool to drive their clients to your software online, and they simply pick up the figures at the end of the week.
Wagner: Exactly correct with one exception. Their clients can bet on the phone via our call center staffed with 250 people OR bet online. Again, they have access to all of the offerings V.O. Group, S.A. provides including call center access, client services and technical support.

Vegas Offshore: Is the PPH product only based on sports wagering?
Wagner: Very good question. The answer is no. We actually have the ability to offer clients two (2) casino products including blackjack, baccarat, slots, 3 card poker, roulette, etc. And, clients may also bet on horses at over 350 tracks worldwide.

Vegas Offshore: Wow! That truly is incredible. So, a bookie that has one client can compete with the likes of MVPsportsbook.com on equal footing?
Wagner: Exactly correct. But, there are two keys to the success of the product launch. The first is being able to compete head to head with the biggest sportsbooks in the world. But, arguably the most important factor is that since the onshore bookie no longer has to answer phones or track plays, he can spend all of his time picking up new clients, NOT writing bets.

Vegas Offshore: I remember that in a prior interview you had stated that the key is not winning, but finding more losers.
Wagner: An embarrassing quote, but true. Amateur bookmakers think that the key is to beat the clients you have. That is a huge mistake. The key to getting rich as a bookmaker is finding a new client every single day of the year. The numbers take care of themselves, the key is to get more and more people playing. That is the TRUE BEAUTY of the PPH software and agency relationship.

Vegas Offshore: Do you have any success stories to tell?
Wagner: I have many, but will tell you one that emphasizes what the PPH program can do. I have a bookmaker that has been working out of San Antonio, Texas for 10 years. The guy has historically made $200,000 a year from his 25 or 30 players. He was introduced to our product in July 2004, and today he is writing over 185 clients. The ability to outsource answering calls and writing tickets to us has allowed him to grow his business 600% and his revenue over 1000%.

Vegas Offshore: Why did his revenue grow more than his business? Is it because your lines are sharper than his?
Wagner: There you go again. That is the obvious mistake that everyone makes. Again, don’t worry about beating the clients. Worry about getting more clients to play. But, to answer your question directly, the ability for his clients to play in the casino, play 24 hours a day, play games/halves/quarters, props, etc. is where the extra revenue comes from. Our lines might be sharper, but the advantage is more clients betting more often.

Vegas Offshore: How many bookmakers do you currently work with?
Wagner: We are writing business for over 250 bookmakers now. Some represent groups as small as 2 or 3; and others represent groups as large as 500. Again, remember the key is finding as many clients as possible. If you get to where you are writing 15 clients or more, it is almost impossible to lose for an extended period of time.

Vegas Offshore: So how does a bookmaker learn more about this?
Wagner: Simply go to any of the V.O. Group, S.A. sites to learn more. I suggest starting at www.BettorsNet.com or calling 1-877-512-1001 or emailing agents@bettorsnet.com to get more information.

State of the Offshore Gaming Industry

2005-06-13T09:53:00.000-07:00

Q & A with
Dalton Wagner, Founder V.O. Group, S.A.

As Founder and Proprietor of V.O. Group, S.A., Dalton Wagner has the reputation of making one-dollar work like three in online marketing. With formidable competition, Mr. Wagner founded V.O. Group, S.A. in 1998 and has quickly risen to be one of the largest offshore operators in the world with over 50,000 active clients. We had an opportunity to catch up with Dalton Wagner on his last visit to Antigua.

Vegas Offshore: This year has been interesting for V.O. Group, S.A. and the offshore industry as a whole (advertising changes, etc.). How do you feel about the industry at the current time?
Wagner: I feel the industry is as strong as it has ever been. We are still finding that our advertising is bringing us new clients at an ever-growing rate. Some competitors are complaining that the crimp on advertising is affecting the industry, but we are not seeing this trend. Consolidation within the industry has begun, and I am sure you will continue to see smaller competitors gobbled up, and medium to larger companies merging. The weaker will go by the wayside.

Vegas Offshore: What about the rumors of V.O. Group, S.A. being bought?
Wagner: I am glad you used the term 'rumors'. The fact is that in July 2004 three suitors approached V.O. Group, S.A. because of our record-breaking profit year. In two of the cases we did see 'value' in a merger/relationship. However, at the end of the day, we could not come to terms that were agreeable on both sides. At this time V.O. Group, S.A. stands independent. And, we are happy being independent, profitable and growing.

Vegas Offshore: It is rumored that Bet On Sports (PLC; BSS.L) was one of the suitors. Is that true?
Wagner: There were three suitors that approached V.O. Group, S.A. And, in all cases confidentiality agreements were signed. For this reason, I cannot confirm or deny who any of the potential suitors were. What I can tell you is that one was a major sportsbook player, one was a major casino player and two were PLCs. I doubt that helps.

Vegas Offshore: Well, if Bet On Sports was involved, I bet you are happy you weren't involved in their November 24, 2004 stock debacle.
Wagner: The Bet On Sports stock debacle was an over-reaction by the market in my opinion. In one day you saw their stock drop 50% in value. And, the reason was poor performance in a very short period. If you looked at the other publicly held offshore gaming companies, you saw the same thing, just not as drastic as with BSS.L. Simply put, I know the founder of Bet On Sports, I know the management of Bet On Sports, I know the staff at Bet On Sports and I would not hesitate to invest in Bet On Sports. They were, and still are, undervalued in my opinion. They are a great competitor and a great company. I wish I had some of their stock at the new adjusted price. I just don't see how you can lose.

Vegas Offshore: It is rare to hear the 'competition' speaking so highly of a major competitor. Why would you do so?
Wagner: Simple. Bet On Sports is a great company. And, I am sure they would tell you the same about V.O. Group, S.A. The fact is, that there are companies out there that make our industry stronger. And, ones that make it weaker. Bet On Sports and V.O.Group, S.A. strive to make our industry stronger and more legitimate on a daily basis (as do BetCris.com; BoDog.com; etc.). What kind of ambassador for the industry would I be if I downplayed a 'good' competitor like Bet On Sports for my own benefit?

Vegas Offshore: If Bet On Sports did approach you, would you consider a merger?
Wagner: I like being independent. But, simply put, NEVER say NEVER.

Vegas Offshore: What of the rumors that you are looking at Panama as a potential relocation site?
Wagner: Our operation currently has offices in Costa Rica, Antigua and Belize. In Costa Rica we currently have over 400 employees. With employment costs in Costa Rica on the rise, the burden of Caja and the ever-changing political climate, we have looked at several alternative locations. One of the most desirable is/was Panama. However, due to the massive failure of BetPanAm.com and the political debacle involving their past gaming commission, we have decided to stay put for the current time. I think it is very safe to say that our primary location will be Costa Rica for the next 5 years. Pending any crazy licensing or governmental changes.

Vegas Offshore: I have visited your office in Antigua, however had no idea that you employed 400 individuals in Costa Rica. Why do you need such a large staff?
Wagner: Our industry has a reputation for being lazy and getting by on a shoestring budget. Most sportsbooks, casinos, racebooks and poker rooms answer phones when it is convenient and buy computers when the ones they are operating blow-up. Most are simply embarrassing to the legitimate operators. To run an operation like ours, servicing 50,000 bettors, you need lots of space, a 1st rate phone system, fiber and satellite phone and Internet backups, IT professionals, accounting professionals, etc. V.O. Group, S.A. is currently located in 35,000 square feet of space, operating 24-hours a day, 365 days a year, and we are busting at the seems. At the current time, we are negotiating to buy the building next to us for future expansion. I think we will employ 600 people come this same time next year.

Vegas Offshore: I have noticed that you are still branching into other products. How has this strategy worked for V.O. Group, S.A.?
Wagner: In 1998 we started with our sports product. We then branched into the casino product. In 2001 and 2002 we opened several major race books and poker rooms. The fact is that we are profitable in all products. However, our core product, and hub for our advertising, is focused on the sports product. The strategy has worked, but we will not desert our core product. You can look for our re-launch of interactive betting in early 2005 and bingo in mid 2005.

Vegas Offshore: What other changes do you see in 2005 for V.O. Group, S.A.
Wagner: I think you will see V.O. Group, S.A. work more on our Internet interface to make it more gamer-friendly; you will see our company move into land-based gaming; and you will see us acquire two or three smaller competitors. Rather, you will see what you have always seen, expansion.

Vegas Offshore: What other changes do you see in 2005 for the industry as a whole?
Wagner: The strong will get stronger. The weak will get weaker. And, the acquisition race will be on.

$1 000 000 March Madness brackets contest

2005-03-02T09:04:00.000-08:00

Pick the winners for all March Madness games and win $1 000 000.
Popular Poker Sportsbook is once again sponsoring this contest that is fast becoming a tradition in the online gambling industry. Once you sign up with Popular Poker Sportsbook you get a free entry to the contest. All you need to do is fill out your bracket for the whole tournament. Once the Madness is over you are awarded prizes according to the amount of winners you picked.
There is a $1 000 000 grand prize for picking all winners, but there are also thousands of dollars for other winners. It all comes down to how many winners you picked.
But you need to hurry up, all entries for the contest need to filled out before March 17, the day before the action starts for March Madness 2005.
For more information on this and other great opportunities visit Popular Poker Sportsbook or call 1 800 653 9774 .
Popular Poker Sportsbook is an established online gambling operation, with hundreds of years of combined experienced, topping the charts in regards to service quality, variety and accuracy. But their most important feature, and this can be ratified by thousands of satisfied customers, is financial stability.

Make Your Picks Now

Online Super bowl 2005 betting tips

2005-02-03T07:48:00.000-08:00

Super bowl betting tips
Rule NO. 1 Look at the entire season
NFL gamblers seem to place too much weight on a team's recent performance while placing too little on its long term record, according to a 1997 study entitled " Testing Market Efficiency: Evidence from the NFL Sports Betting Market" suggests you look at a team entire season rather than worrying about what team is hot or not , this weeek. So stop getting caught up on a team's recent blunders. Consider the bigger picture when you place your bets.
Super Bowl betting Tips
Rule no. 2: Bet the web
Focus your efforts on the newer betting exchanges, such as those on the internet, that are reducing the "vig" -the charge taken on bets, "Finance, behavioral economics and Sports Betting"
Super Bowl Betting Tips
Rule no 3: Bet the Underdog
According " Market Efficiency and a Profitable Betting Rule", underdogs of more than seven points win frequently enough to produce statistically significant profits. Oddsmakers recognize that most people bet on favorites and subsecuently shade the betting line so that favorites win slightly less than half of the games against the spread. So Take a chance on the big underdogs.
Super Bowl Betting TipsRule No. 4: Wager the "Under"
When NFL Bettors get involved in totals bets ( betting on whether the combined points scored by both teams will come in over or under a level set by the bookmaker), they tend to overbet the "over" for those games where the over/under is 47.5 or greater. These outhors theorize that bettor psychology may drive totals too high, making the "under" bet a consistently profitable one.
Super Bowl Betting TipsRule No. 5. Watch your Wallet
Cardinal Rule: Dont't bet anything you cannot afford to lose!
wagering wager super bowl odds online spread betting sports betting bet sports online bet beting online betting games bowl online betting nfl betting future football wagering

Road to the online superbowl betting will be a Wide-Open Race NFL Sure, no one will argue that the AFC was a better conference than the NFC.

2005-01-21T13:57:00.000-08:00

Road to the online superbowl betting will be a Wide-Open Race NFL Sure, no one will argue that the AFC was a better conference than the NFC. It is also safe to say that there isnt as big of discrepencies between the #1 and #6 seeds as there have been in recent years. This should provide some good drama this postseason. Personally, I like a Pittsburgh-Green Bay matchup for this year's online superbowl betting in Jacksonville. Reason I like Pittsburgh is that they have been the most consistent team in the NFL. It helps them not having to play the Indianapolis Colts and the deadly Peyton Manning in the Divisional playoffs. Plus, if you are Steelers' fan, you have to love the fact that New England has had a a rash of injuries in their secondary. Before you start cursing me for my selection of the Packers, keep in mind they wont have to play Philadelphia in the divisonal playoffs, but rather Atlanta. Michael Vick still has problems grasping the middle and deep routes of the west-coast offense. I truly believe Brett Favre retires after this season so just imagine how focused he will be in the next month. It always helps that Terrell Owens will be out of action.

Why are online superbowl betting Markets Organized so Differently than NFL?

2005-01-13T08:43:00.000-08:00

Why are online superbowl betting Markets Organized so Differently than NFL?
--------------------------------------------------------------------------------Page 2 AbstractThe market for superbowl online superbowl betting is structured very differently than the typical financial market. In superbowl betting, bookmakers announce a price, after which adjustments are small and infrequent. Bookmakers do not play the traditional role of market makers matching buyers and sellers, but rather, take large positions with respect to the outcome of game. Using a unique data set, I demonstrate that this peculiar price-setting mechanism allows bookmakers to achieve substantially higher profits. Bookmakers are more skilled at predicting the outcomes of games than bettors and systematically exploit bettor biases by choosing prices that deviate from the market clearing price. JEL codes: G14, L20--------------------------------------------------------------------------------Page 3 There are many parallels between trading in NFL and superbowl wagering. First, in both settings, investors with heterogeneous beliefs and information seek to profit through trading as uncertainty is resolved over time. Second, superbowl betting, like trading in financial derivatives, is a zero-sum game with one trader on each side of the transaction. Finally, large amounts of money are potentially at stake. The four major British bookmaking firms report turnover of almost £10 billion in 2002, and estimates of wagering on sproring events in the United States go as high as $380 billion annually (National online superbowl betting Impact Study Commission 1999). In light of these similarities, it is surprising that these two types of markets are organized so differently. In most NFL, prices change frequently. The prevailing price is that which equilibrates supply and demand. The primary role of market makers is to match buyers with sellers. With superbowl wagering and also horse racing in the United Kingdom, market makers (i.e. bookmakers) simply announce a “price” (which can be odds to win a horse race, or for sporting events can be odds to win a game or a point spread, e.g., the home team to win an Amerian football game by at least 3.5 points), after which adjustments are typically small and relatively infrequent.1If that price is not the market clearing price, then the bookmakers may be exposed to substantial risk.2If bettors are able to recognize and exploit mispricing on the part of 1In my data on American football, in the five days preceding a game, the posted price changes an average of 1.4 times per game. When the price does change, in 85 percent of the cases the line moves by the minimum increment of one-half of a point. Thus, the posted spread on Tuesday is within one point of the posted spread at kickoff on Sunday in 90 percent of all games. These calculations are based on information on changes in the casino lines reported at www.mvpsportsbook.com. In horse racing, the odds set by bookmakers change more frequently. 2 There are many notorious examples of bookmakers suffering large losses. It is reported that --------------------------------------------------------------------------------Page 4 the bookmaker, the bookmaker can sustain large losses. The risk borne by bookmakers on superbowl betting is categorically different than the casino’s risk on other games of chance such as roulette, keno, or slot machines. In those games of chance, the odds are stacked in favor of the casino and the law of large numbers dictates profits for the house. In contrast, however, if the bookmaker sets the wrong line on sporting events, it can lose money, even in the long run. The presence of a small number of bettors whose skills allow them to achieve positive expected profits could prove financially disastrous to the bookmaker. Such bettors could either amass large bankrolls, or in the presence of credit constraints, sell their information to others. Although the mechanism used for price-setting in superbowl betting seems peculiar, there are at least three scenarios in which the bookmaker can sustain profits implementing it. In the first scenario, bookmakers are extremely good at determining in advance the price which equalizes the quantity of money wagered on each side of the bet. If this occurs, the bookmaker makes money regardless of who wins the game since the bookmaker charges a commission (known as the “vig”) on bets.3Following this strategy, bookmakers do not have to have any particular skill in picking the actual outcome of sporting events, they simply need to be good at forecasting how bettors behave. Popular depictions of bookmaker behavior have stressed this explanation.4over half of all British bookmakers were bankrupted when Airborne won the Epsom Derby in 1946 (the first running after World War II ended) at odds of 50 to 1 (Smith 2002). In 1996, popular jockey Frankie Dettori won seven straight races, costing bookmakers an estimated £30 million (online superbowl betting Magazine 1999). Coral Eurobet reported losses of ₤12 million on internet betting in quarter-final round of Euro 2000 soccer championship (Smith 2002). 3 Typically, bettors must pay the casino 110 units if a bet loses, but are paid only 100 units if the bet wins.4For instance, a website devoted to educating novice gamblers (www.mvpsportsbook.com) writes, “A superbowl bettor needs to realize that the point spread on a game is NOT a prediction by an oddsmaker on the outcome of a game. Rather, the odds are designed so that equal money is bet --------------------------------------------------------------------------------Page 5 An alternative scenario under which this price-setting mechanism could persist is if bookmakers are systematically better than gamblers in predicting the outcomes of games. If that were the case, the bookmaker could set the “correct” price (i.e. the one which equalizes the probability that a bet placed on either side of a wager is a winner). Although the money bet on any individual game would not be equalized, on average the bookmaker will earn the amount of the commission charged to the bettors. Unlike the first scenario above, however, if prices are set in this manner, the bookmaker will lose if gamblers are actually more skilled in determining the outcome of games than is the bookmaker. The third possible scenario combines elements of the two situations described in the preceding paragraphs. If bookmakers are not only better at predicting game outcomes, but also proficient at predicting bettors preferences, they can do even better in expectation than to simply collect the commission. By systematically setting the “wrong” prices in a manner that takes advantage of bettor preferences, bookmakers can increase profits. For instance, if bookmakers know that local bettors prefer local teams, they can skew the odds against the local team. There are constraints on the magnitude of this distortion, however, since bettors who know the “correct” price can generate positive returns if the posted price deviates too much from the true odds.5on both sides of the game. If more money is bet on one of the teams, the superbowl book runs the risk of losing money if that team were to win. Bookmakers are not gamblers--they want to make money on every bet regardless of the outcome of the game.” Similarly, Lee and Smith (forthcoming) write, “Bookies do not want their profits to depend on the outcome of the game. Their objective is to set the point spread to equalize the number of dollars wagered on each teamand to set the total line to equalize the number of dollars wagered over and under. If they achieve this objective, then the losers pay the winners $10 and pay the bookmaker $1, no matter how the game turns out. This $1 profit (the ‘vigorish’) presumably compensates bookmakers for making a market and for the risk they bear that the point spread or total line may be set incorrectly.” 5This assumes that bookmakers are unable to offer different prices to different bettors. --------------------------------------------------------------------------------Page 6 In this paper, I attempt to better understand the structure of the market for superbowl online superbowl betting by exploiting a data set of approximately 20,000 wagers on the National Football League, the premier league of American football placed by 285 bettors at an online superbowl book as part of a high-stakes handicapping contest. Two aspects of this data set are unique. First, in contrast to previous studies of betting that only had information on prices, I observe both prices and quantities of bets placed. That information allows me to determine whether the bookmaker appears to be equalizing the amount of bets on each side of a wager. Second, I am able to track the behavior of individual bettors over time, which provides a means of determining whether some bettors are more skillful than others. Although my data are from one bookmaker, the patterns observed here are likely to generalize since all bookmakers offer nearly identical spreads on a given game. A number of results emerge from the analysis. First, I demonstrate that the bookmaker does not appear to be trying to set prices to equalize the amount of money bet on either side of a wager. In almost one-half of all games, at least two-thirds of the bets fall on one side of the gamble. Moreover, the spread chosen systematically fails to incorporate readily available information (e.g. which team is the home team) that would help in equalizing the money bet on either side of a wager.6For instance, in games where the home team is an underdog, on averageIndeed, there is evidence that local bookmakers who deal repeatedly with the same clients are able to exercise some degree of price discrimination. See Strumpf (2002) for empirical evidence that bookmakers both shade the odds against the home team and offer different odds to bettors with different past betting histories. 6Of course, it is not the number of bets on either side of a wager that the bookmaker wants to equalize, but rather, the total dollars bet on either side. In my data, however, all wagers are constrained to have the same dollar value, so the two are equivalent. If there are large bankroll bettors outside my sample who systematically bet against the prevailing sentiment of --------------------------------------------------------------------------------Page 7 two-thirds of the wagers are on the visiting team. These findings argue strongly against the first scenario presented above and the popular depictions of bookmaker behavior. A rationale for this failure to equalize the money emerges in the paper’s second finding: bookmakers appear to be strategically setting prices in order to exploit bettors’ biases, just as DellaVigna and Malmendier (2003) demonstrate health clubs do with their clients. Bettors exhibit a systematic bias toward favorites, and to a lesser extent, towards visiting teams.7Consequently, the bookmakers are able to set odds such that favorites and home teams win less than fifty percent of the time, yet attract more than half of the betting action. By choosing these prices, it appears bookmakers increase their gross profit margins by 20-30 percent over a price-setting policy that attempts to balance the amount of money on either side of the wager. On the dimension of favorites versus underdogs, bookmakers appear to have distorted prices as much as possible without allowing a simple strategy of always betting on underdogs to become profitable. The fact that home teams and underdogs cover the spread a disproportionate share of the time has been well established in the literature (e.g. Golec and Tamarkin 1991, Gray and Gray 1997). My findings provide an explanation for that empirical regularity: it is profit maximizing for the bookmaker who sets the spread. Third, there is little evidence that there are individual bettors who are able to systematically beat the bookmaker.8The distribution of outcomes across bettors other bettors, conclusions based on my sub-sample may be erroneous. 7This finding is not to be confused with the bias towards longshots that has been observed in parimutuel horse-race betting (e.g. Ali 1977, Golec and Tamarkin 1998, Jullien and Salanie 2000, Shin 1991, 1992, 1993, Thaler and Ziemba 1988, Vaughan Williams and Paton 1997). In football, the odds are set to make the chance of each team covering the spread about fifty percent, making this consideration irrelevant. 8While only tangentially related to the issues addressed in this paper, it is worth noting --------------------------------------------------------------------------------Page 8 is consistent with data randomly generated from independent tosses of a 50-50 coin. Moreover, how well a bettor has done up to a certain point in time has no predictive value for future performance. Finally, the evidence is mixed as to whether aggregating bettor preferences has any predictive value in helping to beat the spread. In my sample, there is some weak and ultimately statistically insignificant, evidence that bettors are more likely to predict correctly when there is agreement among them as to which team looks attractive. Altogether, the results are consistent with the conclusion that the bookmakers are at least as good at predicting the outcomes of games than are even the most skilled gamblers in the sample, and the bookmakers exploit their advantage by strategically setting prices to achieve profits that are likely higher than would be possible if they simply acted as market makers letting supply and demand equilibrate prices. The remainder of the paper is organized as follows. Section II provides somebackground on wagering on professional football in the United States. Section III describes the data set used in the paper. Section IV presents the empirical findings. Section V concludes. Section II: Background on American football wagering American football is the most popular sport for wagering in the United States, generating 40 percent of superbowl-betting revenue for legal bookmakers in Nevada (Nevada Gaming Control Board 2002). USA Today reports that half of all Americans have a wager on the outcome of the Super Bowl. The most common type of bet in pro football involves picking the winner of a that there is an extensive academic literature devoted to the question of testing for market efficiency in wagering markets (e.g. Asch, Malkiel, and Quandt 1981, Sauer et al. 1988, Woodland and Woodland (1994), Zuber, Gandar, and Bowers 1985). --------------------------------------------------------------------------------Page 9 game against a point spread (a so-called “straight bet”).9For instance, if the casino posts a betting line with the home team favored by 3 points, a bettor can choose either (1) the home teamto win by more than that amount, or (2) the visiting team to either lose by less than three points or to win outright. In the event the game ends exactly on the point spread, all bets are refunded. Regardless of which team is chosen, the bettor typically pays the casino 110 units if they lose the bet and collects 100 units when victorious. The difference in the amount paid on a loss versus the amount won for a victory is the casino’s commission, known as the “vig.” Because the bettor can take either side of the wager at the same payout rate, the casino needs to pick a betting line that roughly equalizes the probability of the two events occurring (in this case home teamwinning by more than three or more points, or failing to do so). The bettor receives the spread in force at the time a wager is placed, regardless of later adjustments made by the bookmaker. Define terms as follows: p is the probability that the favorite wins a particular game, f is the fraction of the total dollars bet on the game that go to the favorite, and v is the vig or commission charged by the bookmaker, which is paid only on losing bets.10The bookmaker’s 9There are many other types of bets available. For instance, one can bet on whether the total number of points scored in a game is above or below a certain level. One can also bet on which team will win the game (not against the spread), with the payouts appropriately adjusted to reflect the probability of these outcomes. It is also possible to parlay bets on a series of games such that the bettor receives a large payout if correctly picking all the games and receives zero otherwise. Because my data only covers straight bets, I do not focus on these other bet types. Woodland and Woodland (1991) argue that the use of point spreads as opposed to odds that depend only on which teams wins or loses is a bookmaker profit-maximizing response to risk aversion on the part of bettors. 10I define the model in terms of favorites and underdogs simply because this is the most salient dimension empirically. Game outcomes could be characterized along any relevant set of dimensions. --------------------------------------------------------------------------------Page 10 expected gross profit per unit bet11is given by (1)E[Bookmaker profit] = [(1-p)f)+p(1-f)](1+v)- [(1-p)(1-f)+pf] The terms inside the left set of brackets is the fraction of dollars bet in which the bookmaker wins. That amount is multiplied by 1+v to reflect the bookmakers vig. The terms in the right set of brackets are the cases in which the bookmaker loses and has to payout to the bettor. Rearranging terms, equation (1) simplifies to (2)E[Bookmaker profit] = (2+v)(f+p-2pf)-1 11For simplicity, I treat the number of wagers placed as fixed in the analysis. To the extent that changing the odds affects the total volume of bets, the bookmaker’s overall gross profit would be a function of both the number of bets and the gross profit per unit bet. If either the probability the two teams win is equal (p=.5) or the money bet on both teams is equal (f=.5), the bookmakers gross profit simplifies to v/2. In either of these instances, the bookmaker is indifferent about the outcome of the game and earns a profit proportional to the size of the commission charged. As noted in the introduction, therefore, the bookmaker does not need to be able to predict the outcome of the games more accurately than the bettors to ensure a profit. The bookmaker just needs to be able to predict better preferences so as to balance out the money on each side of the wager. --------------------------------------------------------------------------------Page 11 Equations (1) and (2) take f and p as given. Of course, the fraction of money bet on the favorite will be a function of the probability the favorite actually wins, i.e. f=f(p), with ∂f/∂p>0.12Taking the derivative of (2) with respect to p, an optimizing bookmaker will set p such that (3)[1-2f(p)] + (1-2p)∂f/∂p = 0The term in square brackets is the benefit the bookmaker would achieve from distorting the odds if gamblers did not respond to changes in prices. The remaining term captures the impact on profits of the behavioral response of bettors who switch towards the team with bettor odds. Note that if bettors preferences are unbiased in the sense that f(p=.5)=.5, then the bookmaker’s optimum is to choose p=.5, which implies f=.5 as well. If, on the other hand, bettors preferences are biased so that f(p=.5)>.5, as is true empirically in the data set, then the bookmaker can increase profits by reducing p below .5 (Kuypers 2000 makes this same point).13Intuitively, if bettors prefer favorites at fair odds, the bookmaker can offer odds slightly worse than fair on favorites and still attract more than half of the wagers on the favorite, yielding profits that are strictly higher than is the case at p=.5. Mathematically, at p=.5, the term in square brackets in equation (3) is negative, but the other term on the left-hand-side is equal to zero, demonstrating that the bookmaker is not at an optimum. The bookmaker will not want to push p too far away from .5 for two reasons, 12Throughout this analysis, I treat v (the commission charged by the bookmaker) as a parameter rather than a decision variable for the bookmaker. Commissions are virtually always 10 percent. Gaining a better understanding of the reasons for the uniformity of commissions across bookmakers and over time presents an interesting puzzle for future research. 13Although I use the term “bias” to describe bettor preferences, I do not necessarily imply irrationality on their part. If there is more consumption value associated with betting on favorites, then bettor preferences for favorites could be completely rational. --------------------------------------------------------------------------------Page 12 however. First, as p diverges from .5, it becomes increasingly costly to the bookmaker when a bettor switches from the favorite to the underdog. This is because the bet on the favorite is at (increasingly) unfair odds, whereas the bet on the underdog is at (increasingly) better than fair odds. Note that when p is lowered to the point where f(p)=.5, gross profits are back to the level attained when p=.5. Thus, the bookmaker would never want to distort prices to that point. The second reason that the bookmaker cannot distort prices too much is that if some subset of bettors do not have biased preferences, those bettors can exploit the distorted prices. With the standard vig, a bettor must win 52.4 percent of bets to make profit.14One could imagine that the volume of capital available to bettors with positive expected profits could be enormous, both because their bankrolls would grow over time and because the availability of such profits would attract new investors. Thus, it would be surprising to observe price distortions so large that simple strategies (e.g. always bet the underdog) could yield a positive profit. The discussion above assumes that bookmakers have some market power. In a perfectly competitive market, ∂f/∂p will be near infinity and competition will drive the spread back to the point where f(p)=.5 and no excess profits are obtained by bookmakers. Empirically, competitive pressure does not appear to be strong enough to eliminate excess profits. Understanding why this is the case is an important unanswered question of this research. III. The data set The data used in this paper are wagers placed by bettors as part of a handicapping contest 14A bettor breaks even when p-(1+v)(1-p)=0. The solution to that expression is p=.5238. --------------------------------------------------------------------------------Page 13 offered at an online superbowlbook during the 2001-2002 NFL season. In the contest, bettors were required to pick five games per week against the spread for each of the seventeen weeks of the NFL regular season (a total of 85 games). Bettors could choose those five games from any of the 13-15 games being played in a given week. One point was given for each correct pick, and one-half point if a game ended exactly on the spread. The entry fee was $250 per person, and there were 285 entrants. All of the entry fees were returned as prize money, so participants were competing for a total pool of $71,250.15Sixty percent (or $42,750) went to the bettor with the most correct picks. Second through fifth place finishers received declining shares of the pool. The last-place finisher (conditional on making 85 picks) received five percent of the pool. In the data, I observe the ID number of each bettor, all wagers placed as part of this contest, the spread at which the bet was placed, and the outcome of the game. There are a number of potential shortcomings with these data. First, these are not wagers in the traditional sense. The bettor does not receive a direct payoff from winning any particular game; the payoff is only based on the cumulative number of wins. Nonetheless, the presence of large monetary rewards to the winners provides strong incentives to the participants. Presumably, the picks made by bettors in the contest closely parallel the actual betting wagers they were making; anecdotally that is true among the contest participants known to the author.1615The apparent purpose of the contest was to ensure that the bettors had a reason to visit the website each week. The fact that the worst-place finisher (conditional on having bet each week) received a payoff, reinforces this point. 16One important way in which the contest wagers might be expected to differ fromactual wagers placed is that there is an incentive in the contest to pick outcomes that the bettor believes will be unpopular with other gamblers. That is because of the tournament structure of the contest in which rewards are great in the extreme right-hand tail, but no differentiation is made elsewhere in the distribution. --------------------------------------------------------------------------------Page 14 Second, the nature of the data make it impossible to ascertain the intensity of preferences across games since all selections receive equal weighting. Bettors may have much stronger preferences for their most favored pick of the week than would be the case for the fifth-favorite pick. Third, there is substantial attrition in the sample. Of the 285 bettors who entered the contest, 100 (a little more than one-third) made their entries all 17 weeks of the season. More than 60 percent participated in at least 15 of the 17 weeks of the season. Less than 10 percent of the contestants recorded data for fewer than eight weeks. For bettors in the middle of the pack as the season progresses, the incentive to continue participating decreases substantially. Bettors who miss a week receive zero points, greatly reducing their likelihood of winning the contest, and disqualifying themselves from eligibility for the last-place prize. It should be noted, however, that attrition in this context adversely affects my ability to test only one of the hypotheses: what the overall distribution of bettor success rates looks like. Because attrition is non-random, the set of bettors who continue to the end will be skewed. Fourth, the spreads used in the handicapping contest are fixed on the Tuesday preceding the game and do not fluctuate with the actual spread, even though a bettor’s contest picks are not due until the Friday before the game. As a consequence, in some games, the actual spread and the contest spread differ at the time an entry is made. All of the results of the paper, however, are robust to dropping games in which there are substantial fluctuations in the spread between Tuesday and Friday. Finally, these data are not the universe of bets placed at the superbowl book (although they are the universe of bets in this contest), much less at superbowl books in general. Nor, as discussed below, are the bettors who participated in the contest likely to be a random subset of all bettors. On the other hand, these data do offer enormous advantages over that which is typically --------------------------------------------------------------------------------Page 15 available. Virtually all previous analyses of superbowl wagering have focused exclusively on prices, but have not had access to any information about the quantity of bets on each side of the wager (see, for example, Avery and Chevalier 1999, Golec and Tamarkin 1991, Gray and Gray 1997, Kuypers 2000).17In my sample, there are a total of 19,770 bets in the data set covering 242 different games. An average of 80.5 different bettors make a selection on a game, with the minimum and maximum number of bets on a game ranging from 28 to 146. Although the bettors included in my sample are not a random selection of all gamblers, they represent a particularly interesting subset. These bettors are likely to be relatively sophisticated, serious bettors. Because they are betting at an online superbowl book, they are likely to be geographically quite diverse. In signing up for the contest, they are indicating an expectation that they plan to visit (and presumably bet at) the internet superbowl book every week of the season. In addition, to the extent there are differences in skill across bettors, this contest should attract the most skillful players because it rewards exceptional long-run performance. Section IV: Empirical Results 17In his innovative work, Strumpf (2002) does have actual bets based on seized bookmaker records. Although his data are extremely informative on a number of questions, they are less than ideal for the questions posed in this paper both because they cover a short time period and are geographically localized. I begin the analysis by addressing the issue of whether the spread is set so as to equalize the wagers on either side, as well as testing the predictions of a model of profit-maximizing --------------------------------------------------------------------------------Page 16 bookmakers who exploit biases on the part of bettors. The analysis then turns to the question of whether bettors differ in their skill at picking winners. Finally, I examine whether aggregating information across bettors provides any valuable information. How are prices set?The first question addressed is whether bookmakers set prices so as to equalize the amount of money on either side of the wager. Although I have data for only one bookmaker, it is important to note that the prices (i.e. spreads) offered by this superbowl book are virtually identical to those at any bookmaker online or at Las Vegas casinos. Thus, in practice individual bookmakers are not actively setting prices, but rather, following the lead of a handful of influential oddsmakers who are paid by large Las Vegas casinos for their services. Figure I presents a histogram of the fraction of the total wagers placed on the team that bettors most prefer. By definition, this fraction must lie between .5 and 1. If the bookmaker balances bets, these values will be concentrated near .50. In the data, however, this is clearly not the case. In only twenty percent of the games are 50-55 percent of the wagers placed on the preferred team. In the median game, almost two-thirds of the bets fall on one side. In almost 10 percent of the games, more than 80 percent of the bets go one direction. The dispersion in Figure I is not simply the result of sampling error due to the fact that I observe only 80 bets per game on average. If bettor choices were independent and each bettor had a fifty percent chance of picking either team, than one would expect the preferred team to garner between 50 and 55 percent of the wagers in nearly two-thirds of the games, compared to only twenty percent in the data. Furthermore, if the sample is divided in half based on the number of bets placed, the dispersion of the fraction of wagers placed on the preferred team is --------------------------------------------------------------------------------Page 17 actually greater in the games with more total bets.18Two alternative hypotheses could explain the failure of the bookmaker to equalize wagers on the two sides of the spread. The first possibility is that the bookmaker would like to balance the bets, but is unable to do so because it is difficult to accurately predict what teambettors will prefer. The second hypothesis is that balancing the wagers is not the objective of the bookmaker. Indeed, as demonstrated earlier, if bettors exhibit systematic biases, a profit maximizing bookmaker does not want to equalize the money bet on both sides. Rather, the bookmaker intentionally skews the odds such that the preferred team attracts more wagers, but wins less than half of the time. If bookmakers are attempting to balance the money bet on each side of the wager, one would expect that observable characteristics of a team or game would have no power in predicting the fraction of bettors preferring that team. Otherwise, the bookmaker could have used that information to set a spread that would have better equalized the distribution of bets. If the bookmaker is attempting to exploit bettor biases by setting skewed odds, however, the opposite is true. Dimensions along which bettors exhibit bias should be systematically positively related to bet shares (and as demonstrated below, will also be systematically negatively related to win percentages). 18Further confirmation of the patterns in Figure I come from data available at www.mvpsportsbook.com. At that website, visitors make hypothetical wagers on game outcomes as part of small-payoff contests. The breakdown of bets on each team is available. The patterns in that data are strikingly similar. --------------------------------------------------------------------------------Page 18 Figures II and III provide some initial evidence on the question of whether observable characteristics are correlated with the distribution of wagers on a game. Figure II presents a histogram of the fraction of bets placed on the home team for the subset of games in which the home team is the favorite.19Since the identity of the home team is readily observable, the histogram in Figure II should be centered around 50 percent if the bookmaker is attempting to equalize bets on either side of the wager. The distribution is clearly skewed to the right, implying that home teams systematically attract more than half of the bets in games in which they are favored. In almost three-quarters of the games, more bets are placed on home favorites than on their opponents. In the median game, roughly 58 percent of the bets go to the home favorite. Figure III is identical to Figure II, except that the sample is games in which the visiting team is favored, and the values in the figure are the fraction of bets placed on the visiting team. In these games, the distribution of bets is even more skewed. In more than 90 percent of games with a visiting favorite, more bets are placed on the visitor than the home team. In the median game in Figure III, two-thirds of the money is wagered on the visitor. Thus, Figures II and III demonstrate quite definitively that the spreads are set such that substantially more than half of19Note that despite the similarity in the word “favorite” and the phrase “bettors’ preferred team,” these are two completely different concepts. “ Favorite” refers to the teamjudged most likely to win the game by the bookmaker. The “bettors’ preferred team,” on the other hand, is the team that the bettors think is most likely to cover the spread. In other words, the bettors are making their choices conditional on the bookmaker already setting a spread that ostensibly equalizes the chance of the favorite and the underdog covering. --------------------------------------------------------------------------------Page 19 the bets are placed on favorites. Table I further explores the issue of whether observable characteristics are correlated with betting patterns. The dependent variable in Table I is the percent of bettors who choose the favorite. The unit of observation is a game. The method of estimation is weighted least squares, with the weights proportional to the total number of bets placed on the game. The first columnof the table demonstrates that, not controlling for anything else, 60.6 percent of the bets accrue to the favorites. The standard error on that point estimate is .009, so the null hypothesis that 50 percent of the bets are placed on favorites is strongly rejected. Column II adds five indicator variables corresponding to which team is favored and by how much (the omitted category is games in which the visitor is favored by more than 6 points). Consistent with the figures presented earlier, home favorites do not attract as high a fraction of the bets as do visiting favorites. These five variables capturing the spread are jointly highly statistically significant, as reported in the bottom of the table. The third column adds dummy variables for each week of the season. These week dummies are also jointly statistically significant at the .01 level. Although not shown individually, the dummies suggest that a greater fraction of the bets are placed on favorites early in the season. Finally, the last column adds thirty-one indicator variables corresponding to each team in the league. These variables take the value of one if a team is favored and -1 if the team is an underdog. Once again, the team variables are highly statistically significant.20The R2in the final column of the table is .483, implying that these 20Based on these estimates, the five teams that attracted the most bettors, controlling for other factors, were Green Bay, Kansas City, New Orleans, Oakland, and San Francisco. The least popular teams were Baltimore, Chicago, Jacksonville, Pittsburgh, and Saint Louis. --------------------------------------------------------------------------------Page 20 observable characteristics explain a great deal of the variation in the fraction of money bet on the favorite. The results in Table I uniformly argue against the hypothesis that bookmakers are doing the best they can to even out the bets on each game, suggesting instead that the imbalance is intentional. If that is true, then the model presented earlier makes strong predictions about the expected pattern of winning percentages for bets of different kinds. In particular, teams with attributes that attract a disproportionate share of the money from bettors (e.g., being the favorite) should cover the spread less than fifty percent of the time. But, the deviation from 50-50 cannot be too large (more than a few percentage points), or bettors who do not suffer from biases can profitably exploit the price distortion. Table II presents evidence consistent with those predictions. The unit of observation in the table is an individual bet. The first three columns capture the fraction of the bets placed on a team; columns 4-6 are the percentage of those wagers in which the bettor correctly picks the winners. Bets are categorized according to whether they are placed on games in which the home team is favored (top row) or the visiting team is favored (second row), and whether the bet is for the favorite or the underdog (the columns in table). Mirroring the results presented earlier, visiting favorites attract a disproportionate share of the bets placed in such games: 68.2 percent of the total.21To a lesser degree, home favorites also attract excess bets (56.1 percent). Shifting focus to columns 4 and 5, the model predicts that a high fraction of bets in columns 1 and 2 will be associated with low winning percentages in columns 4 and 5. Indeed, the results confirm the 21By definition, the sum of the fraction of bets on visiting favorites (row 2, column 1) and home underdogs (row 1, column 2) must add up to 100 percent. The same holds for home favorites and visiting underdogs. --------------------------------------------------------------------------------Page 21 prediction. Across the four categories considered, the rank-order correlation between the percentage of bets placed and the win percentage is -1. Bets placed on visiting favorites win only 47.8 percent of the time. Bets on home favorites are successful in 49.1 percent of cases. Bets on underdogs, which are under-represented in the data, win more than half of the time. Notably, bets on home underdogs have win rates of 57.7 percent, well above the threshold required for a bettor to break even.22One might be tempted to discount the observed relationship between a high fraction of bets made and low winning percentages since the results are based on the outcomes of only 236 games of a single season. Although it is impossible to obtain the data on quantities for earlier years, outcomes of games relative to the spread are readily available. Assuming that the strong tendencies towards betting on favorites are persistent across years, one would expect the patterns in winning percentages to also be persistent. Indeed, the winning percentage patterns have previously been documented by Golec and Tamarkin (1991) and Gray and Gray (1997). They are also present in Table III, which includes results on game outcomes for twenty-one seasons of data (1980-2001), covering almost 5,000 games. The first column of the table presents the percentage of bets made on teams of a particular type in the 2001-02 data; the second column is the win percentage over the past twenty-one years. The results are quite consistent with those of Table II. Overall, favorites win less than half their games. The null hypothesis of a 50 percent win rate for favorites is rejected at roughly the .01 level. As predicted, visiting favorites (who attract a greater share of the bets) do especially poorly, winning only 46.7 percent, again 22The win percentages on visiting favorites and home underdogs need not sum to 100 percent because the fraction of the bets on the favorite and the underdog varies across games. --------------------------------------------------------------------------------Page 22 rejecting the null of a .50 win percent at approximately the .01 level. Given this win percentage, a naive strategy of always betting against visiting favorites would actually have yielded positive profits over these two decades (as was also true in 2001-02).23In light of the consistent lack of success of favorites, especially visiting favorites, it is remarkable that the strong bettor bias towards such bets persists. The bettor bias is not concentrated among a small fraction of bettors, either. Three-fourths of the contestants chose favorites more often than underdogs. Only two percent of bettors chose visiting underdogs for at least half of their picks. Similar results are also obtained from betting on other American sporting leagues. Homeunderdogs covered the spread in 53.2 percent of the National Collegiate Athletic Association (NCAA) college football games played in 2002, as well as 53.0 percent of professional basketball games in the 2002 season of the National Basketball Association (NBA).23Although not shown in tabular form, the year-by-year patterns in the data confirm the overall findings. In only 4 of the last 21 years have favorites covered the spread in as many of 50 percent of the games. The likelihood of that occurring if the true win likelihood is .50 is less than 1in 300. Only 3 times in 21 years have visiting favorites won against the spread in 50 percent of games. --------------------------------------------------------------------------------Page 23 Just how much do bookmakers increase their profits by exploiting bettor biases in professional football? Assuming that (1) the total distribution of bets in this sample is representative of overall betting, and (2) there is no information in aggregate bettor preferences (the evidence presented below cannot reject this), it is straightforward using the values in Table II to calculate that the way spreads are currently set, bettors should win 49.45 percent of their bets.24Given the standard “vig” of bettors risking 110 units to win 100, a bookmaker who wins half his bets has a gross profit rate of 5.0 percent. If bettors win only 49.45 percent of their bets, the expected gross profit rate jumps to 6.16 percent (.5055*110-.4945*100). Thus, in expectation, this seemingly minor distortion of the win rate increases gross profits by 23 percent. It is true, of course, that the bookmaker must bear some risk when the bets are not balanced on both sides of the wager. Because game outcomes are likely to be independent, however, the risk is minimized as the number of games played increases. For instance, in the case where 63 percent of the money is one side of each wager and that team wins 48 percent of the time, over the course of the NFL season (roughly 250 games) the bookmaker’s expected gross profit rate is 6.1 percent with a standard deviation of 2.5 percent. Thus, the bookmaker would be expected to make negative gross profits less than once every one hundred seasons. If one looks over a five-year time frame, the standard deviation drops to 1.1 percent. So the probability of a bookmaker losing over any given five-year period in this scenario is less than one in 10,000. Relative to a 23 percent increase in gross profit, the costs associated with bearing this level of risk appear 24The number 49.45 is obtained by multiplying the probability that the home favorite wins a game times the percent of total bets on home favorites plus the probability that the visiting underdog wins times the percent of overall bets on visiting underdogs, etc. --------------------------------------------------------------------------------Page 24 minimal.25 One cost to bookmakers of bearing risk, however, is the need to have substantial liquid capital available to them in case of an adverse shock. The last two panels of Table III explore the relationship between other factors and betting on favorites. In the 2001-02 data, a higher fraction of bets were placed on favorites in the first half of the season than in the second half (63.4 percent versus 56.3 percent), and the week of the season was highly statistically significant in predicting bet shares in Table I. Whether this is simply an idiosyncracy of the 2001-02 season is uncertain. Consistent with the theory, the win percentages for favorites over twenty-years are higher in the second half of the season (48.5 versus 47.7), although the differences is not statistically significant. Finally, the bottom panel of the table demonstrates that the size of the spread has little impact on the distribution of bets on the favorite; correspondingly, there is little apparent difference in win percentages across these games in column 2. Is there evidence that some bettors are especially skillful in picking winners?In order for the current system of price-setting (in which the bookmakers set a price and do not adjust that price to equilibrate supply and demand) to survive, there cannot exist a sufficient number of bettors with an ability to pick winners that exceeds that of the bookmaker. This is particularly true when the bookmaker distorts prices to exploit the subset of bettors with biases. In that case, a sophisticated bettor only needs to be slightly better than the bookmaker in 25A major puzzle in this industry is the rarity of price competition, i.e. the vig is almost universally 10 percent. It is possible that the bearing of risk somehow supports this equilibrium. One website, www.mvpsportsbook.com, acts as a traditional financial market-maker, matching buyers and sellers, but taking no positions on game outcomes. The commission charged for this match-making service is less than 1 percent of the bet – far smaller than the traditional vig. --------------------------------------------------------------------------------Page 25 determining the true odds to turn a profit.26Indeed, Strumpf (2002) argues that much of the internal structure of bookmaker organizations is designed to protect the bookmaker against adverse selection by these talented bettors. Testing for bettor skill is complicated in my data set by the fact that there is a great deal of attrition over the course of the sample, and the attrition is not random. Bettors who have performed poorly up to that point in time are much more likely to leave the sample since the chances of receiving a prize are very low for these contestants. I consider two possible approaches for testing for heterogeneity in skill across bettors. The first approach is to look at the overall distribution of games won over the course of the season and to test whether that distribution is consistent with that which would have been generated by homogeneous bettors.27Because of attrition, however, these data are incomplete. Approximately 82 percent of possible bets were actually placed. Under the assumption that the outcomes of bets on the missing games can be modeled as being generated by independent coin tosses with probability .5, it is possible to simulate what the distribution of wins would have been without attrition. This approach has the obvious drawback that the simulated portion of the data is generated by the process that I have defined as the null hypothesis against which to test. Thus, such a test is biased against rejecting the null of no heterogeneity.28Figure IV presents a 26And, as demonstrated above, even a naive strategy of betting against all visiting favorites has been marginally profitable. 27Because the worst-place finisher gets a payoff, those near the bottom have an incentive to try to intentionally pick losers. If they have some ability to do this, that will exaggerate the bottom tail, exacerbating deviations from normality. 28In defense of the manner in which the missing data are generated, other results presented below suggest that there is no evidence of serial correlation across weeks in a given --------------------------------------------------------------------------------Page 26 representative histogram of the distribution of these simulated final win totals. Superimposed on the histogram is the corresponding normal distribution which the data would be expected to approximate if generated by i.i.d. coin tosses with a win probability of .50. Visually, the observed distribution closely mirrors the normal distribution. P-values for the three generally applied tests of normality (skew test, Shapiro-Francis, and Shapiro-Wilk) are well within the acceptable range. Thus, with the caveat that the test is biased against rejection due to the simulated data, there is no evidence to reject the null hypothesis of no differences in skill across bettors in the sample. The second approach to testing whether there is heterogeneity across bettors in ability to pick winners is to look for persistence in win rates. A priori, it is not known who the skillful bettors are. Success early in the season, however, is likely to be a (possibly noisy) signal of talent. Thus, in the presence of heterogeneity in skill, one would expect those who do better at the beginning of the season to also outperform later in the season. Like the test of homogeneity of skill presented above, this approach is not robust to particular sources of attrition. If high-skilled bettors who have been unlucky in the early part of the season are less likely to quit than similarly placed low-skilled bettors, than this approach will be biased against finding heterogeneity in skill across bettors. The poor-performing bettors who persist will be disproportionately drawn from the high-skilled group and thus will be expected to perform well on average later in the season. bettors ability to pick winners. Bearing in mind this caveat, I estimate equations of the form--------------------------------------------------------------------------------Page 27 (4)WINbwg=α + β1Hbw+ β2(Hbw)2 + β3(Hbw)3 + β4(Hbw)4+ γXbwεbwgwhere b, w, and g denote bettors, weeks of the season, and specific games respectively. WINis a variable equal to one if the bettor picks the game and covers the spread, .5 if the game is a push, and zero otherwise. The variable H represents the bettors cumulative historical winning percentage across all games played thus far in the season. The vector X captures other predictors of whether or not the bet is won, for example if the team chosen is a visiting favorite. The quartic in the cumulative winning percentage is designed to non-parametrically capture the serial correlation across betting performances. The first week of the season is omitted from the regression because there is no bettor history. The equation is estimated using weighted least squares, with weights determined by the number of games making up the history. The reported standard errors have been corrected through clustering to account for the fact that H is correlated across games for a given contestant. Table IV reports results of the estimation. In the first column, the cumulative betting success rate is constrained to enter linearly. Although not statistically significant, the point estimate implies that bettors who have been more successful up until that point in the season are predicted to do slightly worse in the current week. This argues against heterogeneity in skill across bettors, which would lead to a positive coefficient. The bottom panel of the table reports the predicted success rate for bettors with varying win percentages up to this point in the season. Bettors across the entire range are predicted to get slightly more than fifty percent of the bets correct. The second equation adds the quartic in betting history. Although the history variables are jointly statistically significant, the R2is very low (.0004). Bettors averaging 40 or 50 percent correct up to this point are predicted to perform right around 50 percent, bettors who have won --------------------------------------------------------------------------------Page 28 60 percent of games prior to this week are projected to win only 48.7 percent of games. Once again, these results argue against persistent differences in skill. Adding covariates in column 3 has little impact on the conclusions. Columns 4 and 5 restrict the sample to exclude the first five and ten weeks respectively, on the rationale that cumulative win percentages early in the season may not be very informative. The results provide no evidence that strong past performance predicts wins today. In summary, there is little in the data to suggest that, at least in this particular sample, there is heterogeneity in skill across bettors. This result may be due partially to the relative sophistication of bettors in the sample – perhaps the most naive bettors are unlikely to frequent internet bookmakers. 29Does pooling information across bettor preferences help in predicting the outcome of games? In other contexts, it has been argued that aggregating information across agents provides valuable information in predicting future outcomes. For example, Clemen and Winkler (1986) and Fomby and Samant (1991) find that the consensus estimate of future GNP growth is a better predictor than any one individual’s estimate. One might also expect such a pattern to be present in superbowl betting, especially because price is set unilaterally by the bookmaker. To the extent that the bookmaker sometimes makes mistakes, one would expect that many bettors will simultaneously recognize the presence of the mistake and disproportionately pick one team. 29Strumpf (2002), for instance, reports the existence of a fraction of New York bettors who always bet on the Yankees, even though the bookmakers, knowing their preferences, systematically offer these bettors substantially worse odds than other clients. --------------------------------------------------------------------------------Page 29 There is one simple result in my data which suggests that aggregating opinions across bettors may carry valuable information: despite the fact that more than half the money is bet on favorites and the bookmaker set the odds so that favorites win less than half the games, the overall winning percentage for bets placed is 50.1 percent. As noted earlier, based on the odds offered by the bookmaker and the distribution of money bet, one would expect 49.45 percent of all bets to win if there was no correlation between the percentage of bettors choosing a game and the game’s outcome. The difference between 50.1 percent and 49.45 percent implies that games in which a greater fraction of bettors choose the favorite (or alternatively the underdog) are more likely to be won by the favorite (underdog). Thus, in principle one might believe that knowledge of aggregate bettor preferences might be useful in prediction, making access to quantity data (which is in general very difficult to obtain, but is available prior to the start of the games through this contest) valuable.30Table V provides additional suggestive evidence on this issue. Bets are divided into four categories: home favorites, visiting favorites, home underdogs, and visiting underdogs. Within each category, bets are divided into quartiles according to how popular that team was with bettors. For instance, the top quartile of bets on home favorites would capture those bets in 30Interestingly, the win percentage for mock bets placed at www.mvpsportsbook.com were 49.5 percent in 2001 – precisely what would be expected if aggregate bettor preferences contained no information. One possible explanation for the difference in win rates between the bettors in mysample and those at wagerline is that my sample contains a more sophisticated set of gamblers who have much more money at risk. --------------------------------------------------------------------------------Page 30 which an especially high share of the people making selections in that game chose the favorite. Results presented in the table show the total for all games (the first column) and then broken down by quartile (columns 2-5). If aggregating bettor preferences is valuable in predicting performance against the spread, then one would expect to observe a monotonic, declining win percentage moving from the top quartile to the bottom quartile of bets by popularity. In practice, bets in the second quartile yield the highest winning percentage (52.9 percent). The top and third quartiles also win more than 50 percent of bets. The worst performance by far (46.6 percent) is for those selections that are least popular with bettors. Thus, while the evidence is mixed, there is at least a suggestion in the data of the possibility that popular bets fare better, at least relative to the least popular choices in the bottom quartile. Table VI tackles this question more rigorously in a regression framework that estimates the impact of a bet’s popularity non-parametrically using a fourth-order polynomial and controlling for other factors such as week of the season dummies and team-fixed effects. In addition, Table VI reports standard errors corrected for the fact that although there are over 19,000 bets made, there are only a few hundred game outcomes; for a given game, the bet outcomes will be perfectly correlated, so effectively there are only a few hundred degrees of freedom. The results of the table provide suggestive, but ultimately quite weak, evidence in support of the value of aggregating preferences. In columns 1 and 2, a bet’s popularity among bettors enters linearly. The coefficient is .11, implying that a ten percentage point increase in the number of bettors favoring a particular gamble is associated with a 1.1 percentage point increase in the likelihood that team wins. While substantively large, the estimate is not statistically significant. When allowing the coefficient on a bet’s popularity to vary in a non-linear fashion, --------------------------------------------------------------------------------Page 31 the individual coefficients are no longer readily interpretable, so I report average win percents by quartile in the bottom panel of the table. The patterns observed in columns 3-5 are similar to those in the raw data, with the bets in the top half of the distribution winning more than fifty percent and bets in the bottom quartile doing especially poorly. It is important to note, however, that the coefficients on bet popularity are in all cases jointly insignificant (as reflected in the P-values reported near the bottom of the table). Thus, great caution must be exercised in drawing any conclusions from these regressions.Section V: ConclusionsThis paper utilizes a unique data set that includes information on both the prices and quantities of bets placed to analyze the workings of the market for online superbowl betting on NFL football. The results provide an answer to the apparent puzzle of why these markets do not look much like markets at all, characterized by small and infrequent price changes. The answer is that the bookmakers are better at predicting game outcomes than the typical bettor. As a consequence, the bookmakers are able to set prices in order to exploit their greater talent, and apparently yielding greater profits than could be obtained if the bookmakers acted like traditional market makers and attempted to equilibrate supply and demand, avoiding taking large stakes in the outcomes of games. In the presence of some bettors who are as skillful in picking games as are the bookmakers, there are limits to how much prices can be distorted without creating profit opportunities. It appears that the bookmakers have distorted prices to the point where they bump up against this limit. In order to support this mode of price-setting, it cannot be the case that a significant --------------------------------------------------------------------------------Page 32 fraction of bettors (or more precisely dollars bet) have better information than the bookmaker. In that case, it would be the bookmaker who would be exploited by this system. At least within mysample of bettors, I see little evidence of heterogeneity in ability to pick winners, suggesting that the bookmaker might be as good or better at picking game outcomes than any of the bettors. Given the incentives for the bookmaker to get the spread right, it is hardly surprising that the most talented individuals would be employed as the odds makers. Perhaps, then, a fundamental difference between online superbowl betting and NFL is that it is possible in online superbowl betting to find and hire a small set of individuals (the odds makers) who can systematically do better in predicting game outcomes than can bettors overall. In NFL, on the other hand, the flow of inside information or the inherent complexity in valuing companies may make it impossible for one individual to do better than the market, meaning that a market maker who acted like a bookmaker would do worse than one who simply equilibrated supply and demand and took advantage of the bid-ask spread. The weight of the evidence regarding the inability of fund managers to systematically beat the market indexes is consistent with this conjecture. ---------------------------------------------------------------------------------------------------------------------------------------------------------------Page 36 Table I: Predicting the Fraction of Bets Placed on the FavoriteDependent variable: Percent of bettors placing bets on the team that is favored Variable (1) (2) (3) (4) Constant .606 (.009) .689 (.025) ----- ----- Home team favored by more than 6 points ----- -.129 (.031) -.131 (.031) -.144 (.031) Home team favored by 3.5 to six points ----- -.127 (.033) -.123 (.032) -.136 (.037) Home team favored by 3 or fewer points ----- -.126 (.031) -.126 (.031) -.123 (.043) Visiting team favored by 3 or fewer points -.005 (.030) -.026 (.030) -.057 (.033) Visiting team favored by 3.5 to 6 points -.016 (.035) -.002 (.034) -.002 (.034) Week of season dummies included?No No Yes Yes Team dummies included?No No No Yes R-squared ----- .165 .299 .484 P-value of test of joint significance of: Spread variables ----- <.01 <.01 <.01 Week dummies ----- ----- <.01 <.01 Team dummies ----- ----- ----- <.01 Notes: Omitted category for the spread variables are games in which the visiting team is favored by ten or more points. The unit of observation is a game. The number of observations is equal to 242 in all columns. Standard errors are in parentheses. The method of estimation is weighted least squares, with the weights proportional to the total number of bets placed on the game.--------------------------------------------------------------------------------Page 37 Table II: Bets Placed and Won on Favorites and UnderdogsWhich teamis favored in the game? Percent of total bets on the game that are placed on: Percent of bets placed that win (i.e. cover the spread) when a team is: (1) (2) (3) (4) (5) (6) Favorite Underdog Total, favorite and underdog Favorite Underdog Total, favorite and underdog Home team56.1 [N=12,011] 31.8 [N=7,190] 47.0 [N=19,201] 49.1 [N=6,741] 57.7 [N=2,286] 51.2 [N=9,027] Visiting team68.2 [N=7,190] 43.9 [N=12,011] 53.0 [N=19,201] 47.8 [N=4,904] 50.4 [N=5,270] 49.1 [N=10,174] Total, home and visiting team60.6 [N=19,201] 39.4 [N=19,201] 50.0 [N=19,201] 48.5 [N=11,645] 52.6 [N=7,556] 50.1 [N=19,201] Notes: The values reported in the first three columns of the table are the percentage of total bets placed on the named team (e.g. home favorite in row 1, column 1). The values reported in the last three columns of the table are the fraction of bets placed that win. The unit of analysis is a bet. The number in square brackets is the total number of bets placed in each cell. The results in this table exclude the six games where the spread was equal to zero, i.e. neither team was favored.--------------------------------------------------------------------------------Page 38 Table III: A Comparison of Betting Frequencies in my Sample and Game Outcomes against the Spread over the Last 21YearsType of gamePercent of bets in my sample that are placed on the favorite Percent of games in which favorite wins over last 21 NFL seasons All games 60.6 48.2 [N=4,793] Games in which home teamis favored 56.1 48.8 [N=3,310] Games in which visiting teamis favored 68.2 46.7 [N=1,483] Games played in first half of season 63.4 47.7 [N=2,209] Games played in second half of season 56.3 48.5 [N=2,584] Games in which the point spread is greater than six points 60.4 48.5 [N=1,759] Games in which the point spread is 3.5 to six points 59.5 48.1 [N=1,475] Games in which the point spread is less than or equal to three points 61.5 47.8 [N=1,559] Notes: Values in the first column are based on bets placed for the sample of 285 bettors for the 2001 season used throughout the paper. Values in the second column are game outcomes for the 21 NFL seasons covering the period 1980-2001.--------------------------------------------------------------------------------Page 39 Table IV: Bettor Performance as a Function of Cumulative Win Percentage up to that Point in the SeasonDependent variable=bettor wins this bet Variable (1) (2) (3) (4) (5) Bettor’s cumulative ranking on win percentage (0=last place, 1=first place) -.012 (.015) -.06 (.32) .03 (.34) .10 (.36) .07 (.47) Cumulative rank^2 ----- .54 (1.31) .25 (1.36) -.07 (1.47) .21 (1.91) Cumulative rank^3 ----- -1.04 (1.97) -.69 (2.04) -.18 (2.20) -.68 (2.86) Cumulative rank^4 ----- .56 (.98) .41 (1.01) .15 (1.09) .41 (1.41) Bettor bets on home favorite ----- ----- -.048 (.010) -.064 (.014) -.037 (.016) Bettor bets on home underdog ----- ----- .046 (.012) .026 (.014) -.010 (.018) Bettor bets on visiting favorite ----- ----- -.005 (.010) -.002 (.013) .121 (.016) Constant .506 (.009) .499 (.024) .497 (.026) .534 (.030) .598 (.036) R-squared .0001 .0002 .0041 .0048 .0131 Number of observations 18,345 18,345 18,345 12, 985 6,785 Sample of games used in estimation Weeks 2-17 Weeks 2-17 Weeks 2-17 Weeks 6-17 Weeks 11-17 P-value of test of joint significance of cumulative rank variables ----- .64 .51 .73 .72 Predicted win percentage if cumulative ranking on win percentage falls in: Top quartile 49.4 49.0 49.0 49.7 49.0 Second quartile 49.8 50.2 50.1 50.4 50.5 Third quartile 50.1 50.7 50.8 51.0 51.2 Bottom quartile 50.4 49.9 49.9 50.1 49.4 Notes: The dependent variable in all regressions is equal to one if the bettor wins the bet, zero otherwise. The unit of observation is a bet. A bettor’s cumulative ranking on win percentage is calculated by taking an unweighted average of that bettor’s win percentage on all bets up to that point in time, and then computing the bettor’s rank among all bettors. The method of estimation is weighted least squares, with the weights proportional to the total number of games making up the year-to-date history.Standard errors (in parentheses) are corrected to take into account correlation in the cumulative rank variable across bets for a given bettor. --------------------------------------------------------------------------------Page 40 Table V: Win Percentages as a Function of whether a Bet is Popular Among BettorsWin percentage as a function of the degree to which bettors prefer this selection (relative to other teams in the same category): Category of games: All games, regardless of degree preferred bybettors Top quartile (i.e. selections most popular with bettors) Second quartile Third quartile Bottom quartile (i.e. selections least popular with bettors) Home favorites 49.1 [N=6,741]46.0 [N=1,666]44.6 [N=1,697]57.4 [N=1,696]48.2 [N=1,682]Visiting favorites 47.8 [N=4,904]52.0 [N=1,240]57.4 [N=1,210]43.9 [N=1,206]39.2 [N=1,205]Home underdogs 57.5 [N=2,286]71.0 [N=568]52.6 [N=579]59.9 [N=564]47.5 [N=575]Visiting underdogs 50.4 [N=5,270]45.5 [N=1,325]59.6 [N=1,288]45.6 [N=1,318]51.0 [N=1,339]All favorites 48.6 [N=11,645]48.6 [N=2,906]49.9 [N=2,907]51.8 [N=2,902]44.4 [N=2,887]All underdogs 52.6 [N=7,556]53.2 [N=1,893]57.4 [N=1,867]49.9 [N=1,882]49.9 [N=1,914]All bets (favorites and underdogs) 50.1 [N=19,201]50.4 [N=4,799]52.9 [N=4,744]51.0 [N=4,784]46.6 [N=4,801]Notes: The values reported in the table are fraction of bets won. The unit of observation is a bet. Games are categorized into quartiles within bet type (e.g. home favorites, visiting favorites, home underdogs, visiting underdogs) according to the fraction of bets on the team in the named category. Top quartile bets correspond to those cases where the greatest fraction of bettors selected the team in the named category.--------------------------------------------------------------------------------Page 41 Table VI: Does Pooling Information Across Bettors Help in Picking Winners?Dependent variable=bettor wins gameVariable (1) (2) (3) (4) (5) Rank of this bet’s popularity (relative to other games of this type) (0= least popular; 1= most popular) .11 (.11) .11 (.11) -1.02 (1.82) -1.16 (1.84) -1.04 (1.85) Rank of bet’s popularity^2 ----- ----- 4.81 (7.61) 5.48 (7.63) 5.18 (7.55) Rank of bet’s popularity^3 ----- ----- -6.77 (11.20) -7.92 (11.23) -7.57 (10.94) Rank of bet’s popularity^4 ----- ----- 3.01 (5.41) 3.64 (5.44) 3.46 (5.21) Bet on home favorite ----- -.013 (.084) -.013 (.083) -.011 (.083) -.013 (.082) Bet on home underdog ----- .073 (.072) .073 (.072) .080 (.074) .055 (.078) Bet on visiting favorite ----- -.026 (.073) -.026 (.073) -.022 (.073) -.054 (.070) Constant .447 (.048) .450 (.062) .499 (.119) ----- ----- R-squared .004 .008 .010 .022 .048 Include week of season dummies?No No No Yes Yes Include team-fixed effects?No No No No Yes P-value of joint significance of bettor preference variables ----- ----- .81 .80 .78 Predicted win percentage by quartile of popularity of this bet: Top quartile (i.e. selections most popular with bettors) 54.2 54.1 52.3 52.3 52.1 Second quartile 51.5 51.4 54.2 54.1 54.0 Third quartile 48.8 48.8 49.4 49.5 49.8 Bottom quartile 46.1 44.5 44.6 44.6 46.1 --------------------------------------------------------------------------------Page 42 Notes: The dependent variable in all regressions is equal to one if the bettor wins the bet, zero otherwise. The unit of observation is a bet. The rank of a bet’s popularity is calculated relative to other games of this type (e.g. home underdog). The method of estimation is weighted least squares, with the weights proportional to the total number of bets placed on the game. Standard errors (in parentheses) are clustered to take into account correlation in the rank of bet popularity variables for different bets placed on the same game.--------------------------------------------------------------------------------

1. online superbowl betting 2. betting 3. football 6. nfl 7. sports 8. superbowl betting 9. gambling 10. odds 11. online betting 15. football betting

2005-01-12T13:12:00.000-08:00

18. superbowl 19. online sports betting 22. sports book 23. vegas 24. sportsbook 31. online 34. bet 38. online casino gambling 40. super bowl betting 42. superbowl odds 44. sports lines 45. betting odds 46. super bowl gambling 47. betting lines 48. football betting spread 49. wagering 50. wager 52. super bowl odds 53. online spread betting 55. sports betting 69. bet on sports online 74. online bet 77. beting 78. online betting games 91. bowl 92. online betting 93. nfl betting 100. futurefootballwagering 118. gamblingfootball 122. gamblingnfl 124. gamblinglines 125. futuresfootballwagering 126. gamblingodds 127. nflfootballplayoffs 128. nflfootballs 129. gamblingline 131. gambling on football 134. futuresfootballlines 135. footballspread 161. nflfootballspread 169. footballbettingline 170. footballbettinglines 171. footballbookie 172. footballbettings 173. footballbettingodds 178. nflbookie 179. nfl beting

Superbowl Betting Guide

2005-01-11T16:19:00.000-08:00

Super Bowl Betting Guide

Above is an example of what the line might look like in Super Bowl. The computer
betting number for the Falcons is 501; for the Broncos it is 502. The Broncos
are a seven point favorite. This is indicated by the -7, to the right of their
name. The Falcons, therefore, are a seven point underdog. The number 47 indicates
the total (over/under). The +180 and -200 represents the money line. The game
will start at 1:00 p.m. Nevada time. The bottom team listed is the
home team.

In Super Bowl wagering, a point spread is used to attempt to equalize
two teams. In the above example, the Broncos are a seven point favorite over
the Falcons. If you wanted to wager on the Broncos, they would have to win
the game by more than seven points in order for you to win your bet.

If you would have decided to wager on the Falcons, then in order for you
to win your bet the Falcons would have to either win the game or lose
by less than seven points. Should the broncos win by exactly seven points,
it is classified as a tie and all wagers on both teams would be refunded.

The number 47 is the total (over/under) number. You may wager on whether
the combined final score on both teams will be either over or under
this number. Should the total combined final score end exactly on 47, then
all bets would be classified as a tie, and all wagers would be refunded. All
points scored in sudden death overtime are counted in the final score.

When betting the over/under, it does not matter who
wins the game, only how many points are scored. In Super Bowl, when betting
on the point spread or the totals, the player wagers $11.00 to win $10.00.

The money line bet is indicated by +180 on the Falcons and -200 on
the Broncos. Money line is a wager for a particular team to win
the game without consideration of the point spread. For example: if you bet
the Broncos to win the game you would wager $2.00 to win $1.00 for a total
return of $3.00. If you bet the Falcons to win the game, you would wager $1.00
to
win $1.80 for a return of $2.80.

Super Bowl Parlays

You may combine two or more teams or totals in a single wager.
In parlay wagering, as in teaser wagering, ALL included in
a parlay must win by a margin as adjusted by the point spread. The more teams
included in a parlay, the greater the payoff. In the event of a tie in any
game, or total, the parlay is reduced to the next betting bracket. Thus, a
four team parlay with one tie would simply become a three team
parlay. Minimum parlay bets are $10.00.

Super Bowl parlay payoff prices are as follows:

Two team parlay	13 to 5
Three team parlay	6 to 1
Four team parlay	10 to 1
Five team parlay	20 to 1
Six team parlay	40 to 1

Again, in a parlay, all teams must cover their point
spread, as printed on your ticket.

Super Bowl Teasers

The generally a sportsbook offers 6, 6 1/2 and 7 point teasers.
You may choose from two to six teams in a teaser.
A teaser wager allows you to adjust the point spread in your favor by 6, 6
1/2
or 7 points. All teams selected in a teaser combination
must win by a margin as adjusted by the teaser
point spread. In a two team teaser involving a tie, wager is considered no
action
and money is refunded.

online superbowl betting

2005-01-11T07:35:00.000-08:00

online superbowl betting