MLB Totals Betting
by Robert Ferringo - 06/02/2008
Maybe it's the fact that the Boys of Summer are off the juice. Maybe it's the maple bats. Maybe it's the weather. Or the pitching. Maybe chicks stopped digging the long ball and now they dig the double play. I'm not certain exactly why, but there it is clear that scoring in Major League Baseball is down this year and that we've seen the end of the Steroid Era.
Over the next week I'm going to be breaking down the recent scoring situation in Major League Baseball and how it's altered the way that bettors and handicappers need to approach wagering on the games.
Now, betting is a results-oriented business so I'm going to start at the end of this topic and work my way back. Later on we'll discuss possible reasons for the lower scoring numbers and even some factors that may suggest that the lower scoring is just a two-month trend that will correct itself. But before we get through the hows and the whys of the apparent power outage my practicality beckons me to address the end result of this lower scoring first. And we'll do so in the area where scoring has the greatest impact - MLB totals betting.
First, as of June 1 there was an average of 9.00 runs scored per game - 8.8 runs per game in the American League and 9.2 per game in the National League. That number, as it stands, would represent the lowest-scoring season in a decade. I'm clearly not the first person to point all of this out. But while baseball writers and bobbleheads seem to worry about the lower scoring's impact on overall interest, attendance, records, et cetera, I am curious as to whether or not the sportsbooks and oddsmakers had adjusted to the current landscape of baseball and what impact it's had on betting MLB totals. The results were fascinating.
As of the start of business on June 1 there had been a total of 848 games played in the Major Leagues. Discounting pushes, teams have combined to go 373-for-814 against the total to give them a 45.8 percent 'over' rate. On the surface, and to the common observer, this percentage may seem a little low or a little odd. But to any handicapper or baseball numbers cruncher these numbers are amazing and they clearly differ from what's occurred in the recent past.
By comparison, testing for performance enhancing drugs began in 2005. That season saw an average of just 9.19 runs per game scored and the 'over' on MLB totals hit at just 48.2 percent, the lowest rate of the current century. Clearly the drug policing had an impact, and that impact was reflected both on the field and in the sportsbooks. But both the players and the totals market seemed to correct themselves. In 2006 there were 9.71 runs scored per game - the highest total of this decade - and the 'over' hit at a 50.7-percent clip and in 2007 there were 9.59 runs per game and the 'over' hit at a nearly flawless 50.1-percent rate.
Then, just like the housing market, the bottom fell out.
According to Covers.com, in the four seasons played between 2004 and 2007 there were a combined total of 9,207 games whose total didn't end in a push. Over that four-year the 'over' hit on 4,615 of those 9,207 games. That gave the oddsmakers an absolutely incredible 50.1 percent rate. It is incredible because it shows no significant statistical variance from the 50/50 chance that any given game has of going 'over' or 'under' and it creates nearly the perfect situation for profit for the sportsbooks. This percentage shows yet again what a cool, efficient machine books and oddsmakers are.
Further, those four-year totals support other research and studies conducted on the market of MLB totals. The 2002 Brown-Abraham Study at the University of Northern Iowa concluded that the over-under betting market for baseball is efficient. Or at least it was in the five seasons that they examined (1996-2000). What this means is that the totals that are released by Las Vegas Sportsbooks are an accurate reflection of expected results and therefore no profitable strategy exists that can consistently produce a profit.
In layman's terms, this study showed that the oddsmakers are as accurate as a Greg Maddux fastball. Since the oddsmakers have proven that the 'over' and the 'under' each hit about 50 percent of the time the market is efficient because you can't blindly bet one side or another and consistently produce a profit. The oddsmakers are so good at setting and moving totals that they've basically made it a coin-flip. It's doesn't mean that you can't make money betting totals, but it means that going into it the gambler does not have any statistical advantage.
Which brings us back to why this year's numbers are such an amazing anomaly.
The break-even point on a wager with 10-cent juice (an average wager at -110) is 52.6 percent. That means that if you bet 1,000 games you would need to hit 526 of them just to break even. This season blindly betting the 'under' would have given you an exceptional 54.2 percent rate of success. That means that if you had bet $100 on the 'under' of every game this season you would already be up $6,470 assuming a 10-cent vig on each game.
Basically, oddsmakers proved over the last 10 years that they had complete control over the totals market. You had a 50/50 chance of success, which creates the perfect opportunity for profit for the sportsbooks. The average totals player was screwed. But then this year even the oddsmakers are, at least at the moment, getting completely dominated by the lower-scoring games. Obviously the books are losing money on this deal. But what's happened is that there appears to be a loophole or trap door open for astute gamblers to sneak in and grab some extra cash.
Remember: the sportsbooks and oddsmakers aren't great at what they do because they are know the most about sports. They are great at what they do because, as mathematicians and accountants, they are better with a calculator than Albert Pujols is with a baseball bat. But right now their calculators appear to be broken.
Naturally, if there's a door open to gamblers they're going to slam it shut. And they've already started moving in that direction. In the recent past a posted total of 7.5 was usually pretty rare. Those totals were generally reserved for games featuring two of the top five or 10 pitchers in baseball, and usually involved National League teams. However, now it's not at all surprising to see two or three totals of 7.5 listed every day. In the last week alone - from May 26 through June 1 - there were an astounding 12 totals that closed at 7.5 or 7.0, including four in the American League. But even more remarkable is that only TWO of those 12 games actually went 'over' the total and the A.L. games went 0-4 against the number!
That leads us to the issue moving forward: how much more can and will the totals market correct itself? Will we start to see regular totals of 6.0 or 6.5? Will we see a total of 5.5? There are still roughly 1,600 games remaining in this season. Over that time frame it's a virtual certainty that the 'over' percentages will drift back toward 50 percent. This phenomenon is referred to as "regression toward the mean". But because we have seen such an extraordinarily low percentage of games going 'over' in this early part of the season the only way that this regression will take place is if games start to go 'over' at an equally prodigious pace. The result: there should be a period during this season where playing the 'over' blindly will yield a tidy profit.
Again, statisticians and economists have studied MLB totals and determined that this is a highly stable, razor-sharp, efficient market. Yet this year, mathematically speaking, the symbolic floodgates are open! I won't bore you with the high school math of it - standard deviations and variance and all that; for more info on this check out Doc's Sports Blog - but if the numbers hold (I don't believe they will, but if they do) this season would be the type of random, weird, exception that happens only about twice a century and the savvy bettor could be in a spectacular position to take advantage of it.
For more information on Ferringo's members picks, check out his Insider Page here.