NFL Expectation Analysis - Last Five Seasons

[deljzc]

As many of you know, I have been using "money line" data on football games to establish a correct win expectation level entering week-to-week games in the NFL. So for example, the current money line on the Steelers first game of this upcoming season (Cleveland at home) is -$240 +$200. This correlates mathematically at the Steelers winning the game 68.6% of the time and Cleveland winning 32.4% of the time.

Over the course of a season(s) you can mathematically keep track of both a) the trend in expectation level and b) whether the team is winning to those expectations. For example, a team that played every game as a "pick 'em" has a 50% win probability for all 16 games and would be EXPECTED to win 8 games. If that team wins 7 games, they are UNDERACHIEVING. If that team wins 9 games, they are OVERACHIEVING.

The data below is REGULAR SEASON ONLY. I have data on playoffs, but those sample sizes are very small and might not be representative of skill.


AFC EAST

New England
Wins Expected (1 year, 3 years, 5 years) = 10.13, 34.48, 55.80
Actual Wins = 12, 37, 61

New York Jets
Wins Expected = 6.00, 22.38, 40.60
Actual Wins = 8, 22, 42

Miami Dolphins:
Wins Expected = 7.65, 21.26, 35.67
Actual Wins = 8, 21, 35

Buffalo Bills:
Wins Expected = 6.17, 19.76, 31.10
Actual Wins = 6, 18, 28


AFC NORTH

Baltimore Ravens
Wins Expected: 7.92, 28.17, 49.15
Actual Wins: 8, 30, 51

Pittsburgh Steelers
Wins Expected: 8.14, 28.49, 49.14
Actual Wins: 8, 28, 49

Cincinnati Bengals
Wins Expected: 9.57, 25.77, 40.75
Actual Wins: 11, 30, 44

Cleveland Browns
Wins Expected: 6.42, 18.35, 28.55
Actual Wins: 4, 13, 23


AFC SOUTH

Indianapolis Colts
Wins Expected: 8.59, 19.47, 39.93
Actual Wins: 11, 24, 48

Houston Texans
Wins Expected: 7.90, 28.91, 45.86
Actual Wins: 2, 24, 39

Tennessee Titans
Wins Expected: 7.41, 21.64, 37.99
Actual Wins: 7, 22, 36

Jacksonville Jaguars
Wins Expected: 3.42, 13.30, 27.40
Actual Wins: 4, 11, 26


AFC WEST

San Diego Chargers
Wins Expected: 8.04, 25.49, 47.02
Actual Wins: 9, 24, 46

Denver Broncos
Wins Expected: 12.60, 29.75, 43.80
Actual Wins: 13, 34, 46

Kansas City Chiefs
Wins Expected: 9.11, 19.36, 32.22
Actual Wins: 11, 20, 34

Oakland Raiders
Wins Expected: 4.80, 17.91, 28.07
Actual Wins: 4, 16, 29


NFC EAST:

Philadelphia Eagles
Wins Expected: 9.22, 26.20, 46.00
Actual Wins: 10, 22, 43

Dallas Cowboys
Wins Expected: 8.26, 26.48, 45.38
Actual Wins: 8, 24, 41

New York Giants
Wins Expected: 7.11, 24.68, 44.24
Actual Wins: 7, 25, 43

Washington Redskins
Wins Expected: 6.57, 20.65, 33.19
Actual Wins: 3, 18, 28


NFC SOUTH

New Orleans Saints
Wins Expected: 10.30, 30.07, 52.77
Actual Wins: 11, 31, 55

Atlanta Falcons
Wins Expected: 7.08, 25.99, 44.01
Actual Wins: 4, 27, 49

Carolina Panthers
Wins Expected: 9.60, 24.11, 36.13
Actual Wins: 12, 25, 35

Tampa Bay Buccaneers
Wins Expected: 5.77, 19.36, 30.29
Actual Wins: 4, 15, 28


NFC NORTH

Green Bay Packers
Wins Expected: 9.24, 32.09, 52.61
Actual Wins: 8.5, 34.5, 55.5

Chicago Bears
Wins Expected: 8.15, 25.24, 40.61
Actual Wins: 8, 26, 44

Minnesota Vikings
Wins Expected: 5.79, 19.04, 37.72
Actual Wins: 5.5, 18.5, 36.5

Detroit Lions
Wins Expected: 9.29, 26.74, 35.55
Actual Wins: 7, 21, 29


NFC WEST

San Francisco 49ers
Wins Expected: 10.69, 31.15, 47.28
Actual Wins: 12, 36.5, 50.5

Seattle Seahawks
Wins Expected: 11.80, 26.38, 39.65
Actual Wins: 13, 31, 43

Arizona Cardinals
Wins Expected: 7.40, 19.25, 35.07
Actual Wins: 10, 23, 38

St. Louis Rams
Wins Expected: 5.89, 16.01, 25.99
Actual Wins: 7, 16.5, 24.5



Remember, a team under management can fail in two ways: they can LOWER the expectation of the team with lack of talent and quality wins and they can UNDERACHIEVE against the expectations neutral observers think of them. Both are bad. Both go hand-in-hand. But it's worth realizing they are separate things and are somewhat controlled by different parts of the organization and/or coaching staff.

 

[Wingman]

So your 5 yr number for the steelers is 49 wins out of 80 games giving us a record of 49 and 31 over 5 years, compared to darth hoodie of 61 and 19 over the same time frame. Now what is it projected for the future this seasons projected would be good fodder for discussion if it is available.

 

[deljzc]

So your 5 yr number for the steelers is 49 wins out of 80 games giving us a record of 49 and 31 over 5 years, compared to darth hoodie of 61 and 19 over the same time frame. Now what is it projected for the future this seasons projected would be good fodder for discussion if it is available.

I'm confused at the question.

If the Patriots had performed "average" vs. their expectations they would have won 56 games over the least 5 seasons. The Patriots ACTUALLY won 61 regular season games. That's very, very good. Not only do they maintain high expectations (which is hard to do) but they consistently reach them in the regular season.

If the Steelers performed "average" vs. their expectation level they would have won 49 games and in fact, they have won exactly 49 games over the past five regular seasons. What is more concerning however is the expectation level of the team is declining (10+ wins 5 years ago to 8 wins in 2013). Falling expectations are a major issue with this team right now and it will take a solid season of playoff football to put us back into the discussion of a team "expected" to be playoff contenders year-to-year-to-year.

I can not give you projected expectation level for this season because there are too many variables. Each week injuries, weather conditions, hot/cold streaks, off-field issues, etc. effect the team's perceived chances at winning/losing that week's game. That's why I use week-to-week money line gambling records to calculate win expectation rates and not something like pre-season +/- win totals.

 

[Drink IRON City]

So your 5 yr number for the steelers is 49 wins out of 80 games giving us a record of 49 and 31 over 5 years, compared to darth hoodie of 61 and 19 over the same time frame. Now what is it projected for the future this seasons projected would be good fodder for discussion if it is available.

What a reference LOL I've heard many, but not that one lately. thank you wingman


Salute the nation

 

[slashsteel]

I'm confused at the question.

If the Patriots had performed "average" vs. their expectations they would have won 56 games over the least 5 seasons. The Patriots ACTUALLY won 61 regular season games. That's very, very good. Not only do they maintain high expectations (which is hard to do) but they consistently reach them in the regular season.

If the Steelers performed "average" vs. their expectation level they would have won 49 games and in fact, they have won exactly 49 games over the past five regular seasons. What is more concerning however is the expectation level of the team is declining (10+ wins 5 years ago to 8 wins in 2013). Falling expectations are a major issue with this team right now and it will take a solid season of playoff football to put us back into the discussion of a team "expected" to be playoff contenders year-to-year-to-year.

I can not give you projected expectation level for this season because there are too many variables. Each week injuries, weather conditions, hot/cold streaks, off-field issues, etc. effect the team's perceived chances at winning/losing that week's game. That's why I use week-to-week money line gambling records to calculate win expectation rates and not something like pre-season +/- win totals.

Patriots are not the norm, and I can't honestly say that they are coming by their record honestly. I still think there is some home game cheating going on.

But yeah the Pats have exceeded and the Steelers have been average. But a bounce back year or two coming off two 8 and 8s sure could change things up a bit.


For the Steelers I think the arrow is back to pointing up.

 

[Wingman]

I'm confused at the question.

If the Patriots had performed "average" vs. their expectations they would have won 56 games over the least 5 seasons. The Patriots ACTUALLY won 61 regular season games. That's very, very good. Not only do they maintain high expectations (which is hard to do) but they consistently reach them in the regular season.

If the Steelers performed "average" vs. their expectation level they would have won 49 games and in fact, they have won exactly 49 games over the past five regular seasons. What is more concerning however is the expectation level of the team is declining (10+ wins 5 years ago to 8 wins in 2013). Falling expectations are a major issue with this team right now and it will take a solid season of playoff football to put us back into the discussion of a team "expected" to be playoff contenders year-to-year-to-year.

I can not give you projected expectation level for this season because there are too many variables. Each week injuries, weather conditions, hot/cold streaks, off-field issues, etc. effect the team's perceived chances at winning/losing that week's game. That's why I use week-to-week money line gambling records to calculate win expectation rates and not something like pre-season +/- win totals.

Del I was not trying to be a smartass, although it does come naturally. I do not gamble or follow lines or any of that stuff but it appears that you do. So I was wondering if the odds makers had some projections on what they expected the season to be based on the numbers. I thought it might make for a good discussion topic, that's all, especially with the present lack of things to talk about, unless it will be archer.

 

[deljzc]

Del I was not trying to be a smartass, although it does come naturally. I do not gamble or follow lines or any of that stuff but it appears that you do. So I was wondering if the odds makers had some projections on what they expected the season to be based on the numbers. I thought it might make for a good discussion topic, that's all, especially with the present lack of things to talk about, unless it will be archer.

The only thing available is the over/under win total bets on all 32 teams, which has been released by Vegas Hilton here:

http://www.nationalfootballpost.com/LVH-releases-2014-NFL-season-win-totals.html

It is important to understand this is a snap-shot NOW, where my method is a snap shot week-to-week, game-to-game. The more information you have (week-to-week injuries or weather for instance), the more accurate Vegas gets.

You have to realize there is no "accurate" over-under number on that list either. They pick something to encourage bets and then make adjustments using the vigorish.

For example, you pick "Over 11" on Denver and you have to bet $140 to win $100. If you pick "Under 11", you bet $100 you win $120. Those are drastically different payouts. That means Vegas thinks Denver is going to win MORE than 11 game, but not quite enough to put the line at 11.5 or 12.

Based on the over/under wins for the Steelers (8.5) and the payouts (more money to bet the under), I would suspect Vegas right now pegs the Steelers at a 9 win team. I actually think that's a pretty good bet, since I'm picking the Steelers to win 10 or 11 this year.

When I create an "expected win percentage" on a specific game, there is no difference in vigorish. I average out the vigorish both ways so that the win probability of a game exactly adds up to 1 win total (for example, in week 1 the Steelers have a 68.63% win expectation and the Cleveland Browns have a 31.37% win expectation = 100%).

Also note the link above does NOT add up to 256 wins (it adds up to 258), which is impossible. Vegas finds more people like to bet "overs" than "unders" and thus adjusts things in their favor. That type of thinking does not effect their weekly "odds bets" (which is just a bet on who wins each game).