Free Money? Bet Against Regression to Mean for the Pats

[SlowGettingUp]

So here's an interesting article in the NY Times comparing a betting site's projected over/under on wins to last year's records.

Not surprisingly, the betting site banks on regression to the mean - that last year's bad teams will do better and last year's good teams will do worse. That will undoubtedly prove correct in aggregate. But does anyone here seriously think the Pats will score under 11.5 wins?

Will Your N.F.L. Team Get Better or Worse? You Can Bet on It
https://www.nytimes.com/2017/05/02/sports/football/nfl-schedule-records-betting.html?

(The reason regression to the mean happens in any process involving chance is that some prior underperformance simply involved bad luck and some overperformance good luck. On average, the luck will even out this year).

 

[TB_Helmet]

I don't know about you guys but I'm thinking weekend road trip to South Point Casino! (Wherever that is).

 

[IndianPat]

11.5 wins is a decent prediction. 12 wins are tough to get, even though Pats make it seem routine. The fan base that should be really pissed off are the Dolphins who are projected to win 7 games. A bet on the Dolphins to beat the projected outcome carries the best risk/reward potential.

 

[stinkypete]

11.5 wins is a decent prediction. 12 wins are tough to get, even though Pats make it seem routine. The fan base that should be really pissed off are the Dolphins who are projected to win 7 games. A bet on the Dolphins to beat the projected outcome carries the best risk/reward potential.

I don't know if 7 is an easy bet for the Dolphins. The AFC West and NFC South look like the two toughest divisions in football right now. The Patriots should win at least 6 or 7 of games against those divisions. I don't think Miami is built to keep pace with that schedule.

11.5 is a deflated number. The numbers may suggest that the Patriots regress 2.5 wins, but they've set 12 wins as the mean since 2010.

 

[Ring 6]

11.5 wins is a decent prediction. 12 wins are tough to get, even though Pats make it seem routine. The fan base that should be really pissed off are the Dolphins who are projected to win 7 games. A bet on the Dolphins to beat the projected outcome carries the best risk/reward potential.

Patriots are more likely to win 12 or more than the Dolphins are to win 8 or more.

 

[patsfan13]

Tom Brady has Averaged 12.47 wins for every 16 games he has played, So 12.5 games is TB12's mean, he regresses to.

 

[brdmaverick]

But does anyone here seriously think the Pats will score under 11.5 wins?

Your statement seems to imply that winning less than 12 games would be crazy.

For the record, I agree with your statement. I think 12 wins is the floor with this team, but just for fun let's soak in the historical context of this statement.

Here it is in 2017 we are saying that 12 wins is our baseline FLOOR when prior to Brady/Belichick the team had NEVER had 12 wins.

 

[SlowGettingUp]

Your statement seems to imply that winning less than 12 games would be crazy.

Well not crazy, but less than a 50% shot. And of course it could happen - a rash of injuries at one position (OL?), a couple of tipped balls, some extra fumbles that are lost, and some lucky catches by opposing receivers can turn around a lot of games that we "should" win.

But the team going into the season with a healthy Gronk is clearly stronger than the one that won the Super Bowl. Clearly stronger at TE, stronger at receiver, stronger at RB, stronger at corner. So I think 12 games is indeed a realistic floor.

 

[Ice_Ice_Brady]

11.5 wins is a decent prediction. 12 wins are tough to get, even though Pats make it seem routine. The fan base that should be really pissed off are the Dolphins who are projected to win 7 games. A bet on the Dolphins to beat the projected outcome carries the best risk/reward potential.

They can be "pissed off" all they want. The Dolphins were outscored last season and went something like 8-2 in close games. 7 is the right number. I've been saying this offseason that the Dolphins are likely to finish below .500. They addressed nothing this offseason, content with their luck boosted record.

 

[BobDigital]

I will be more inclined to bet on 12 wins or more than 11 or less for this team. All you want to know is how Brady looks before the year starts. If he looks good bet heavily on the over.

Yes 12 wins are generally hard to get but to get against it is or more a much longer shot IMO.

 

[n1997y]

I will be more inclined to bet on 12 wins or more than 11 or less for this team. All you want to know is how Brady looks before the year starts. If he looks good bet heavily on the over.

Yes 12 wins are generally hard to get but to get against it is or more a much longer shot IMO.

You really don't even need to worry about how Brady looks. Point of keeping JAG is that there is another QB on the roster who is quite capable of running the offense with the weapons they have. No reason to flip from betting over to betting under just because he's starting.

 

[RayClay]

Excellence in football is not a random occurrence.

 

[stinkypete]

Excellence in football is not a random occurrence.

All statistical distributions tend toward a mean. All statistical distributions have outliers as well.

Pythagorean win projections and other regression formulas are a good predictor for most NFL teams, but the data seems to force teams into a bell curve that ranges from 3 wins to 11 wins. One could conclude that, based on based on these projections, that a 12+ season is an outlier.

These projections fail when measuring the greatest outlier in NFL history - the Belechick/Brady Patriots. The team has won 12+ games in 11 out of the last 14 seasons. Clearly, the "mean" as it applies to this franchise must be defined differently relative to the rest of the NFL.

I

 

[RayClay]

All statistical distributions tend toward a mean. All statistical distributions have outliers as well.

Pythagorean win projections and other regression formulas are a good predictor for most NFL teams, but the data seems to force teams into a bell curve that ranges from 3 wins to 11 wins. One could conclude that, based on based on these projections, that a 12+ season is an outlier.

These projections fail when measuring the greatest outlier in NFL history - the Belechick/Brady Patriots. The team has won 12+ games in 11 out of the last 14 seasons. Clearly, the "mean" as it applies to this franchise must be defined differently relative to the rest of the NFL.

I

Following one team who picks their players and their coaches is not a statistical distribution.

 

[RayClay]

All statistical distributions tend toward a mean. All statistical distributions have outliers as well.

Pythagorean win projections and other regression formulas are a good predictor for most NFL teams, but the data seems to force teams into a bell curve that ranges from 3 wins to 11 wins. One could conclude that, based on based on these projections, that a 12+ season is an outlier.

These projections fail when measuring the greatest outlier in NFL history - the Belechick/Brady Patriots. The team has won 12+ games in 11 out of the last 14 seasons. Clearly, the "mean" as it applies to this franchise must be defined differently relative to the rest of the NFL.

I

If there is a statistical mean that teams regress toward, it shouldn't matter what team.

I'm sure you can find a lousy team who's record also fails to regress to a "statistical mean." It's like saying a grandmother and a world series of poker champion will eventually regress to a mean level of success at that game.

 

[Adam Seward]

All these posts and no link to the free money!?!?

 

[BobDigital]

You really don't even need to worry about how Brady looks. Point of keeping JAG is that there is another QB on the roster who is quite capable of running the offense with the weapons they have. No reason to flip from betting over to betting under just because he's starting.

.

Well U think a lot of JG but with Brady i think 12-4 is the absolute floor if he is right. With JG I don't feel the same way. I think the offense and team will be fine but maybe they go 11-5 or 10-6. There will be a drop off between what we are use to getting from Brady and JG of course we just don't know how big it will be yet.

 

[TruthSeeker]

Since the author brought up the Patriots several times, I thought he was going to point out their record vs. predictions in the Brady era (which shows no "regression to the mean"). Alas, no such thoughtful analysis.

 

[luuked]

11.5 wins is a decent prediction. 12 wins are tough to get, even though Pats make it seem routine. The fan base that should be really pissed off are the Dolphins who are projected to win 7 games. A bet on the Dolphins to beat the projected outcome carries the best risk/reward potential.

I disagree competely.

I think they completely over-performed last year in terms of pulling out wins in close games and could have easily gone 5-10 or worse if the football had bounced differently a few times (49ers fourth down RZ stop, being not able to do anything agains the Rams for 57 minutes, needing 4 INTs from Rivers to pull out a win against the Chargers, kickoff return for TD against the Jets late in the fourth when the game was tied, Cardinals being stopped multiple times on offense by torrential rain while the Dolphins could drive for the winner, the Browns missing game winning FGs and some more..).

And then they got totally stomped by a Steelers team in the playoffs which in turn got stomped by us.

Based on last season I think Miami might be one of the more overrated teams going into this season. They have a lot of areas to improve upon to even repeat a playoff berth. I honestly think they are a .500 team that might regress a bit to the mean in terms of success in close games.

 

[RayClay]

All these posts and no link to the free money!?!?

I was hoping to get an offer. BTW mean in the Brady era is 12.25 if I got it right.

Where you been, by the way? We could use a seasoned LB. Still announcing games in Chinese?