Ask the Capper: Unders the Trend In College Hoops?
by Robert Ferringo - 02/06/2007
THIS WEEK'S ASK THE CAPPER QUESTION:
Bill from Oklahoma asks: Have your handicapping experts noticed that a higher percentage of college basketball games are going 'under' this year?
I received this e-mail from Bill on Wednesday, Jan. 31 and my initial reaction was to tell Bill that it's very unlikely to see a major trend like this take place in college basketball, where the sheer volume of contests only serves to fortify the statistical norms. I'm sure he may just feel like there have been less scoring or more 'unders' coming through because he was on the wrong end of a few totals plays over the past few weeks.
However, after some due diligence on my part I actually discovered that between Jan. 1 and Feb. 4 there was actually a higher rate of games staying 'under' the total this season than occurred last year. The statistical variance from 2007 to 2008 was slight but it definitely occurred.
Before I get to my findings, I have to say that this wasn't the most scientific of studies. I took a not-so-random-sample of games played during that period, using only games involving Top 25 teams. There are some confounding factors that may have played into more of these games staying 'under' the posted total. But, nonetheless, I figured this was as good of a sample as I would find for this little project.
During the 2006-07 season there was a total of 213 games played that involved at least one team that, at the time of tip-off, was in the Top 25. The 'over' in those games complied a 97-114-2 record, a 45.97 percent success rate. Someone blindly playing the 'over' during that time would have only experienced six winning days of the 35 sampled and would have been minus-$2,840 if they had wagered $100 per bet.
Conversely, and 'under' bettor would have experienced a 54.03 percent rate of success and would have banked $730 if they had played $100 per game.
Over the same time period - Jan. 1 to Feb. 4 - in the current 2007-08 season there were 195 games played involving at least one Top 25 school. The 'over' came in at a 89-106-2 mark against the total for a 45.64 percent success rate. Again, blindly playing the 'over' during that stretch would have yielded a net loss of $3,060 after factoring in the standard 10-percent juice on a losing wager.
And for the second year in a row a blind 'under' bettor would have turned a profit. This year they would have taken in a net profit of $810.
The simple answer to Bill's question would be that yes, there are more games staying 'under' the posted total this year. But as you can see the percentages are miniscule and would really only make a difference if someone were wagering large sums of money on every single Top 25 game. Again, you can argue that using the Top 25 isn't a good sample or that there were 16 more games played last year and that could skew the numbers. However, the bottom line is that over the past two seasons the 'under' has come in at a combined 54.29 percent rate.
The break-even point against a standard 10-percent vigorish is 52.38 percent - meaning that you have to win at least 52.4 percent of your wagers to show any profit - so, technically, we've found a profitable trend. But like I mentioned earlier, this info really is of little or no use to a handicapper or a recreational gambler because the only way to really show a profit (and a negligible one at that) is to wager vast amounts on every single Top 25 game played each day. That's not very practical.
Finally, I did notice a few small tendencies for both 'overs' and 'unders' over the past two years. But I would need much more research in order to refine these tendencies into any concrete information. And in case you are curious, the 'over' showed a significantly higher success rate after Feb. 4 and actually was a profitable proposition in the final month of the season.
>From Feb. 5 to the national championship game on April 2 there was a total of 266 games played involving either a team from the Top 25 or a team participating in the NCAA or NIT tournaments. The 'over' went 137-129 overall for a 51.5-percent clip. That would have been a losing bet for both the blind 'over' and blind 'under' player. However, the 'over' went 54-47 from March 1 to April 2 and would have earned a dollar bettor ($100 per game) all of $230 in net profit.
Even though it seems like a lot of games this was still a relatively small sample we were working with. The statistical norm for the 'under' is somewhere below 50 percent. I will continue to monitor the numbers for the rest of this season. Picking up a couple hundred bucks here or there for a month's work definitely doesn't seem like much. But if these numbers are consistent then we may have found something that can produce a lot of little profits that can become a big return.