Thursday, February 11, 2021

2020 NFL Game Predictions: Season Recap

Welcome!  I intend this to be an ongoing project of predicting NFL game outcomes, point spreads, and final scores.  My hope is that my models will improve over time as they become more sophisticated and use better data.  I will try to regularly publish predictions to keep myself accountable, and will make observations along the way about what is working and what isn't.  See below for latest updates.  Enjoy!




Previous 2020 Predictions:

________________________________

Season Recap:

Here we are at the end of the 2020 season.  I have run predictions from week 2 all the way through the Super Bowl.  I have used two models to make these predictions, I have compared with the models at 538, and I have made my own predictions about particular games.  So how did I do? 

Model Statistics

Let's look at the stats:
  • Whole Season accuracy % (95% Confidence Interval)
    1. 538 QB: 68% (0.62, 0.74)
    2. Personal: 66% (0.60, 0.72)
    3. 538 Traditional: 65% (0.59, 0.71)
    4. LR Model: 63% (0.57, 0.69)
    5. RF Model: 63% (0.56, 0.69)
  • Regular Season accuracy % (95% Confidence Interval)
    1. 538 QB: 68% (0.62, 0.74)
    2. Personal: 67% (0.60, 0.72)
    3. 538 Traditional: 66% (0.59, 0.71)
    4. RF Model: 63% (0.56, 0.70)
    5. LR Model: 63% (0.57, 0.69)
  • Playoffs accuracy % (95% Confidence Interval)
    1. 538 QB: 62% (0.31, 0.86) - not significant
    2. 538 Traditional 62% (0.31, 0.86) - not significant
    3. LR Model: 62% (0.31, 0.86) - not significant
    4. Personal: 62% (0.31, 0.86) - not significant
    5. RF Model: 54% (0.25, 0.81) - not significant
538's QB adjusted model performed the best at 68% accuracy over the course of the whole season.  My personal predictions were second best at 66% accuracy.  The 538 traditional model took third at 65% accuracy.  My LR model barely edged out the RF model with both getting 63% accuracy.  So while my home grown models did NOT beat the professional models, they still did pretty well.  Sure, it would be nice to beat the professional models but that is unlikely without some major investment of time/energy in this beyond what is already being done.

The whole season and regular season results are statistically significant (two-tailed binomial test) while the playoff results in isolation are not. But that is not surprising with so few playoff games.  Also, this exercise does suggest once again that 70% seems to be an upper bound beyond which no model can accurately predict due to the randomness inherent in the game.  Anything above that would be unsustainable luck.

Trends

Week to week the predictions tended to stay between 40% and 80% accuracy with an average of just above 60%.  That changes in the playoffs with divergence in both directions, but throughout the regular season the models were fairly consistent.  A couple exceptions:
  • Week 2: 538 QB got 15 of 16 games correct, the highest weekly accuracy during the regular season of any model.
  • Week 9: the LR model gets 12 of 14 games correct.
  • Week 11: all of the models (except for the 538 traditional) performed worse than 50%, with the LR model going down to 29% (4 of 14 games correct).

The models start off in a larger spread of accuracy at the beginning of the season, but over time the cumulative accuracy stabilizes until the models are in a tight cluster by the end of the season.


Probability vs. Actuality of Win

How does probability of a win relate to actuality of a win?  Presumably, the greater the probability of a win, the more likely the actuality of a win.  Did this happen in the models?

In the LR model, this turned out to be true.  The average predicted probability for predicted wins that turned into actual wins was 66% compared with 61% for actual losses.  Similarly, the probability for predicted losses that turned into actual losses was 34% compared with 39% for actual wins.  That is, probabilities that were higher did tend towards actual wins while probabilities that were lower tended toward actual losses.

The LR model did tend to overestimate the probabilities, predicting an average of 36% for losses and 64% for wins.  In actuality, the average predicted probability for actual losses was 44% and 56% for actual wins, much closer together than what the LR model was predicting.



If I break the probabilities into clustered groups, we have the below view.  Generally speaking, the clusters that had the lowest predicted probabilities had the lowest actual probabilities.  But this was not always true.  For example, for teams with a predicted probability between 15% to 20% of winning a game, the actual probability was 36%, although the 20% to 25% category was only 31% actual probability.


How did the RF model do?  For one, the RF model did not have such extreme probabilities.  It produced an average probability of 38% for predicted losses that were actual losses compared with 41% for actual wins, and 62% for predicted wins that were actual wins compared with 59% for actual losses.  Thus, there is much less separation between predicted outcome probabilities (39%, 61%) and actual outcome probabilities (46%, 54%).




With the clustered probabilities, the RF model does a better job matching the predicted probability of outcome with the actual probability of outcome this season.  The lowest predicted probabilities do tend to match the lowest actual probabilities.


How do my models compare with 538's QB Adjusted model, which produced the best results this season?  The 538 QB Adjusted model looks a lot more like the LR model than the RF model.  When predicted to be a loss, the actual losses had an average probability of 34% compared with actual wins of 38%.  Similarly, predicted wins that resulted in actual wins had an average probability of 69% compared with 64% actual losses.  



Likewise, the 538 QB model tended to generally have predicted probability clusters that mirrored the actual probabilities, but not as well as the RF model.  Again, this was more similar to the LR model.


In short, all models tended to have average predicted probabilities that, more or less, mapped correctly to average actual probabilities this season.  Probability of win did tend to correlate with actuality of win.

Probability, Point Difference, and Team Score

What is the relationship between the probability of a win, the actual point difference, and the actual team score?  The relationship between the probability of a win and the actual point difference is not strong in my models, which is not too surprising as the models are designed to predict a win (binary: 1 or 0).  

The LR model gives an R^2 value of 0.12 with an equation of PD = 27.774*prob - 13.887 while the RF model gives an R^2 of 0.13 with an equation of PD = 39.038*prob - 19.519.  So a probability of 0% will predict a rounded point difference of -14 (LR) or -20 (RF), while a probability of 100% will predict a rounded point difference of 14 (LR) or 20 (RF), with both models predicting a point difference of 0 for 50%.  Not great but its something.




Team score is much more promising.  By forcing the intercept of the models to 0, the LR model yields an R^2 of 0.82 with an equation of score = 45.467*prob, while the RF model yields an R^2 of 0.84 with an equation of score = 47.15*prob.  Thus, a probability of 0% will predict a score of 0 in both models while a probability of 100% will predict a rounded score of 45 (LR) or 47 (RF).




So there is a fairly strong relationship between the probability of a win and the team score, although the probability of a win and the resulting point difference are not strongly related (and a different model would be required).

Team Model Stats

So much for the models themselves.  Now on to the teams.  What did my models think of  each team throughout the season in terms of their likelihood of winning each game?  Below are some interesting stats:

Regular season
  • The Chiefs had the highest expected value (LR: 10.73, RF: 8.30) for total games.  The Jaguars had the lowest (LR: 3.71; RF: 4.08)
  • The Chiefs had the highest average probability (LR: 72%; RF: 64%).  The Jaguars had the lowest (LR: 25%; RF 31%)
  • The Chiefs and the Vikings had the highest LR max predicted probability (92%) for a single match up.  The Jets had the lowest (50%), meaning they were never predicted to win a game with more than 50% likelihood.
  • The Colts had the highest RF predicted max probability (86%) for a single matchup.  The Broncos had the lowest (51%).
  • The Packers had the highest minimum LR predicted probability (49%), meaning that they were never predicted to have a probability of winning a game less than 49%.  The Jaguars and the Panthers had the lowest minimum predicted LR probability at 8%.
  • The Chiefs had the highest minimum RF predicted probability (54%).  They were never predicted to lose by the RF model.  The Ravens and the Packers were also never predicted to lose by RF.  The Jaguars had the lowest minimum predicted RF probability at 14%.
  • The Raiders had the highest Actual Wins - LR Predicted wins with 4 (predicted 3 wins; 7 actual).  The Cowboys had the lowest with -5 (11 predicted wins; 6 actual).  The Vikings also had -5 (12 predicted; 7 actual).
  • The Browns, Giants, and Lions had the highest Actual Wins - RF Predicted wins with 4.  The Ravens and the Vikings had the lowest with -3.
  • The Buccaneers had the highest Actual - LR Probabilities with 3.39.  They were expected to have 7.61 wins but ended up with 11.  My model clearly underestimated them.  The Bills had 3.37 (8.63 expected, but 12 actual).  The Jaguars had the lowest with -3.71 (3.71 expected, 0 actual).
  • The Chiefs had the highest Actual - RF Probabilities with 4.70 (8.30 expected, 13 actual). The  Bills had 4.69.  The Jaguars had the lowest with -4.08 (4.08 expected, 0 actual).
In short, the models judged the Chiefs to be the best team during the regular season and the Jaguars to be the worst team of the regular season.  The Raiders, Buccaneers, and Bills overperformed.  The Vikings and Ravens underperformed.


Playoffs
  • The Buccaneers had the highest LR expected value (1.99), but they played the most games.  The Bills were second (1.62) and the Chiefs were third (1.61).  The Bears were the lowest (0.32) with only one game and worst probability for that single game.
  • The Buccaneers had the highest RF expected value (1.92) with the Chiefs second (1.62).  The Steelers had the lowest (0.37) along with Washington (0.37).
  • The Chiefs had the most predicted LR wins (3) with the Buccaneers at 2.  For RF, the Chiefs, Bills, Packers, Saints, and Ravens tied for 2.  The Buccaneers were 1.
  • The Buccaneers had the most actual wins with 4.  That's a lot of wins in the playoffs.
  • The Seahawks had the highest average LR probability with 0.63 (only 1 game which they lost).  The Bears had the lowest (0.32 - 1 game).
  • The Saints had the highest average RF probability with 61%.  The Steelers and Washington had the lowest at 0.37.
  • The Saints had the highest max LR probability at 68% (against the Bears).  The Bears had the lowest (32%).
  • The Browns and Buccaneers had the highest max RF probability at 63%, against the two teams with the lowest: the Steelers and Washington (37%).  
  • The Seahawks had the highest minimum LR probability (63%), the Bears had the lowest (32%).
  • The Saints had the highest minimum RF probability (61%), while the Steelers and Washington had the lowest (37%).
  • The Buccaneers had the highest actual - LR predicted wins with 2 (2 predicted wins, 4 actual).  Several teams tied for -1: the Steelers, Titans, Chiefs, Seahawks, and Packers.
  • The Buccaneers had the highest actual - RF predicted wins with 3 (1 predicted, 4 actual).  Several teams tie for -1: the Packers, Saints, Ravens, and Seahawks.
  • The Buccaneers had the highest Actual - LR probabilities with 2.01 (1.99 expected, 4 actual).  The next highest was the Chiefs with 0.40.  The lowest was the Seahawks with -0.63.
  • The Buccaneers had the highest Actual - RF probabilities with 2.08 (1.92 expected, 4 actual).  Seahawks again had the lowest with -0.58.
In short, the Chiefs were the favorite in the playoffs.  The Buccaneers were the underdog, the overperformers, and the biggest surprise to my models.  The Seahawks were the biggest upset and playoff disappointment.  The Bears were the worst team in the playoffs.  


The Best Team:

Which was the "best" team this season?  If each team were to play every other team once, which team would have the most wins?  By summing up the probabilities of each game matchup generated by the LR model, the 32 game season would have a final top 5 rank of:
  1. Chiefs
  2. Packers
  3. Ravens
  4. Saints
  5. Bills
Chiefs remain on top at 1.  Packers remain at 2.  Ravens jump up to 3.  Saints move to 4 from 5.  Bills slide from 3 to 5.  Buccaneers (oddly) move from 4 to 8, contrary to their crushing Super Bowl victory against the Chiefs.

By summing the outcome of each game matchup (1 point per win), the final top 5 rank would be:
  1. Chiefs
  2. Packers
  3. Buccaneers
  4. Ravens
  5. Bills
Chiefs remain at 1.  Packers move up to 2 from 3.  Buccaneers move from 2 to 3.  Ravens take 4 and Bills slide to 5.  

What does the RF model say?  The top 5 by sum of probabilities of each game matchup are:
  1. Packers
  2. Chiefs
  3. Buccaneers
  4. Bills
  5. Saints
The RF model is largely in agreement with the LR model.  Packers are number 1 instead of the Chiefs, but barely.  Chiefs are 2 instead of 1.  Buccaneers are 3 instead of 8.  Bills take 4 (instead of 5) and Saints take 5 (instead of 4).  The Ravens are placed at 9.  

The top 5 by outcome of each game matchup are:
  1. Buccaneers
  2. Packers
  3. Chiefs
  4. Saints
  5. Bills
The Buccaneers take the number 1 spot (finally) instead of 3.  Packers take 2 as they did before.  Chiefs are 3 instead of 1.  Ravens move to 8 instead of 4.  Saints are 4 instead of 6.  Bills are 5 in both.

So how do we arrive at an answer to our question?  One way is to average the ranks into a single rank to balance out the strengths/weaknesses of each model and way of ranking.  When we do that we find the following:
  • 1 and 2. Chiefs/Packers: on average, they tie in rank.  They both do well in rank by expected value and in rank by total wins.  Yes, both teams were beat by the Buccaneers, but the models suggest that both teams would have more expected wins and total wins than the Buccaneers (and in the LR model, both teams are still predicted to win against the Buccaneers in a single game matchup).  If pushed, I would choose the Chiefs as the best team, following the LR model which performed better than the RF model throughout the season and in line with other statistics already mentioned.
  • 3. Buccaneers : The Buccaneers are hard to judge.  They won the Super Bowl and beat both the Chiefs and the Packers.  However, they were a 5th seed coming in to the playoffs and finished the season at 11-5.  They rank poorly by expected wins for the LR model and are 3rd for the RF model.  However, they perform well for total wins: 3rd for LR and 1st for RF.  So the models judge them to win a lot of games by a narrow margin.  Balancing all of these factors out places them in 3rd place behind the Chiefs and Packers.  So yes, that would mean that the better teams (Chiefs and Packers) lost to a worse team (Buccaneers) in the playoffs,  and the "best" team did NOT win the Super Bowl, but that's always a possibility in the playoffs and in football.
  • 4 and 5.  Saints and Bills: both tie in average rank.  Both teams are consistently 4 and 5 in expected value and total wins.
  • 6-10. In order of average rank: Ravens, Seahawks, Steelers, Colts, and Titans.
In short, the best teams were the Chiefs/Packers, then Buccaneers, then Saints/Bills.  The numbers are below, sorted by combined average rank.



The Worst Team:

Based on the above rankings, the "winners" for worst team were the Jets and Jaguars.  The LR model chose the Jaguars as the worst, followed by the Jets, while the RF model picked the Jets as worst, followed by the Jaguars.  When averaged, they tied for last.  Based on other stats I would probably choose the Jaguars as the worst team if forced to choose between the two.

The Best Team to Miss the Playoffs:

Which was the best team to miss the playoffs? I'd suggest that the Dolphins were the best team to miss. The Dolphins ranked ahead of several teams that did make the playoffs (Rams, Washington, Bears).  They finished at 10-6, second in the AFC East behind the Bills.

The Worst Team to Make the Playoffs:

Which was the worst team to make the playoffs?  Here are the two possibilities:
  • Bears - the Bears had an average rank of 19 and finished the regular season at 8-8.  However, they got the 7 seed wild card for finishing second in the NFC North.
  • Washington - Washington had an average rank of 15.25 and finished the regular season at 7-9, winning the NFC East division to get the 4th seed.
Between these two and based on other statistics, I'd choose the Bears as the worst team to make the playoffs.

The Biggest Playoff Upset

The Seahawks were the biggest playoff disappointment.  They had the highest probability for winning (63%) of the teams that didn't win a game in the playoffs and consequently, the biggest negative gap between expected wins and actual wins.  As a Seahawks fan, this is especially disappointing.

Summary and Conclusion

I could endlessly analyze and explore, but this post is already too long and life demands that I move on to other things.  But to summarize my findings:
  • My models were 63% accurate while 538 was 68% accurate.
  • The Chiefs were the best overall, the Jaguars were the worst overall.
  • The Buccaneers, although they won the Super Bowl, were NOT the best team this season.  But credit is due to them for beating the models over and over again. 
  • Don't bet against Tom Brady
So that concludes my predictions and analysis for the 2020 NFL season.  I hope you enjoyed these posts.  See you in a few months for the 2021 season with new predictions, and hopefully some model improvements and additional insights.

Thanks!

Thursday, January 28, 2021

2020 NFL Game Predictions: Super Bowl, Playoffs, Best Team

Welcome!  I intend this to be an ongoing project of predicting NFL game outcomes, point spreads, and final scores.  My hope is that my models will improve over time as they become more sophisticated and use better data.  I will try to regularly publish predictions to keep myself accountable, and will make observations along the way about what is working and what isn't.  See below for latest updates.  Enjoy!




Previous 2020 Predictions:
________________________________
Recap:
0-1 (LR) with 0.52 expected; 0-1 (RF) with 0.52 expected.  Both models were wrong, I was wrong, and 538 splits.  Don't bet against Tom Brady.  Buccaneers soundly defeat the Chiefs, playing in their home stadium, 31-9.  Don't bet against Tom Brady.  

The season is over so its time for the season recap!
________________________________
Super Bowl:
Here we are at the end: the Super Bowl, with the Chiefs playing the Buccaneers.

The Chiefs have a repeat appearance from last year's Super Bowl victory.  They had the best record in the regular season and have been consistently judged by my models to be the best team in the league. They have always been the #1 seed in my playoff predictions and have always been picked to win the Super Bowl.  They certainly deserve to be here.

The Buccaneers under Tom Brady struggled in the middle of the season.  Before their bye in week 13 they were 7-5.  However, they won 4 straight to finish 11-5 with a Wild Card 5th seed and have been winning ever since, knocking off the Saints and Packers on their way.  They have also climbed in my model's estimation.  They started Week 14 as the predicted #6 seed getting knocked out in the Wild Card round, with a ranking of 15th in the league.  By Week 17 they were the predicted 5th seed progressing to the Division round before losing as 13th in the league.  They moved to 8th in Wild Card week, 9th in Division week, and 6th in Conference week as they continued to win again and again.  What rank are they going into the Super Bowl?  See below.

So who will win?

Oddities and Bold Predictions:
  • Chiefs vs. Buccaneers
    • Both models pick the Chiefs.  538 QB chooses the Chiefs while the 538 traditional model chooses the Buccaneers.  The Chiefs won the last matchup in late November, 27 to 24.  I know I'm not supposed to bet against Tom Brady, but I am picking the Chiefs.
Good luck!

Here are the predictions for the Super Bowl:



Playoffs:

How will the playoffs playout using the final seeding and game results?  Using the LR model for one-time head-to-head matchup predictions:
  • Wildcard:
    • AFC
      • Chiefs (1)-bye
      • Bills (2) vs. Colts (7) -> Bills (2)
      • Steelers (3) vs. Browns (6) -> Browns (6)
      • Titans (4) vs. Ravens (5) -> Ravens (5)
    • NFC
      • Packers (1)- bye
      • Saints (2) vs. Bears (7) -> Saints (2)
      • Seahawks (3) vs. Rams (6) -> Rams (6)
      • Washington (4) vs. Buccaneers (5) -> Buccaneers (5)
  • Division:
    • AFC
      • Chiefs (1) vs. Browns (6) -> Chiefs (1)
      • Bills (2) vs. Ravens (5) ->  Bills (2)
    • NFC
      • Packers (1) vs. Rams (6) -> Packers (1)
      • Saints (2) vs. Buccaneers (5) -> Buccaneers (5)
  • Conference:
    • AFC: Chiefs (1) vs. Bills (2) -> Chiefs (1)
    • NFC: Packers (1) vs. Buccaneers (5) -> Buccaneers (5)
  • Super Bowl:
    • Chiefs (1) vs. Buccaneers (5) -> Chiefs (1) - (0.523)

The LR model still predicts that the Chiefs will go all the way and win the Super Bowl, but obviously its now against the Buccaneers. No need to run a Monte Carlo simulation this week as it would only tell us that the Chiefs have a 52.3% chance of winning the Super Bowl compared with the Buccaneer's 47.7% chance, with everyone else at a 0% chance.

Best Team:
What is the "best" team this week?  If each team were to play every other team once, which team would have the most wins?  By summing up the probabilities of each game matchup, the 32 game season would have a final top 5 rank of:
  1. Chiefs
  2. Packers
  3. Bills
  4. Buccaneers
  5. Saints
Packers move back down to 2 from 1 as Chiefs reclaim the top position.  Bills remain at 3.  Buccaneers jump to 4 from 6.  Saints move to 5 from 7.  Seahawks move to 6 from 4.  Colts move to 8 from 5.  The top 10 has become significantly more competitive as the Chiefs drop below 23 games to 22.955 expected. The Dolphins at 10 are at 19.692 expected wins.  Ranks 3 - 5 round to 21 games and 6 - 10 round to 20 games.  So the field is very close together in predicted wins.

By summing the outcome of each game matchup (1 point per win), the final top 5 rank would be:
  1. Chiefs
  2. Buccaneers
  3. Packers
  4. Bills
  5. Browns
Chiefs stay on top.  Buccaneers move from 4 to 2.  Packers slide to 3 from 2.  Bills move to 4 from 3.  Browns remain at 5.  Interestingly and for the first time, we have some ties, as the Titans tie the Seahawks for 6th and the Dolphins tie the Colts for 8th.  This is possible because:
  • Seahawks beat the Titans but lose to the Colts
  • Titans beat the Colts but lose to the Seahawks
Similarly:
  • Dolphins beat the Colts but lose to the Seahawks
  • Colts lose to the Dolphins but beat the Seahawks
I guess the transitive property doesn't hold (i.e., if A beats B and B beats C, then A beats C).

At this point, it looks like we have 4 top teams: the Chiefs, Packers, Bills, and Buccaneers.  Two of those teams are in the Super Bowl and the other two lost in the Conference Championships, so the playoffs are at least matching this ranking pretty well.

The Chiefs are at the top of both lists and get the number 1 spot.  Packers are ahead of the Bills in both lists as well.  So the Buccaneers are the variable as they win lots of games narrowly while the Packers and Bills win fewer games but with a wider margin on average.  So who is better?  Of just those top 4 teams, the Chiefs win against the others (3), the Buccaneers win 2 (Packers-0.532 and Bills-0.53), the Packers win 1 (against the Bills-0.51), and the Bills win 0.  The Packers have a 0.625 expected value margin on the Buccaneers while the Bills have a 0.277 expected value margin.

One can average the rankings and thereby split the difference between the two rankings, meaning that the Chiefs take first, the Packers take second, but the Buccaneers take third with the Bills moving into fourth.  This seems to be a fair compromise.  But in any case, it seems that the Chiefs are on top, with the Packers/Buccaneers likely fighting for second and third, and with the Bills in fourth.

We'll see if that remains true after this final game...


Tuesday, January 19, 2021

2020 NFL Game Predictions: Conference Championships, Playoffs, Best Team

Welcome!  I intend this to be an ongoing project of predicting NFL game outcomes, point spreads, and final scores.  My hope is that my models will improve over time as they become more sophisticated and use better data.  I will try to regularly publish predictions to keep myself accountable, and will make observations along the way about what is working and what isn't.  See below for latest updates.  Enjoy!




Previous 2020 Predictions:
_______________________________________________________
Recap:
1-1 (LR) with 1.05 expected; 0-2 (RF) with 1.078 expected.  Buccaneers upset the Packers but Chiefs beat the Bills as expected.  Tom Brady gets to go to the Super Bowl again.  I guess the lesson now is not so much "don't bet against the Patriots" but "don't bet against Tom Brady".  Personally disappointed that the Packers lost but glad that the Chiefs made it through.
  • Packers vs. Buccaneers
    • Buccaneers scored first, Packers tied, but Buccaneers took the lead again and never let go.  Everyone misses.
  • Chiefs vs. Bills
    • Bills were up 9-0, but Chiefs got 3 consecutive touchdowns and never lost the lead.  LR, myself, 538 correct; RF incorrect.
On to the Super Bowl!
_______________________________________________________
Conference Championships:
Four teams left as we move to the conference championships: Chiefs vs. Bills, Packers vs. Buccaneers.  Two games this week.  The models disagree on 1 of the 2 games.  538 is in agreement between QB and traditional models.  It sides with LR on both games.

Oddities and Bold Predictions:
  • Chiefs vs. Bills
    • LR picks the Chiefs and RF barely picks the Bills.  538 picks the Chiefs.  A major factor in this is whether or not the Chiefs' top QB will be starting after sustaining a concussion.  The models currently assume he will be playing.  I pick the Chiefs if their top QB does play.  If not, I'm not sure....
  • Packers vs. Buccaneers
    • All models pick the Packers.  I agree: Packers will win.
Good luck!

Here are the predictions for the Conference Championships:



Playoffs:

How will the playoffs playout using the final seeding and game results?  Using the LR model for one-time head-to-head matchup predictions:
  • Wildcard:
    • AFC
      • Chiefs (1)-bye
      • Bills (2) vs. Colts (7) -> Bills (2)
      • Steelers (3) vs. Browns (6) -> Browns (6)
      • Titans (4) vs. Ravens (5) -> Ravens (5)
    • NFC
      • Packers (1)- bye
      • Saints (2) vs. Bears (7) -> Saints (2)
      • Seahawks (3) vs. Rams (6) -> Rams (6)
      • Washington (4) vs. Buccaneers (5) -> Buccaneers (5)
  • Division:
    • AFC
      • Chiefs (1) vs. Browns (6) -> Chiefs (1)
      • Bills (2) vs. Ravens (5) ->  Bills (2)
    • NFC
      • Packers (1) vs. Rams (6) -> Packers (1)
      • Saints (2) vs. Buccaneers (5) -> Buccaneers (5)
  • Conference:
    • AFC: Chiefs (1) vs. Bills (2) -> Chiefs (1) - (0.522)
    • NFC: Packers (1) vs. Buccaneers (5) -> Packers (1) - (0.528)
  • Super Bowl:
    • Chiefs (1) vs. Packers (1) -> Chiefs (1) - (0.510)

The LR model still predicts that the Chiefs will go all the way and win the Super Bowl against the Packers.  No changes in game outcomes in Conference Championships or Super Bowl.

If we run this as a Monte Carlo simulation with the seeding fixed, played games fixed, and 50,000 runs, we get the following result:



Chiefs are still on top but are practically tied with the Packers like last week.  Bills stay at third and Buccaneers are fourth still, but are a much closer fourth than last week after beating the Saints.

Best Team:
What is the "best" team this week?  If each team were to play every other team once, which team would have the most wins?  By summing up the probabilities of each game matchup, the 32 game season would have a final top 5 rank of:
  1. Packers
  2. Chiefs
  3. Bills
  4. Seahawks
  5. Colts
Packers take over the top position (from 2) as Chiefs slide to 2 from 1.  Bills stay at 3. Seahawks surprisingly jump to 4 (from 7), and Colts move to 5 (from 10).  Buccaneers move from 9 to 6, Saints move from 5 to 7, and Titans move from 4 to 8.  I should note that the difference between the Seahawks at 4 and the Steelers at 10 is less than a single game.  While slight changes in the probabilities are dramatically changing the ordering, there are very few if any major differences in the quality of these teams as 5 of the 7 would round to the same number of games won (i.e., 21).

By summing the outcome of each game matchup (1 point per win), the final top 5 rank would be:
  1. Chiefs
  2. Packers
  3. Bills
  4. Buccaneers
  5. Browns
Chiefs remain on top.  Packers move to 2 (from 3).  Bills move back to 3 (from 2).  Buccaneers stay at 4.  Browns move up to 5 (from 12).  Titans move to 8 (from 5).  

For the top three teams, the expected value and total win ranks are largely in agreement. However, the Packers are moving up from fighting with the Bills for second, to fighting with the Chiefs for first.  Although the Packers have a slight expected value advantage over the Chiefs, the model still predicts that the Chiefs would beat the Packers in a matchup.  So we have Chiefs and Packers for first/second, and Bills for third.




Tuesday, January 12, 2021

2020 NFL Game Predictions: Division, Playoffs, Best Team

Welcome!  I intend this to be an ongoing project of predicting NFL game outcomes, point spreads, and final scores.  My hope is that my models will improve over time as they become more sophisticated and use better data.  I will try to regularly publish predictions to keep myself accountable, and will make observations along the way about what is working and what isn't.  See below for latest updates.  Enjoy!




Previous 2020 Predictions:

_______________________________________________________
Recap:
4-0 (LR) with 2.345 expected; 2-2 (RF) with 2.319 expected.  I personally went 3-1, as did 538.  LR got lucky and RF got unlucky.
  • Bills beat the Ravens as predicted.  All models except RF correct.
  • Packers solidly beat the Rams.  All models correct.
  • Chiefs beat the Browns, holding on to the lead after losing their starting QB to a concussion.  All models correct.  Whether the QB will play in the next game will be a huge factor in the Chiefs' likelihood of winning that game.
  • Buccaneers upset the Saints as LR is correct; everyone else is incorrect.
Next week is the Conference Championships!
_______________________________________________________
Division:
The Wild Card week is done and we are on to the Divisional round.  The winners of the Wild Card games play each other and the #1 seeds that had a first round bye.  There are 4 games this week.  

The models disagree on 2 of the 4 games.  538 is in agreement between QB and traditional models.  It sides with LR on 1 game and RF on another.  All models are in agreement on 2 games.  So not a straightforward week for predictions.  I'd expect to only get 2 games right.

Oddities and Bold Predictions:
  • Bills vs. Ravens
    • LR picks Bills but RF picks Ravens.  538 (QB and traditional) chooses the Bills.  Bills are 14-3 and Ravens are 12-5.  Bills play at home.  I think this is a pretty good matchup with pretty evenly matched teams, so it could go either way.  I am going to side with the Bills.
  • Packers vs. Rams
    • All models agree: Packers will win.  I am inclined to agree.
  • Saints vs. Buccaneers
    • LR picks Buccaneers but RF picks Saints.  538 (QB and traditional) picks Saints.  Saints play at home and are 13-4.  Buccaneers are 12-5.  Both teams are passing teams and the stats are very similar so it's a close call.  Buccaneers lost both previous matchups this season.  Picking the Saints.
  • Chiefs vs. Browns
    • All models agree: Chiefs will win.  Picking the Chiefs.
Good luck!

Here are the predictions for the Divisional Round:



Playoffs:

How will the playoffs playout using the final seeding and game results?  Using the LR model for one-time head-to-head matchup predictions:
  • Wildcard:
    • AFC
      • Chiefs (1)-bye
      • Bills (2) vs. Colts (7) -> Bills (2)
      • Steelers (3) vs. Browns (6) -> Browns (6)
      • Titans (4) vs. Ravens (5) -> Ravens (5)
    • NFC
      • Packers (1)- bye
      • Saints (2) vs. Bears (7) -> Saints (2)
      • Seahawks (3) vs. Rams (6) -> Rams (6)
      • Washington (4) vs. Buccaneers (5) -> Buccaneers (5)
  • Division:
    • AFC
      • Chiefs (1) vs. Browns (6) -> Chiefs (1) - (0.56)
      • Bills (2) vs. Ravens (5) ->  Bills (2) - (0.618)
    • NFC
      • Packers (1) vs. Rams (6) -> Packers (1) - (0.661)
      • Saints (2) vs. Buccaneers (5) -> Buccaneers (5) - (0.506)
  • Conference:
    • AFC: Chiefs (1) vs. Bills (2) -> Chiefs (1) - (0.512)
    • NFC: Packers (1) vs. Buccaneers (5) -> Packers (1) - (0.526)
  • Super Bowl:
    • Chiefs (1) vs. Packers (1) -> Chiefs (1) - (0.525)

The LR model still predicts that the Chiefs will go all the way and win the Super Bowl against the Packers.

If we run this as a Monte Carlo simulation with the seeding fixed, played games fixed, and 20,000 runs, we get the following result:



The Chiefs still win most often at 17.8% (down from 20%), followed by the Packers at 17.6% (up from 16%), so they are virtually tied.  The Bills, having won in the first round, are now at 16.8% (up from 9%).  All of the teams that played and won in the first round have a boost in likelihood as they now distribute among themselves the collective probabilities once assigned to their Wild Card opponents.  Also, the advantage once given to the Chiefs and Packers due to their having first round byes no longer applies, meaning that the probabilities given now better reflect only the strength of each team as measured by the model (as well as matchup advantages/disadvantages due to seeding).

Best Team:
What is the "best" team this week?  If each team were to play every other team once, which team would have the most wins?  By summing up the probabilities of each game matchup, the 32 game season would have a final top 5 rank of:
  1. Chiefs
  2. Packers
  3. Bills
  4. Titans
  5. Saints
Chiefs remain on top. Packers move from 5 to 2.  Bills move from 2 to 3.  Steelers get bumped from 3 to 8.  Seahawks get bumped from 4 to 7.  Titans move up to 4 from 13.  Saints move up from 7 to 5.  Ravens move from 9 to 6.  Buccaneers move from 8 to 9, Browns from 6 to 12, and Rams from 14 to 13.  

By summing the outcome of each game matchup (1 point per win), the final top 5 rank would be:
  1. Chiefs
  2. Bills
  3. Packers
  4. Buccaneers
  5. Titans
Chiefs stay on top.  Bills remain in second. Packers move from 4 to 3.  Buccaneers move from 5 to 4.  Steelers move from 3 to 7.  Titans move from 11 to 5.  Saints move from 9 to 6.  Ravens move from 12 to 9.  Browns move from 7 to 12.  Rams move from 14 to 13.  

For the top three teams, the expected value and total win ranks are largely in agreement.  In short, the Chiefs still remain the "best" team with the Bills and Packers fighting it out for second and third.



Tuesday, January 5, 2021

2020 NFL Game Predictions: Wild Card, Playoffs, Best Team

Welcome!  I intend this to be an ongoing project of predicting NFL game outcomes, point spreads, and final scores.  My hope is that my models will improve over time as they become more sophisticated and use better data.  I will try to regularly publish predictions to keep myself accountable, and will make observations along the way about what is working and what isn't.  See below for latest updates.  Enjoy!




Previous 2020 Predictions:
________________________________________________________
Recap:
3-3 (LR) with 3.406 expected; 5-1 (RF) with 3.548 expected.  With those probabilities, I had a 52% chance of getting 3 or fewer games correct with the LR model, so not surprising.  I think I got lucky with the RF model.  Personally, I go 4-2.  538 QB is 4-2 and 538 traditional is 3-3.
  • Bills beat the Colts.  Colts scored first but Bills took lead late in 1st quarter and never gave it up.  Everybody correct. 
  • Rams beat the Seahawks.  Rams scored first, Seahawks tied, but Rams took lead again and never gave it up.  Several interceptions and dropped passes by the Seahawks as the Rams decisively win over the Seahawks.  Everybody incorrect.
  • Buccaneers beat Washington.  Buccaneers took and never lost the lead.  Everybody correct.  
  • Ravens beat Titans.  Titans took early lead but Ravens tied it by the half.  Ravens took lead in 3rd quarter and never lost it after that.  RF, 538 QB, and myself correct; LR and 538 traditional incorrect.
  • Saints beat the Bears as expected.  Saints took and never lost the lead.  Everybody correct.
  • Browns beat the Steelers.  Browns took 28-0 lead in first quarter with a missed Steelers snap turned into a Browns touchdown and several interceptions.  Steelers closed in to 11 points by the end of the game but never caught up.  Lots of scoring for a final 48-37 Browns victory in which Steelers QB throws for 501 yards but 4 interceptions.  RF correct; LR, myself, 538 QB, and 538 traditional incorrect.
Playoff predictions are hard.  Better and more evenly matched teams mean that the probabilities are closer together, making any prediction more like a 50-50 guess.  I suppose that is good as it makes for an exciting playoff experience.

On to the divisional round!
________________________________________________________
Wild Card:
At last we are in the playoffs with the Wild Card round.  Seeds 2-7 face off while the number 1 seeds have a bye.  There are 6 games this week.  The models disagree on 2 of the 6 games.  538 is split on one of those as well.  Here we go...

Oddities and Bold Predictions:
  • Steelers vs. Browns
    • LR favors Steelers while RF picks Browns.  538 (QB and traditional) picks Steelers.  The Browns and Steelers just played, and though the Browns won, they won barely, and the Steelers were sitting a lot of their starters.  Steelers play at home.  Steelers are 12-4 and Browns are 11-5.  Picking the Steelers.
  • Titans vs. Ravens
    • LR narrowly picks Titans while RF narrowly picks Ravens.  538 QB picks Ravens while 538 traditional picks Titans.  So this is a toss up.  Titans play at home and are 11-5.  Ravens are also 11-5.  Titans won the previous matchup this season.  Ravens look like a slightly better team on paper.  Picking the Ravens.
The other predictions look correct, but I have a funny feeling about the Colts upsetting the Bills... Good luck!

Here are the predictions for Wild Card Week:



Playoffs:
The playoff seeds are now fixed as follows:
  • AFC
    1. Chiefs: West Champ
    2. Bills: East Champ
    3. Steelers: North Champ
    4. Titans: South Champ
    5. Ravens: Wild Card #1
    6. Browns: Wild Card #2
    7. Colts: Wild Card #3
  • NFC
    1. Packers: North Champ
    2. Saints: South Champ
    3. Seahawks: West Champ
    4. Washington: East Champ
    5. Buccaneers: Wild Card #1
    6. Rams: Wild Card #2
    7. Bears: Wild Card #3
How good were my playoff predictions?  I started predicting playoffs in week 14.  So there are 4 weeks of playoff seed predictions.  These are the summary results:
  • Week 14:
    • Made Playoffs: AFC - 6/7; NFC - 5/7
    • Correct Seed: AFC - 2/7; NFC - 0/7 
  • Week 15:
    • Made Playoffs: AFC - 6/7; NFC - 6/7
    • Correct Seed:  AFC - 2/7; NFC - 3/7
  • Week 16:
    • Made Playoffs: AFC - 6/7; NFC - 6/7
    • Correct Seed: AFC - 2/7; NFC - 4/7
  • Week 17:
    • Made Playoffs: AFC - 6/7; NFC - 7/7
    • Correct Seed: AFC - 5/7; NFC - 7/7
The consistent misses in making the playoffs were the Dolphins in the AFC and the Cardinals in the NFC, but overall this was not too challenging.  The correct seeding is much more difficult.  Seeds #1-4 remained largely the same, but the ordering of #5-7 changed nearly every week.

How will the playoffs playout using the final seeding and game results?  Using the LR model for one-time head-to-head matchup predictions:
  • Wildcard:
    • AFC
      • Chiefs (1)-bye
      • Bills (2) vs. Colts (7) -> Bills (2) - (0.528)
      • Steelers (3) vs. Browns (6) -> Steelers (3) - (0.521)
      • Titans (4) vs. Ravens (5) -> Titans (4) - (0.509)
    • NFC
      • Packers (1)- bye
      • Saints (2) vs. Bears (7) -> Saints (2) - (0.681)
      • Seahawks (3) vs. Rams (6) -> Seahawks (3) - (0.631)
      • Washington (4) vs. Buccaneers (5) -> Buccaneers (5) - (0.536)
  • Division:
    • AFC
      • Chiefs (1) vs. Titans (4) -> Chiefs (1) - (0.613)
      • Bills (2) vs. Steelers (3) ->  Bills (2) - (0.513)
    • NFC
      • Packers (1) vs. Buccaneers (5) -> Packers (1) - (0.507)
      • Saints (2) vs. Seahawks (3) -> Seahawks (3) - (0.53)
  • Conference:
    • AFC: Chiefs (1) vs. Bills (2) -> Chiefs (1) - (0.539)
    • NFC: Packers (1) vs. Seahawks (3) -> Packers (1) - (0.51)
  • Super Bowl:
    • Chiefs (1) vs. Packers (1) -> Chiefs (1) - (0.559)

The LR model still predicts that the Chiefs will go all the way and win the Super Bowl against the Packers.  In fact, the outcomes are the exact same as the predictions from last week.  There are slight probabilistic changes, but the predicted winner of each game in the playoffs is the same.

If we run this as a Monte Carlo simulation with the seeding fixed and 20,000 runs, we get the following result:



The Chiefs still win most often at 20% of the time (down from 24%), followed by the Packers at 16% (down from 20%) and the Bills at 9% (stay the same).  The downward change in the Chiefs is most likely due to sitting their starters and losing their last game.  We know that this doesn't really reflect how good the team is, but the model doesn't know that.  The previous 24% is probably a better reflection of the Chief's chances of winning the Super Bowl.

Best Team:
What is the "best" team this week?  If each team were to play every other team once, which team would have the most wins?  By summing up the probabilities of each game matchup, the 32 game season would have a final top 5 rank of:
  1. Chiefs
  2. Bills
  3. Steelers
  4. Seahawks
  5. Packers
Chiefs remain on top. Bills remain in second.  Steelers move from 8 to 3.  Packers move from 3 to 5.  Seahawks move from 5 to 4.   Saints move from 4 to 7.  In all this, it is important to remember that only the Chiefs are ahead of the next team by a full game.  In fact, the difference between the Bills and the Saints is only 1.37 games.  A very slight change in the probabilities can significantly change the ordering of the top teams, as it continues to do so.

By summing the outcome of each game matchup (1 point per win), the final top 5 rank would be:
  1. Chiefs
  2. Bills
  3. Steelers
  4. Packers
  5. Buccaneers
Chiefs stay on top.  Bills remain in second. Steelers move from 4 to 3.  Packers move from 3 to 4.  Buccaneers jump from 11 to 5.  Seahawks remain at 6.  Dolphins move from 4 to 10.

For the top three teams, the expected value and total win ranks are in agreement.  In short, the Chiefs still remain the "best" team, the Bills remain in second, and the Steelers move up to third.