NFL Analytics for 2021 Season

[KillerGoose]

This thread is going to be used to house some fun numbers projects that I am running this season. To begin, we will start with a Pythagorean projection of all NFL teams, Points over Expected (PoE) for all teams and SRS for all teams. This will probably grow into other metrics as the season progresses and my projects do as well, but that is the starting point. Please note, these numbers don't tell us a story YET. They won't begin to take much meaning until next week because our dataset is too small right now.

Pythagorean Projections  - If you're a baseball fan, you'll be familiar with this. This is a formula that can be used to predict what a teams final record will be by using points for and points against. The formula looks like this...

A = Points for/Points against

(A)^2.37/(A)^2.37 + 1

To build confidence in this, we can look at last season. The 2020 Cincinnati Bengals scored 311 points and allowed 424 points. If you take 311/424, you're going to get 0.733. Now, take 0.733 and apply an exponent of 2.37. This will bring a result of 0.478. Now, take 0.478 and divide it by 1.478 and you will get 0.32. This model predicts that the 2020 Cincinnati Bengals won 32% of their games, or five games. In reality, Cincinnati won 28% of their games, or four games. The model predicts that the 2020 Oakland Raiders won 44% of their games (seven games). In reality, they won 8. You're going to find some teams that 'overperformed' or 'underperformed' by a couple of games due to the limited number of games in an NFL season.  The Steelers, Chiefs and Bills are all an example of this. The model is pretty accurate at getting within a game or two of how a team should finish, usually. The Browns are the major exception. The model would predict them to be a .500 team, roughly, and they won 11 games plus a playoff victory. That's football, however. Volatility is real and 'any given Sunday' is as well. Attached is what the predictions currently look like. Keep in mind, these mean NOTHING right now. We don't have enough data to say anything with any confidence, because there has only been two games. 





Points over Expected (PoE) - This metric is used to try to describe how well a team is performing relative to expectations. Say it is week 10, and you have Team A vs Team B. Below are their points for and points against...

Team A || 100 points for || 140 points against

Team B || 160 points for || 120 points against

By having these numbers, we can predict a final score of 15-11 in favor of team B. The 15 points is the average of team B's points per game and team A's points allowed per game (16,14). The 11 points is the average of team A's points per game and team B's points per game allowed (10, 12). Now, if team A wins the game 20-13, then team A is going to see their offensive PoE increase by +9 and their defensive PoE decrease by -1 (negative numbers are good). So, overall, team A's PoE would increase by 10. 

I am unable to show an example of what this looks like currently because you need at least three games to get the data that you want for this, but this will come out next week. 

SRS - Otherwise known as Simple Rating System, this takes into account your margin of victory (MOV) and strength of schedule (SOS) to get a simple team rating. The SOS is calculated by adding up the MOV for all opponents that have been played, but not future opponents. So, this is descriptive only. Please see attached for current SRS numbers. Again, these won't start meaning much until next week where we start building more data into the formulas. You're going to notice that despite the MNF game not being played, it will show the Lions and Packers as having played two games. I did that just to get this template out on my work from home day. I wasn't worried too much about the accuracy due to the aforementioned dataset issues. This is primarily to familiarize everyone with what we are going to be working with.




As we move forward, I will probably create some Tableau presentations of already gathered data such as yards per play, points per play, pressure percentage allowed etc. 

If anyone has any questions, please let me know. Happy to discuss further. 

[KillerGoose]

I am adding another metric that I will keep track of in this thread. This metric is an adjusted passer rating and is based upon work that Brian Burke did along with some adjustments that Wayne Winston made, the author of "Mathletics: How Gamblers, Managers, and Sports Enthusiasts Use Mathematics in Baseball, Basketball, and Football". The formula is as follows...

(1.543(passing yards - sack yards)/passing attempts) - 50.0957(interceptions thrown/passing attempts)

The coefficients of this formula come from Brian Burke's work, which is a bit dated at the moment (2007) but the results still hold up well. If I can find the time, I would like to re-run the linear regression to get new coefficients for this formula to attempt to make it more accurate for today's game, but for now, this will work. The result of this formula brings us an 'adjusted' passer rating which attempts to paint a picture of which QBs are helping their team win more games via their passing abilities. 

Here are the current rankings by this adjusted rating.




And for reference, here are last seasons ratings. 

[EatonFan]

I love numbers, but this may be too geeky even for me!

[KillerGoose]

Here are some of the numbers following week 3. I have made an adjustment to my QB rating to factor in TD%. The base of this work was done by Brian Burke, but I ran a regression on wins and TD% and found there was strong enough correlation to include the coefficient into our formula. Quite frankly, I think these rankings make much more sense. And here we go...

Pythagorean Projections - Week 3



This data is still very early BUT our Cincinnati Bengals are playing at a predicted 11 win pace if they can keep this performance up. We are still dealing with a sample size issue, which should be apparent with the model predicting seven wins for GB and only eight wins for KC/Baltimore. Through the first three weeks, we have a surprise with Denver and Carolina making up two of the top three spots (Denver playing at a 16-1 pace!) I certainly don't expect either of these teams to keep this pace up but they are playing much better football this year than they did last year. 

SRS - Week 3 






Our Cincinnati Bengals are also ranked as top 10 team through week 3, logging a 7.4 SRS rating. Remember, SRS is trying to tell you how many points BETTER that team is against a theoretical average NFL team. Keeping this in mind, it really puts into perspective how bad Atlanta, New York and Detroit has played through these first few weeks. Carolina again is a surprise team topping the charts along with a curious visit from the Vikings. Why are the Vikings ranked higher than Cincinnati despite losing to them? Because they played Cincinnati AND Arizona (both top 10 teams by this model) close while borderline thumping the Seahawks. Essentially, it is saying that Minnesota has had a tougher schedule and has been very competitive throughout. Like everything else, this will normalize throughout the season.

Adjusted QB Rating - Week 3






The changes that I made to this formula have made this list make more sense. The previous formula wasn't factoring in TD% to this rating. That formula was from Brian Burke, and the data is outdated (I still need to redo the regression on this). I did find time to do a regression on TD% and wins and found a strong enough correlation to make the change, so we now have a stronger (in my opinion) formula that tells you who is playing well and who isn't. Burrow's stellar TD% launches him into the top 10 (previously 24th) and there is a surprise visit from Heinicke to round out the top 10. Winston is a curious addition in the top 5 but it is likely due to sample size. He hasn't been asked to do much but has thrown seven TDs on 63 attempts for the highest TD% in the league (11.3%).

I am still trying to build a framework to keep track of other metrics that would be fun to look at as the season goes on and this should be much easier to do now that we have a large enough dataset to matter. If anyone has any questions, feel free to ask.

[*bfine32*]

[Bengalitis]

I knew gambling was legal now in the NFL, but scientific research in the NFL, now THAT i did not expect! WOW!

pst..pst... give me the lottery numbers..

[KillerGoose]

Week 4 - Pythagorean Projections



Denver has relinquished their top spot after receiving a 23-7 shellacking by the Baltimore Ravens, leaving the Buffalo Bills atop our projections with projected 16-1 record. One thing to keep in mind when judging the Bills, while they have played fantastic football these first four weeks, they have also had the easiest strength of schedule. This doesn't completely diminish their accomplishments; they have a margin of victory of 22 points per game which is incredible in the NFL, but they will certainly level out as the season continues. They have a showdown with the Chiefs this weekend. 

Cincinnati comes in with a projected 11-6 season while having the 9th easiest schedule in the league. Going by a projected wins and hardest SOS combo, Dallas and LAC have the most projected wins (11) with the 4th and 5th most difficult schedule in the league so far. Kansas City is currently projected for 9 wins with the third hardest schedule so far. 

Week 4 SRS






The Bills top our SRS rankings as well, in large part to their insane margin of victory. The aforementioned Dallas and Chargers squeeze into the top five right after the Browns. Cincinnati is currently ranked the 14th best team due to their weak SOS.

Week 4 Adjusted QB Rating






I made another tweak to this formula, but only aesthetically. I changed the constant that I multiply by to give this a more 'normal' scale to work with. The ranking methodology is the same as last week. Mahomes has taken the top spot with a ridiculous five TD performance this past weekend, leading the Chiefs past the Eagles 42-30. Burrow jumps up a spot and sticks into the top five with Taylor Heinicke sticking around in the top 10 after a 290 yard, 3 TD performance in a win over the Falcons. 

[kronadon]

Thanks for doing this. I would vote for this to be sticky.

[KillerGoose]

(10-05-2021, 12:10 PM)kronadon Wrote: Thanks for doing this. I would vote for this to be sticky.

You're welcome, I enjoy doing it! I'm happy you enjoy it.

[Truck_1_0_1_]

Would be perfectly content with an 11-win season

[KillerGoose]

Week 5 Pythagorean Projections




The Buffalo Bills top the list, playing at a blistering 16-1 pace after their domination of the Chiefs this past weekend. As we get deeper into the season, we are starting to see some of the top teams take shape and it is looking like Arizona, Dallas, New Orleans and Tampa Bay are potentiasl contenders along with Buffalo. Our Cincinnati Bengals dropped to a 10 win pace after the loss to Green Bay, still sporting a positive point differential. 

Week 5 SRS





The top five in our SRS rankings are very similar to our projected wins with Buffalo, Dallas, Arizona and Tampa Bay making up four of the top five. The one difference is the Chargers coming in at #4 with the Saints ranking 8th. Why this difference? The Chargers have played Dallas, Kansas City, and Cleveland. All three of these teams rank as a top 10 team by this model, and the Chargers beat two of them. Cincinnati ranks as a completely average team by coming in at 15th with a zero SRS. Remember, SRS is telling you how many points that team is better than when comparing to a theoretical average team. 

Week 5 Adjusted Receiving Yards - NEW 






This is a metric I have been spending some time on to try to get right. This is taking a receivers yards and TDs to combine them into one number using the formula provided by PFR for adjusted passing yards (receiving yards + (20 * TDs)). This shows you an overall number, which is sorted by largest to smallest and then shows the percentage of passing offense that player has contributed. Ja'Marr Chase, Davante Adams and Deebo Samuel are having incredible years so far, all accounting for over 50% of their respective teams passing offense. Corey Davis and Darnell Mooney top the list when looking at percentage of offense contributed, with Davis being 74% of the Jets passing offense and Mooney being 84%(!!!) of the Bears offense.

Week 5 Adjusted QB Ratings






I made one final adjustment to this formula and we are going to let it ride from here until the end of the season. I re-did the regressions on the metrics Brian Burke used in order to make it more modernized. I was using his coefficients, which were from the early 2000's. We now see Brady and Herbert catapult into the top 10 with huge performances this past weekend and Burrow drops out of the top 10 all the way down to 14th. Why did this happen? Well, when I re-ran the numbers, I found a larger penalty for interceptions. This could likely be due to how pass heavy offenses are, and how much better offenses are in general. Because of this larger penalty, the formula drives Burrow down the list because of his interception percentage, which ranks as the 4th highest INT % in the league. Only Zach Wilson, Davis Mills and Trevor Lawrence have a higher INT % than Burrow. 

[KillerGoose]

Week 6 Adjusted Ratings






Stafford, Murray and Prescott continue balling out this season with dominating week six performances. Winston sticks around at #3 due to low volume and having a bye week. Wilson is at the top but will drop off next week more than likely due to his injury taking him out of the minimum attempts filter. Burrow jumps a couple of spots from 14 to 12 as he puts up a strong game, but the turnovers keep coming. Burrow now has the third highest INT% in the league, jumping Trevor Lawrence.

Week 6 Adjusting Receiving Yards






Cooper Kupp continues his domination of the league after notching a 133 yard and two touchdown performance this past weekend. Ja'Marr Chase, our rookie phenom, has turned into a one man show. He ranks fourth in the league with 653 adjusted receiving yards and has accounted for 63% of the Bengals passing offense. Burrow is looking for Chase and no one can cover him - yet. Tyler Boyd and Tee Higgins, on the other hand, are not producing very much at the moment. Boyd is ranked as the 55th WR out of 87 with 310 adjusted yards and Higgins is ranked as the 76th WR out of 87 with 234 adjusted yards. 

Week 6 Pythagorean Projections






The Bills still top our projected records with 14 wins after losing to Tennessee last night but the Cardinals are hot on their trail. Arizona is undefeated at 6-0 and just finished absolutely obliterating the Browns in a 37-14 rout this past weekend. In the Queen City, the Bengals find themselves remaining consistent as they track towards an 11 win season. A huge showdown in Baltimore this weekend as the Bengals take on the Ravens to decide who gets to take early control of the division. 

Week 6 SRS 






The Cardinals unseat the Bills as the top rated team in the league with the Rams, Cowboys and Ravens finishing out the top five. Minnesota, despite being 3-3, rank as the 7th team after an OT victory against the Panthers, who are the 13th ranked team by SRS. Cincinnati breaks into the top 10 after they smote the Lions this past weekend. 

[*Nate (formerly eliminate08)*]

I have to bump this thread as it is interesting and too much for me to take in at the moment lol

[kronadon]

Hope there is an update to this soon.

[Dr.Z]

Just now seeing this thread.  I'm going to have to get you to do some analytics on some market trades for me.   My options trades have been very unprofitable this week so far. 

[KillerGoose]

Week 7 Adjusted QB Ratings






After a legendary game against the Ravens, Joey Bengal catapults himself into the top 10 and lands at #7, BARELY missing out on the #6 spot. I don't believe that Jameis is going to stay in the top 10 for long as the season goes on. I have dropped Russell Wilson from the list due to his injury (he previously was the top QB). Tom Brady continues defying all logic as he sits at #4, continuing his dominance of the league. On the other end of the spectrum, Justin Fields still sits at the bottom of the list after a miserable game, continuing to solidify himself as far-and-away the worst starting QB in the league. 

Week 7 Adjusted Receiving Yards






We are going to continue our Bengals lovefest with the highlighting of Ja'Marr Chase. He ranks second in the league with 874 adjusted receiving yards while accounting for just over 50% of the Bengals passing offense. The most impressive thing? He is doing it while only being targeted 51 times (24th in the league). To compare this to Cooper Kupp, who leads the league in adjusted yards, Kupp is averaging 12.2 adjusted yards per target. Chase is averaging 17.13 adjusted yards per target. In short, he is dominating the NFL. On the other hand, Kupp is producing over 57% of the Rams high-flying passing attack. 

Two names to keep an eye on as the season progresses is Deebo Samuel (49ers) and D.J. Moore (Panthers). Despite being on disappointing teams, these guys are putting up great numbers.

Week 7 SRS


[img]blob:https://imgur.com/526a6d9c-0fa6-431f-932c-9199ab200d03[/img]



The still undefeated Cardinals sit atop our SRS rankings after a 31-5 shellacking of the Texans. Cincinnati has soared to 6th after two dominant weeks in a row. There may still be plenty of season to go but one thing is for sure - this team is playing like a top team and making their case to be called a contender. By SRS, the favorites for a SB appearance would be the Cardinals and Bills if we ended today. 

Week 7 Pythagorean Projection






After beating the Ravens this weekend, Cincinnati is on pace for a 12 win season (who would have imagined?) and are currently outpacing the Rams, Cowboys AND Ravens. 

[*Nate (formerly eliminate08)*]

Awesome thread, thought we would of saw a huge jump after the Ravens game. Hope all Bengals keep climbing.

[KillerGoose]

Adding a new metric to the mix, another QB metric. We are going to be looking at EPA (Expected Points Added). I will be posting a list and then a graph of plotted data with EPA and my adjusted rating. Here is the top QBs in the league by EPA...



And the data plotted...

[grampahol]

Is it just me or does anyone else think these math geeks should be shot on sight?  
Actually I'm glad someone thinks about this stuff because I've always sucked with number. I'm good with abstract concepts, something math generally is not..

[KillerGoose]

(10-31-2021, 12:07 PM)grampahol Wrote: Is it just me or does anyone else think these math geeks should be shot on sight?  
Actually I'm glad someone thinks about this stuff because I've always sucked with number. I'm good with abstract concepts, something math generally is not..

I used to struggle with math in high school, but once I got to college I started loving numbers. Marrying that with my favorite sport is a ton of fun for me. A lot of football conversations tend to be abstract and I don't like that when there are definitive answers available  

Now, here is some winning percentages for this weeks games! Predicted winners and their chance of victory...