Pinny close lines

Alan Tongue

Pretty much a regular
Been stuck in planes and airports the past couple of days and been bored. I decided to look at how the season played out based on the Pinnacle close.

I used H2H prices and ran 15,000 simulations. Results are



Pinn Close.PNG


The most over rated team were the Browns, with the market expecting them to win just under 5 games. But that number will always be off when H2H odds are offered and they don't win a game. Next most overrated were the Broncos, with the market off by a tick under 3 games. Giants were just behind the Broncos.

Vikings and Eagles were the most underrated, with the market nearly 3.5 wins off their final number.

The teams most likely to make the Super Bowl from each conference are there.

What does this all mean? Not sure. If you subscribe to the theory that NFL closing prices are efficient, does that mean the teams that outperformed the market should regress next season, the opposite for teams that under performed (the betting market expectation, not the actual number)? Could do. I might run the numbers from previous seasons and see how they went. The season wins distribution were quite varied also
 
Looked at the 2016 numbers and compared them to the 2017 season just completed.

Capture1.PNG

Interesting stuff. The top 9 teams that outperformed the betting market in 2016 (by a total of over 25 games) won less games the following season. These ranged from 1 less win to 8 less wins (35 wins in total, or 3.89 wins on average).

Of the 13 teams that performed below the market, 11 of them won more games the following season. Of the other two teams, the Jets won the same amount and Cleveland went from 1 to 0. These 11 teams won between one and 7 more games (47 in total, average of 4.27).

Of the massive sample size of 32 teams and one season it shows the teams that outperformed the betting market regressed the following season, while the teams that under performed were much better the next season. I'll run more and see if this continues
 
pretty good stuff, Alan. That's an approach I hadn't really thought of, but I like it. Would like to see how well this holds up going back a few years. Would be very interesting for baseball as well.
 
2015 and 2014 had similar trends

Pinn Close2015.PNG

Pinn Close2016.PNG


Over the 3 year sample there were 16 teams that out performed the betting market by at lease 2 wins. 15 of these 16 teams had less wins the following season (average 4.4 wins)

Pinn Close2+.PNG

18 teams under performed the market by at least 2 games, and only 2 had a worse record the following season (Cleveland from 3 to 1 to 0). The remaining 16 teams won on average 3.75 games the next season.

Pinn Close2-.PNG

So looking at this seasons numbers, MIN, PHI, LAR, CAR and PIT should see a sizable cumulative drop, while HOU, NYG, DEN and CLE should see a cumulative increase

Pinn Close2017.PNG

How to profit? Well, we don't know what the season win totals would be, but if they are close to the actual numbers from this season I'll be looking at some boxed parlays. If you are assessing teams for next season maybe give these numbers a tick up or down
 
pretty good stuff, Alan. That's an approach I hadn't really thought of, but I like it. Would like to see how well this holds up going back a few years. Would be very interesting for baseball as well.

Baseball would be interesting. It will be a bit more work to use teams and starting pitchers
 
Back
Top