Week 17 stats, and first round NFL Playoff Odds

The regular season has ended and the playoffs have begun. It would be useful to have a set of playoff grade data to do playoff probabilities, and though I’ve been down and out this season (no job at times, foot stress fracture at times, and a bad right shoulder), I currently have some time off my new job, a new laptop, and enough time to grind through some playoff numbers.

NFL stats at the end of the regular season:

To explain the columns above, Median is a median point spread, and can be used to get a feel for how good a team is without overly weighting a blowout win or blowout loss. HS is Brian Burke’s Homemade Sagarin, as implemented in Maggie Xiong’s PDL::Stats. Pred is the predicted Pythagorean expectation. The exponent for this measure is fitted to the data set itself. SOS, SRS, and MOV are the simple ranking components, analyzed via this Perl implementation. MOV is margin of victory, or point spread divided by games played. SOS is strength of schedule. SRS is the simple ranking.

Playoff Odds are calculated according to this model:

logit P  =  0.668 + 0.348*(delta SOS) + 0.434*(delta Playoff Experience)

The results are given below, as a “score” in logits:

2013 NFL Playoff Teams, C&F Playoff Model Worksheet.
NFC
Rank	Name	Home Field Advantage	Prev. Playoff Experience	Strength of Schedule	Total Score
1	Seattle Seahawks	0.406	0.434	0.494	1.334
2	Carolina Panthers	0.406	0.0	0.484	0.889
3	Philadelphia Eagles	0.406	0.0	-0.661	-0.256
4	Green Bay Packers	0.406	0.434	-0.842	-0.003
5	San Francisco 49ers	0.0	0.434	0.612	1.046
6	New Orleans Saints	0.0	0.0	0.658	0.658
AFC
1	Denver Broncos	0.406	0.434	-0.546	0.293
2	NE Patriots	0.406	0.434	-0.258	0.582
3	Cancinnati Bengals	0.406	0.434	-0.856	-0.017
4	Indianapolis Colts	0.406	0.434	0.209	1.048
5	Kansas City Chiefs	0.0	0.0	-0.602	-0.602
6	San Diego Chargers	0.0	0.0	-0.118	-0.118

 

The total score of a particular team is used as a base. Subtract the score of the opponent and the result is the logit of the win probability for that game. You can use the inverse logit (see Wolfram Alpha to do this easily) to get the probability, and you can multiply the logit of the win probability by 7.4 to get the estimated point spread.

For the first week of the playoffs, I’ve done all this for you, in the table below. Odds are presented from the home team’s point of view.

First Round Playoff Odds
Home Team	Visiting Team	Score Diff	Win Prob	Est. Point Spread
Philadelphia Eagles	New Orleans Saints	-0.914	0.286	-6.8
Green Bay Packers	SF 49ers	-1.049	0.259	-7.8
Cincinnati Bengals	San Diego Chargers	0.101	0.525	0.7
Indianapolis Colts	Kansas City Chiefs	1.650	0.839	12.2

 

Some general conclusions from the data above: the teams my model favors most are the Seattle Seahawks, the Indianapolis Colts, the 49ers, the Carolina Panthers, and then the New Orleans Saints, mostly NFC teams. Since the Super Bowl itself does not have a home team, the odds change once you actually reach the Super Bowl. The sum of the SOS column and the Previous Playoff Experience column can be used to estimate odds of winning “the big one”. The strongest team in a Super Bowl setting would be the San Francisco 49ers, with a total score, less HFA, of 1.049. The Indianapolis Colts, with a total score of 0.643 less HFA, would be the strongest possible AFC contender.

A point I’d like the reader to consider is this question: should the New Orleans Saints be granted an exception to the previous playoff experience rule of “last year only counts” and given the 0.434 advantage of a playoff team? 2012 was an aberration as the coach was suspended. I’m not calculating this variation into the formula at this point, but I’ll note that this is an issue that you, the reader, need to resolve for yourself.

The road to the playoffs is not easy, a topic that can be studied by trying to calculate the path to the playoffs of the Indianapolis colts, a team that would be favored in every matchup along the way. Let’s calculate the odds of Indianapolis actually winning all three games.

Odds of Indianapolis reaching the Super Bowl
WP versus Kansas City	WP versus Denver Broncos	WP versus NE Pats	Cume Probability
0.839	0.586	0.515	0.253

 

Three teams from the NFC would be favored over any possible AFC contender. Those are San Francisco, Seattle, and the New Orleans Saints. Carolina would be favored over any AFC contender except the Indianapolis Colts.

 

NFL Season 2013: Week 13

Sorry about any delays in publication. I was between jobs at the time.

Week 13 NFL Stats:

To explain the columns above, Median is a median point spread, and can be used to get a feel for how good a team is without overly weighting a blowout win or blowout loss. HS is Brian Burke’s Homemade Sagarin, as implemented in Maggie Xiong’s PDL::Stats. Pred is the predicted Pythagorean expectation. The exponent for this measure is fitted to the data set itself. SOS, SRS, and MOV are the simple ranking components, analyzed via this Perl implementation. MOV is margin of victory, or point spread divided by games played. SOS is strength of schedule. SRS is the simple ranking.

OSRS and DSRS stats look like this:

The two most impressive teams so far, IMO, are Seattle and Carolina. New Orleans may win the division but right now Carolina is something of a statistical darling.

 

NFL Season 2013: Week 8

Conversation around Atlanta is that the Falcons are effectively out of the playoff hunt, as they would need to go 8-1 to be back in it. Personally, I don’t see how they can become a better team than Carolina at this point, much less New Orleans.

Week 8 NFL Stats:

To explain the columns above, Median is a median point spread, and can be used to get a feel for how good a team is without overly weighting a blowout win or blowout loss. HS is Brian Burke’s Homemade Sagarin, as implemented in Maggie Xiong’s PDL::Stats. Pred is the predicted Pythagorean expectation. The exponent for this measure is fitted to the data set itself. SOS, SRS, and MOV are the simple ranking components, analyzed via this Perl implementation. MOV is margin of victory, or point spread divided by games played. SOS is strength of schedule. SRS is the simple ranking.

OSRS and DSRS stats look like this:

I’m very tempted to start figuring out who would win playoffs if results were frozen and teams were to go into rounds of games at this time. That’s something for a future post, though.

 

NFL Season 2013: Week 7

As #1 teams go, Kansas City has had an exceptionally weak set of opponents and as a #1, Kansas City looks to be had. The schedule doesn’t give them very many hard opponents. Denver looms in week 11 and 13, and Indianapolis in week 16, but otherwise the schedule favors this team — until the playoffs.

Week 7 NFL Stats:

To explain the columns above, Median is a median point spread, and can be used to get a feel for how good a team is without overly weighting a blowout win or blowout loss. HS is Brian Burke’s Homemade Sagarin, as implemented in Maggie Xiong’s PDL::Stats. Pred is the predicted Pythagorean expectation. The exponent for this measure is fitted to the data set itself. SOS, SRS, and MOV are the simple ranking components, analyzed via this Perl implementation. MOV is margin of victory, or point spread divided by games played. SOS is strength of schedule. SRS is the simple ranking.

OSRS and DSRS stats look like this:

The OSRS and SDSR stats are calculated as described here. The top 5 teams in OSRS turn out to be Denver, Chicago, Indianapolis, Dallas, and Green Bay. The top 5 teams in DSRS are Carolina, Kansas City, Seattle, San Francisco, and New Orleans. Carolina’s stats in general are notable, as they have the second best Pythagorean in the league.

 

NFL Season 2013: Week 6

Week 6 stats:

To explain the columns above, Median is a median point spread, and can be used to get a feel for how good a team is without overly weighting a blowout win or blowout loss. HS is Brian Burke’s Homemade Sagarin, as implemented in Maggie Xiong’s PDL::Stats. Pred is the predicted Pythagorean expectation. The exponent for this measure is fitted to the data set itself. SOS, SRS, and MOV are the simple ranking components, analyzed via this Perl implementation. MOV is margin of victory, or point spread divided by games played. SOS is strength of schedule. SRS is the simple ranking.

Week 6 OSRS and DSRS stats:

It’s interesting that while the #1 DSRS is Kansas City (expected), the #2 is Carolina (not so expected). The #1 OSRS is Denver, and the #2 for now appears to be Dallas. Dallas’s offense SRS can’t be assigned entirely to the offense. Monte Kiffin’s defense emphasizes turnovers and the Dallas special teams are scoring as well.

 

NFL Season 2013: Week 5

Week 5, NFL scoring stats.

To explain the columns above, Median is a median point spread, and can be used to get a feel for how good a team is without overly weighting a blowout win or blowout loss. HS is Brian Burke’s Homemade Sagarin, as implemented in Maggie Xiong’s PDL::Stats. Pred is the predicted Pythagorean expectation. The exponent for this measure is fitted to the data set itself. SOS, SRS, and MOV are the simple ranking components, analyzed via this Perl implementation. MOV is margin of victory, or point spread divided by games played. SOS is strength of schedule. SRS is the simple ranking.

I don’t like working stats the first couple weeks of the season, because they don’t mean very much. But for those who might want some week 4 stats, here they are.

For those with sharp eyes, Pro Football Reference’s 2013 stats, specifically their SOS and SRS stats, have been busted. This is a screen shot of page http://www.pro-football-reference.com/years/2013/ around 8:30am on this October 11, 2013.

The simple ranking (SRS) is supposed to equal MOV + SOS. On page above, it does not. (seen October 11, 2013).

Simply put, MOV + SOS is supposed to equal SRS. It, in general, does not. Normally my SRS values are within one tenth of a point of PFR’s, but not today, and the failure of the calculation on their part is pretty plain to see.

Nathan Oyler, via Twitter, started asking me the following question, how can my code be used to calculate offensive SRS and defensive SRS? I looked a bit and there doesn’t seem to be a good definition for the term. Some comments by Chase Stuart (1) lead to the notion that you start with “points for” and “points against” stats, and that gives us some useful clues.

If SRS can be broken into two components, OSRS and DSRS, for offensive and defensive SRS, then so can margin of victory. The sum of offensive margin of victory and defensive margin of victory have to equal margin of victory, so it then becomes possible to define such a term.

If you take “points scored” or “points for” and add them all up, divide by the total number of games played, you get the average number of points scored, by one side, in a game. We’re going to call this the average score for now. You can then define the offensive and defensive mov in this way:

offensive mov = (“points for” – games_played(team)* avg_score ) / games_played(team)

defensive mov = ( games_played(team) * avg_score – “points against”) / games_played(team)

These definitions work the way they should. Add the two together and you get margin of victory. Put these terms in place of the MOV in a SRS calculation, and you have a way to calculate an offensive SOS and a defensive SOS. There are still some wrinkles, and those belong in another publication. That’s enough for now, though.

Notes

1. See Chase’s reply to Racer1.

 

Thoughts on NFL week 11

Over some five years, the whole of the Matt Ryan – Mike Smith era, Atlanta has had a habit of outperforming its Pythagoreans:

Atlanta outperforming its Pythagoreans
Year	WL%	Pythag	Delta
2008	69	62	7
2009	56	56	0
2010	81	72	9
2011	63	59	4
2012 (to date)	90	71	19

 

But they’ve never outperformed their Pythagoreans as substantially as they have this year. It can’t be blamed on early season New Orleans collapse, as their only loss was inflicted by New Orleans. New Orleans has only hindered this process. Is it turnover that are causing all this? While the 2010 team had a +14 turnover ratio and the 2011 team had a +8 turnover ratio, the 2012 team has only a +5 turnover ratio at this point and the 2008 team had a -3 turnover ratio. No, it’s something else. For now, perhaps noting that this team tends to outperform its Pythagoreans is enough.

Week 11 scoring stats:

Chicago’s biggest weakness was on display this Monday night, as Aldon Smith had a career day. Aaron Schatz (@FO_Schatz) has sent digging into his archives for the biggest DVOA blowouts of all time. The 32-7 demolition of the Bears by the 49ers wasn’t the worst, but it clearly evoked the worst.

The game plan was heavy on traps and wham blocks, and would have warmed the hearts of anyone who ever played NFL Strategy against a blitz heavy opponent.

It does lead to the question of whether Chicago is in the same downward spiral they experienced last year. At this point, however, you would expect Jay Cutler to return and thus slow down the bleeding.

 

Week Ten thoughts

I believed, in the immediate aftermath of the 2011 season, that with Jason Peters at left tackle, the least of Philadelphia’s worries would have been the tackle position. Instead, he was injured in the off season. In September, Philadelphia center Jason Celce went down with a season ending injury. In the New Orleans game, Todd Herremans suffered a season ending injury, and going into the Dallas game, starting guard Danny Watkins had been out with a sprained ankle.

Losing Todd Herremans: deal breaker for the Eagles? (Image from Wikimedia).

So, in week 10, the Eagles had one healthy starting caliber player, and 4 backups playing on the offensive line. This loss of talent was profound, even in comparison with Dallas, which had 1 backup on the line – though Dallas RG Mackenzie Bernadeau has been pretty marginal as a starter. Simplified, losing tackles is much worse than losing a guard and a center. Result? A markedly ineffective Vick, a thoroughbred offense reduced to dog-sled pace.

No wonder announcers were hyping this as the “end of a season” for one of these teams. Most any cold blooded announcer could have figured out what was about to happen. The only question was how best to pitch it so people would actually watch.

Atlanta: I’ve been comparing the 2012 Atlanta Falcons to the 1976 Oakland Raiders, to make the case that Atlanta has a chance. But the 1976 Raiders had made it to three previous Conference Championship games, while the Mike Smith squads have never gone that far. They lack the deep playoff experience of those 1970s Raiders squads.

The fact is, all scoring stats suggest Atlanta has benefited from plenty of luck. I think, because of a better Julio Jones, that this is a better Falcons team than the 2011 team, but the coaching changes in New Orleans markedly benefited this squad. Yes, Atlanta can be beaten.

Week 9 scoring stats:

Week 10 scoring stats:

If we use the median point spread as a measure of how good Atlanta is, and select the teams within 2 points of their value, you end up with a group that includes San Francisco, New England, Minnesota, and the New York Giants. That’s a talented group of teams, but perhaps not as terrifying as Green Bay, Houston, Denver, and Chicago. Pythagoreans point out three elite teams in Houston, Chicago, and San Francisco, while simple rankings prefer the quartet of Houston, Chicago, Denver and San Francisco.

At this point, perhaps the more appropriate past comparison for the Falcons would be the 1973 Oakland Raiders. Atlanta needs to make some noise in the playoffs first.

Should anyone be worried about the Giants mid season slide? No. They always do this. The question is, will they fully recover in time to make a playoff run. That’s not something that will be entirely answered until week 17.

 

NFL Season 2012: Week 8

Week 8, NFL scoring stats:

To explain the columns above, Median is a median point spread, and can be used to get a feel for how good a team is without overly weighting a blowout win or blowout loss. HS is Brian Burke’s Homemade Sagarin, as implemented in Maggie Xiong’s PDL::Stats. Pred is the predicted Pythagorean expectation. The exponent for this measure is fitted to the data set itself. SOS, SRS, and MOV are the simple ranking components, analyzed via this Perl implementation. MOV is margin of victory, or point spread divided by games played. SOS is strength of schedule. SRS is the simple ranking.

NFC East teams with aspirations of playoff contention, Dallas and Philadelphia, appear to be having them derailed by a single common cause. Neither has an offensive line healthy enough or good enough to let their quarterbacks shine. Further, being 1.5 games behind their wild card competition in the NFC North leaves them with precious few chances. Perhaps they’ll improve, perhaps Philadelphia will find a miracle LT and Dallas will scrape together a center and a right guard, but don’t hold your breath waiting. Washington is dynamic, but needs a few pieces here and there. Early season defensive line injuries did that team little good.

For now, what interests me are things like: who in the AFC can compete with Houston? Baltimore looks as if it’s winning on tradition more than dominance. New England looks great, except when it doesn’t. Denver is a powerful work in progress. In the NFC, it’s entirely possible that the NFC North will field 3 playoff teams. Chicago, Green Bay and Minnesota look good. Atlanta continues to roll up wins, the last perhaps its most impressive so far. The Giants continue to show strength. The San Francisco 49ers may be the best team in football right now, leading in HS, #2 in Simple Ranking, and no more than 0.3% off the top of the Pythagoreans.

Totally off the subject: this is the political season, and some of my favorite bloggers are tweeting some politics these days. One of the most interesting of the lot is @skepticalsports. I don’t share the political sentiments of Benjamin Morris, but polite and political – which he manages to do – is a rare combination, and it actually takes some work to be offended by his tweets.

 

NFL Season 2012: Week 7

Week 7, NFL scoring stats:

To explain the columns above, Median is a median point spread, and can be used to get a feel for how good a team is without overly weighting a blowout win or blowout loss. HS is Brian Burke’s Homemade Sagarin, as implemented in Maggie Xiong’s PDL::Stats. Pred is the predicted Pythagorean expectation. The exponent for this measure is fitted to the data set itself. SOS, SRS, and MOV are the simple ranking components, analyzed via this Perl implementation. MOV is margin of victory, or point spread divided by games played. SOS is strength of schedule. SRS is the simple ranking.

One of the things dogging Atlanta sports talk radio is “just how good are the Atlanta Falcons”? Statistically, they’re in the top 5-10 but not the very top in the various scoring stats. A lot of their success is based on turnover differential, not a good predictor of success over the long term. The Houston Texans, by contrast, shook off their one game blues are are back in the top 2 or 3 once again.

Chicago for now is at the head of the scoring stats, and those of us familiar with Jay Cutler’s ability to have really bad games will be watching to see if he can keep it up. Coming hard are both the Giants and the Packers. Denver is the best looking of the 3-3 teams.