We enter the stretch run of the season coming off another losing week, unfortunately. Losing isn’t fun, but lady luck dictates it will happen sometimes, and those are the times when people might be tempted to question the process.
Do I believe that our Massey-Peabody model is perfect and there are no possible improvements? Of course not. There are all sorts of things we don’t account for: injuries (aside from quarterbacks) offseason transactions and bad calls by refs to name a few. But we like that our framework is simple and understandable (even if the details and code are not).
The market gets more efficient every year–I truly believe that–but I’m still confident that we have an edge. Over a small sample (and one season certainly qualifies as such) line move value is a better predictor of success than win-loss record; I’ve said this countless times in past years. I haven’t examined the closing line value of Massey-Peabody picks closely, but I still engage in my (ahem) day job every week, and have noticed lines moving just as much in my direction.
An understanding of randomness is a huge part of our model. If the public properly appreciated the role randomness played (in football and in their lives) I would need to change careers. A misunderstanding of randomness (and failure to account for survivorship bias) contribute to the tout industry flourishing, as countless people get suckered into buying picks from people who may as well be flipping a coin. Hell, there’s a (I’d like to think very remote) chance that Massey-Peabody numbers hold no *real* predictive value, and we’ve been the ‘false positive’ all these years.
Understanding the role randomness plays in results is just as important as knowing its role in fumble recoveries. We live in a probabilistic world. In the end, a season such as this one is probably beneficial to one of the missions of this site: proselytizing better, unbiased reasoning. Randomness is as ubiquitous as the air we breathe. Our (awful) record this season is unlikely, but so was having five straight seasons of profitable picks. We will spend time looking for ways to improve our model this coming off-season (like we do every off-season) but remember that our model has not changed in the last year; it’s still the same under the hood. If you liked us last year, you should still like us now (unless you liked us because of our success in a small sample, in which case you were probably fooled by randomness in a different way).
Process over outcome, baby. That’s what we believe in.
On to the Week 13 picks.
Big Plays (5-9-2 YTD)
- Minnesota +3 (+100) vs. Dallas [MP Line: DAL -0.6]
- Buffalo +3 (+100) at Oakland [MP Line: OAK -0.7]
- Carolina +7/+7.5 at Seattle [MP Line: CAR +5.4]
- New York Jets +1.5 vs. Indianapolis [MP Line: NYJ -3.5]
Other Plays (9-9.5-2.5 YTD)
- San Francisco +1 at Chicago [MP Line: SF -2.0]
Break-Even or Better (unofficial leans) (8-12-2 YTD)
- Green Bay -6.5 vs. Houston [MP Line: GB -8.6]
- New England -13 vs. Los Angeles [MP Line: NE -15.2]
- San Diego -3.5 vs. Tampa Bay [MP Line: SD -5.9]
Massey-Peabody True Line on Remaining Games
- Denver -3.1 at Jacksonville
- Atlanta -3.4 vs. Kansas City
- Cincinnati -1.0 vs. Philadelphia
- New Orleans -6.9 vs. Detroit
- Baltimore -2.2 vs. Miami
- Arizona -2.8 vs. Washington
- Pittsburgh -6.3 vs. New York Giants