Thursday, February 24, 2011
If I gambled I might try out this theory
First, a quick note about how gambling works. There are basically three types of gamblers: The everyday guys, the big fish and the sophisticated gamblers.
The everyday guy bets his gut and falls prey to all sorts of cognitive biases like confirmation bias. This guy is basically the cash cow that the books make their money on. They don't bet a lot of money, but there are a lot of them. Their bets are no better than guessing, and any time the books get you to that point, they will make their money. The everyday guy will bet on "his team". The bet is usually about a show of loyalty and maybe some sort of misguided belief that they know things that nobody else does, because they read an article that's freely available to the public.
The big fish is more sophisticated, he knows more about which team is likely to cover (or win, depending on the type of bet), but his problem is that he likes huge bets on single outcomes. He does better than guessing, by a little bit. However, he falls prey because if he wins big on one bet he'll just bet more on a subsequent win, until eventually he gets wiped out and has to start from scratch. You can only make money in gambling if you eventually remove money from your "gambling pot" and stop putting it back in the form larger bets. Because everybody loses at some point. For the big fish, it's not so much about making money, it's about the thrill of gambling. Sure, they love to win, but they love the adrenaline rush of the risk more than anything else. And to keep getting that adrenaline rush, you have to keep betting more and more. Also, for some reason they tend to get a bigger rush when they're betting for something instead of against something, so they tend to focus on things they think are undervalued and bet for them. This is a subtle, yet often important point we'll come back to later.
The sophisticated gamblers are those that pick and choose their moments carefully and always spread their risk by betting smaller sums on a LOT of games. They use a lot of statistics and will occasionally bet on games they know nothing about, if the statistics tell them so. These guys can actually make a living betting on sports, as long as there are enough common guys and big fish out there that the books still have to set the lines around them, and not the sophisticated gamblers. The sophisticated gamblers look at gambling as a science and tackle it in a very cold sophisticated way. Many don't even enjoy gambling, in fact they wouldn't be as good at it if they did. For them it's purely a job, they're no different than a market analyst. In fact many of them are former wall street quants.
So to make money gambling, you have to understand a few things:
1) lines aren't set so that the books equal the probability for each team to win. Lines are set to have equal money coming in as much as possible. Because this equals out their risk.
2) You need to find occasions where you think that big fish and regular guys will have some reason to overvalue one side or another and bet against that. You aren't betting against the teams, you're betting against the other gamblers.
3) Regular guys bet with their heart and common information. You can get a good idea of where they will bet based on which team has the larger betting fanbase and by reading popular ESPN articles.
4) Big Fish tend to be a little more sophisticated, mostly betting on things that they think are undervalued. For instance, in baseball they tend to bet on the team with the higher OBP.
5) Sophisticated guys tend to bet against things they think are overvalued.
6) Unless you bet for the joy of it, you want to be the sophisticated guy.
How do you find something that is overvalued? Usually you have to find some statistic that is highly correlated with a team's success, yet also highly unstable, that is not an easily repeatable skill and very much a product of luck.
MLB: batting average with runners in scoring position. For a long time you could make a killing in baseball by getting AGAINST the team with the higher batting average with runners in scoring position. It's been more or less statistically proven that hitting with runners in scoring position isn't really a repeatable skill. Teams almost always regress to what their overall batting average is, plus a couple of points (for reasons too numerous and complicated to go into here, most hitters are going to hit slightly better with runners in scoring position, not because of skill, but because of systematic differences in the situation). Yet hitting with runners in scoring position is highly correlated with success. You can still make SOME money with this approach, but not nearly as much as the good old days when the first thing that made someone like a player or team was clutch hitting.
NFL: Red Zone Offense and Defense. Unlike hitting with runners in scoring position, this is a valuable skill that some teams are going to be systematically better or worse at. However, it's still overvalued. This area is again highly correlated with a team's success, but it's also at least partially, if not primarily based on luck. You're isolating just a few moments in a game and giving them enormous importance in the outcome of the game. Any time this happens, it's often the case that the "skill" becomes overvalued. You can win a lot of games very simply by taking a team's redzone offense and defense rankings, adding them and then betting AGAINST the one with the better composite ranking. Hypothetical Example: Indianapolis v. Jacksonville (Indy 3.5 point favorite) Indianapolis is ranked 12th in red zone offense and 22nd in redzone defense. Jacksonville is ranked 23rd in redzone offense, but 3rd in redzone defense. Indianapolis' composite score is 34 v. Jacksonville's composite score is 25. Bet on Indianapolis. This may seem highly counterintuitive, betting on the statistically worse team, but remember, its not that you are betting on a team, its that you're betting against things that are overvalued. Red Zone offense and defense, being highly important to game outcome, yet largely based on luck, tend to be overvalued.
NBA: 3 point FG % defense. Again, this is a valuable skill, and skill has a lot to do with it, but again, luck has a lot to do with it as well, and it tends to be slightly overvalued when it comes to game outcomes because of how important it is to game outcomes, yet how relatively small the sample size is.
Early in the season, bet against teams with more rabid fanbases. Early in the season the big fish tend to lay low. There isn't yet a lot of information out there on which teams are better, making large single bets especially dangerous. Yet, there is some money to be made if you know how to bet against the regular guys, who are just as likely to bet early in the season as late in the season (excluding playoffs, where everybody bets more). Remember, regular guys bet with their heart, so it's often a good strategy to roughly gauge the rabidity and size of the fanbases and bet against teams whose fanbases vastly outpace their opponents. Early in a season, it's a pretty good idea to bet against the Cowboys in a matchup against the Jaguars. Whtever that line is, you can almost assure it will be slightly too high in favor of the Cowboys to bring in money for the Jaguars, but won't ever quite get "right" because of the Cowboys perpetually large, optimistic and rabid fanbase.
The major key here to being a sophisticated sports gambler is realizing what you're gambling against. You're gambling against the other gamblers. Thus you must focus not on which team you think is better, you must gamble against what you think is overvalued by the other gamblers. Especially the regular guys and the big fish. You also have to realize that this only wins on average, and that your best bet is to find quality strategies and bet small sums on a LOT of games, rather than large sums on a few games. Statistics work best over large samples.
Of course the best strategy is not to gamble and invest your money in bonds and stable stocks with good fundamentals. But I digress.
Posted by FJR at 12:28 PM