Tuesday, December 22, 2009

Gambler’s Fallacy

Suppose an NBA handicapping method has proven over time to produce 55 percent winners on average per season. If the selections are only posting 50 percent winners halfway through a current season, does this confident handicapping method begin to raise the wager amounts with the belief that 55 percent winners will be achieved at season end? And, future plays would need to win at greater than 55 percent of the time over the remainder of the season to raise the success rate from 50 to 55 percent.

The answer is an unequivocal “NO”. Many emotional cappers and bettors hold to the mistaken belief that the chances of winning increase during losing periods. Sports gamblers on a losing streak suffering from gambler’s fallacy will believe that their next wager will have an increased chance of winning. The reason is that these gamblers believe that they will win in the end; therefore they are more inclined to win the next bet when in the midst of a losing period.

Often times this disastrous thought process leads to larger wagers because sports gamblers feel they are “due” to win. Unfortunately, anyone who believes they are due to win is many times only increasing the probability of losing. Progressive betting schemes are the result of gambler’s fallacy and have caused the ruin of many a gambler.

Gambler’s fallacy fails to realize that the handicapping method is an inanimate entity with no memory. Each pointspread selection is unbiased and unprejudiced and independent of the last. Regardless of the outcome of prior wagers, each and every future bet will have the same 55 percent chance of winning, provided the handicapping method is truly capable of producing 55 percent winners over time.

Gambler’s fallacy is not to be confused with a statistical term called “regression to the mean.” Regression to the mean simply means that if an event has a certain statistical chance of occurring, the event should eventually approach the expected occurrence rate over the long term.

Looking at the graph below, it is seen for this hypothetical series of pointspread plays that during the first 100 or so plays, the winning percentage hovers between 48 and 60 percent. Only after about 225 plays does the winning percentage settle in to the 55% target area.




In plain terms, if an NBA handicapping method can produce 55% winners over time, then regardless of how many winning and losing streaks are encountered along the way, 55% winners should ultimately be achieved at some future point in time. Consequently, a winning handicapping method should eventually post an overall winning record. However, notice the operable terms are “ultimately” and “eventually”.

The inability to understand the difference between gambler’s fallacy and regression to the mean is one of the main reasons why many sports bettors bust their bankrolls long before they would have had any chance of turning a profit for the season. As soon as the wager amounts start to progressively increase during losing periods, gamblers place themselves in imminent danger of going bust. Trust me, the next bet has no compassion for the previous losing ones.

There are two major obstacles with a progressive betting scheme anyway. The first is that Fort Knox is needed to fund the venture since a losing streak will cause the wager amounts to increase dramatically since the objective is to win all the losses back plus show a profit when a winning bet does eventually occur. What happens if the bettor loses ten consecutive bets? Imagine the size of the next wager to recoup the losses.

When you think about it, our federal government could certainly embark on a progressive betting scheme in Vegas aimed at eliminating the federal deficit over time. The U.S. government has access to unlimited funds as it could simply print more money as needed. Such a bank vault could theoretically cover any number of consecutive losing streaks.

Unfortunately, the second obstacle poses a problem for Uncle Sam. Once the wager sizes become too large, the sportsbooks can simply refuse to take the bet. House limits provide the casinos with the ability to cut-off a bet size deemed excessive. Therefore, sportsbooks have built-in protection from colossal money sources involved in progressive stakes.

Sports gambling for any NBA contest is an indisputable, uncertain venture because nobody, and I mean nobody, can predict with any degree of certainty how long of a time period is required to realize the regression to the mean result. Going back to the above graph, who is to say when the 55% winning rate is achieved? It may take 300 to 400 pointspread plays to come true.

This is why Matchup Handicapping holds to the conviction that each bet should be the same wager amount during a betting season. Mission impossible is trying to predict winning and losing periods. It is fact that winning seasons experience losing streaks. It is hard enough picking 55% winners over time. It is impossible to pick 55% winners over every week or every month.

In summary, operating in a gambler’s fallacy mode attempts to predict when the regression to the mean outcome will come true. Unfortunately, the short term gambler’s fallacy will run its course long before the realization of the long term regression to the mean. As I have written many times before, the single biggest mistake made in sports betting is viewing it as a short term prospect.

On the contrary, a successful winning strategy involves a long term and disciplined commitment. You are not running a sprint. You are running a marathon.