I was hoping that would get your attention!
Just saying "Bowl Post Mortem" would have been too dreary. My parents went to Ohio State, and I still have family living in the Ohio area. Have to keep spirits up. Too bad they'll never fully know the thrill an Austinite had watching Texas beat USC last year!
I did find an interesting combination of indicators when running through this year's bowl data. I wanted to pass it along to you while the bowls were still fresh in our minds. This isn't one of those "after-the-fact" things that has been the topic of discussion in some Major Wager threads this week (teams who rush for less than 100 yards don't cover). This is a combination that you can look for in advance. This combo only came up nine times in the 32-game bowl slate, and it ended up going 7-2 against the spread.
Yesterday afternoon I went through and catalogued how each of the "category" headers we used in the Bowl Preview Series performed against the spread. I wasn't really expecting to find much. Nobody really thinks "take the team with the better record" is going to hit a home run. It didn't. But, an interesting combo did pop out.
BOWL ATS RECORDS
Team with the better straight up record: 12-13 ATS
Team with the better ATS record: 16-15 ATS
Team that played the toughest schedule: 17-15 ATS
Team that had the better point differential: 18-14 ATS
Team that had the better yardage differential: 17-15 ATS
Team that had the better turnover differential: 13-19 ATS
Team that had the better TD/INT passing ratio: 20-11 ATS
It's those last two that got my attention. Jay Graziani's article yesterday mentioned that some handicappers consider turnovers as being representative of luck, while others consider it being representative of skill. The jury's still out on that one, and probably always will be. It was interesting though that many teams with great turnover differentials laid eggs in their bowl games. Ohio State had a big edge over Florida in this regard (+11 to +1 for the year), but lost the stat 2-0 in the BCS championship game. Rice had one the biggest edges possible in a bowl game (+23 to -9) against Troy, but lost that category 5-1 in the New Orleans Bowl.
On the list above, a positive turnover differential was the only thing that turned out to be a negative indicator to any real degree. This isn't necessarily proof that turnovers are luck. But, it does suggest that teams who benefit from turnover differential during the regular season may not be as good as everybody thinks they are. When matched against opponents who won't self-destruct, these teams lose one of their big edges.
The TD/Interception ratio is something I watched closely this year for the first time. I really liked how it gave you a sense of how a team gets things accomplished on the field (or doesn't). The game that jumps out here would be Kentucky/Clemson. You may remember how those numbers seemed backwards from perceptions. Clemson was supposed to be the high octane power, but had a TD/INT ratio of just 15-10 for the year. Kentucky was supposed to be a pretender, but had a stellar 28-7 mark. When you watched the Music City Bowl, you saw the story those numbers were trying to tell.
I wondered how teams performed this year when BOTH of those indicators lined up. What happened when a team with the better TD/INT margin also was WORSE in the turnover category? That would be a matchup of a team that had done a good job of moving the ball and scoring without risk...facing a team that may be overrated a bit because of the turnover situation.
THAT'S THE COMBO that went 78% in the bowls this year. There were nine teams who qualified. Seven of the nine covered their Las Vegas spreads.
POINTSPREAD WINNERS: San Jose State, Central Michigan, Oklahoma State, Missouri, Georgia Tech, USC, Louisville
POINTSPREAD LOSERS: Texas Tech, Nebraska
That's worth remembering for next year's bowls. And, it's probably something we all should be looking at on a weekly basis during the regular season. Go against teams who gain advantages from the turnover department when they're matched up against teams who throw a lot of TD passes but don't throw many interceptions. Makes a lot of sense when you think about it.
How did the bowls go for you? I can't say I was happy with the overall ATS performance in the Bowl Preview Series. The top percentage games did okay. Games rated at 65% or 70% combined to go 5-2 ATS, which is 71% success (winners were Kentucky, Georgia Tech, Oklahoma State, the Under in Auburn/Nebraska, the Over in Michigan/USC; losers were Oklahoma and Texas A&M). Anything at 55% or less though was basically just an imagined edge that randomly did or didn't pan out, or was just a clear error.
I did notice though that the forecasts did a better job of anticipating the "stat score" rather than the scoreboard result. You may have read about that formula in some of the Mess Hall threads at MajorWager.com. The formula doubles rushing yardage to give a sense of what the score "should have been" if you take out some of the random factors. Stat score: Rushing yards doubled...plus passing yards...multiplied by 0.67...divided by 15.
By my count, the bowl previews had 38 instances of teams or total suggestions rated at 60% or higher. If the stat score had been the final score, those games would have gone 22-14-2, for a 61% success rate. The stat scores WEREN'T the final scores, so that means NADA!
Well, it means that using stat indicators can give you a good sense of what the stats might look like in a game. Hard to know in a small sample if bad turnover luck was involved, or if there's a lot more of the bridge to cross in terms of turning stat assessments into pointspread winners. It was probably a mix of the two indicators this year. A little bit of bad luck, and a lot of learning still to do.
It's fair to say that the whole wagering world still has a lot of learning to do regarding college bowl predictions. The results from this year's MajorWager bowl contest were striking. The sum of the performances suggest that even the sharpest football wagerers don't really have a full sense of what leads to success or failure on the field.
In what is arguably the most handicapping-friendly format ever devised, a collection of 49 contestants (who paid $1,000 each to enter) managed to go 97-138-10 as a composite on their top five selections. That's 41% against the spread, even though:
*Contestants only had to isolate their favorite five games out of 32 bowls *Contestants had a few days to make their decisions against frozen lines *The field was limited to people confident enough in their abilities to pay the entry fee
All of the above only added up to 41% on top releases. The full slate of 15 selections for everyone worked out to 48% as a composite. You'd think the field would go about 50/50 if they had to pick all 32 games against the spread. Trimming more than half the slate actually worsened the record to 48%. Trimming 27 games and going with just the best five yielded only a 41% success rate.
I'm not saying everyone struggled. There were certainly some success stories. Taken as a whole though, "the power of the collective" was a big net negative. The wagering world has a lot to learn about what causes teams to play well, or play poorly in postseason action.
If you ask a bunch of poker professionals whether AK or QQ should be favored in a showdown, they'll all tell you that QQ is the favorite. If you ask a bunch of sports handicappers what's going to happen in a game, you don't get anywhere near a consensus. The oddsmakers do a good job, but even their numbers missed the final score by a TD or more in seven of the last 10 board games. This isn't a case of "the oddsmakers are unbeatable." This is "we're light years away from fully understanding all of the influences."
A collection of 49 informed wagerers couldn't go 97-138 on its cream of the crop selections in this kind of format, unless what the collective knew was dwarfed by what it didn't know.
Maybe the 78% combo uncovered today will help get all of us moving in a direction that yields consistent success for wagerers.
And, maybe we can start working together to find more successful indicators in all sports!