FeinSports.com

2010 FEIN Fantasy Football Projections

zachfein — Sat, 14 Aug 2010 21:41:56 +0000

The NFL preseason is already underway, and fantasy football players are prepping for the most critical part of the season–the draft. While some people choose to rank players based on gut feeling and intuition, I like to go with a more mathematical approach, making projections based on past performance and historical trends.

Last year, my projections beat out three of the top fantasy football Web sites, CBSSportsline, ESPN, and FFToday. This year, my projections should only get better. Here are what’s changed in my projections from last year to this:

1. Stadium factors. I had previously dabbled with this subject, but the data shown on pro-football-reference’s game logs (just points and yards) were not enough to include into my projections. But with PFR’s new team game finder, I was able to cull out the detailed stats for every game since 2005, which allowed me to get a stadium factor for every stat I projected (including, say, rushing touchdown percentage).

2. A touch-up to the aging adjustments. I now age adjust each season to age 27, and then adjust from age 27 to the player’s 2010 age for their projection. Steve Smith of the Giants had a breakout year last year, but he was also turning 24, the age at which WRs see their biggest statistical increase. This step gives a little of Smith’s breakout to his age.

3. Projections only take into account seasons from 2005 onward. There’s little effect of dropping 2002-2004 (since the years are weighted toward the recent, anyway), and it’s easier for me to work with the spreadsheet this way.

Playing time is still based on the projections of the three Web sites listed above, as it is downright impossible to objectively forecast playing time, especially for players like Kevin Kolb, who are taking over the starting role this year.

Rookies are also projected the same way as last year–use projected playing time as a proxy for 2009 stats, and calculate projections normally from there, with stadium factors included.

Without further ado….

Anyone who was projected to get playing time by any of the three Web sites is included. You can see a larger version here.

A few key notes to remember as you peruse the forecasts:

-Wide receivers have their biggest increase from 23 to 24. DeSean Jackson, Sidney Rice, and Mike Wallace are among those turning 24.

-For someone like DeSean Jackson, remember that the effect of McNabb’s departure is minimal–less than a three fantasy-point difference.

-All the projections know about Ryan Matthews, for instance, is that RBs who have 257 carries have a YPC of about 4.14 and that Matthews is 6′0″, 218 pounds and 23 years old, and playing for the Chargers. Nothing about his draft position, college stats, or Combine scouting report is included.

Fantasy Football: How do Players Perform when Their Teammates Change?

zachfein — Mon, 02 Aug 2010 23:00:56 +0000

With marquee players signing with new teams this offseason–LaDainian Tomlinson to the Jets, Terrell Owens to the Bengals, Anquan Boldin’s being traded to Baltimore–I thought it would be interesting to see how their old and new teammates’ performances change with their departure or arrival. For instance, how much does the arrival of a good WR help Joe Flacco?

I ran my projections for every player-season since 1980. I then looked at all players since 1980 who had 300 pass attempts, 150 carries, or 50 receptions, and switched teams from one year to the next. I compared my projections to the projections of the teammates’ of the player who switched. (I prorated my projections to each player’s actual playing time in the projected year, so that I only looked at the effect on per-attempt performance, and because there’s no reliable way to project playing time for football.)

Here are the results. ** denotes differences of more than five fantasy points; * denotes a difference of more than two fantasy points.

(Here’s how to read the first row: There were 41 quarterbacks whose running back left; they were projected to score 211.0 fantasy points, and they actually scored 210.6. That’s a difference of -0.4, meaning they performed slightly worse than expected.)

QBs	#	Proj	Act	Diff
RB left	41	211.0	210.6	-0.4
WR left	45	230.0	224.9	-5.0**
RB arrived	23	200.4	188.5	-11.9**
WR arrived	45	210.3	211.5	1.2

RBs	#	Proj	Act	Diff
QB left	37	111.1	112.7	1.6
RB left	50	95.4	89.3	-6.1**
WR left	50	132.9	134.6	1.7
QB arrived	34	127.8	130.0	2.3*
RB arrived	14	92.8	92.3	-0.5
WR arrived	58	134.8	136.0	1.2

WR	#	Proj	Act	Diff
QB left	87	86.1	83.7	-2.4*
RB left	123	74.7	77.6	2.9*
WR left	119	83.7	86.4	2.7*
QB arrived	74	87.6	87.1	-0.5
RB arrived	70	81.8	82.9	1.2
WR arrived	88	74.7	77.4	2.6*

So, what can we take away from the data? Not much. There were only three instances of a difference of more than five fantasy points:

-There were 45 quarterbacks whose wide receiver left, and they performed five points worse than expected. This makes intuitive sense.

-Quarterbacks who had a new RB performed 12 points worse than expected, but with a sample size of just 23, most of that is noise. Note that projections are based on actual pass attempts, so the difference would come in a QB’s per-attempt numbers–that would mean the new RB didn’t help take pressure off the quarterback, at least for the sample.

-Running backs who had a 150-carry back leave their team were six points worse than their projection the following year. This could be because they had a specialized role with the 150-carry RB on the team, and the following year they had to take on newer situations.

As for Carson Palmer and Joe Flacco, Owens’ and Boldin’s new quarterbacks? Passers who had a new receiver only performed 1.2 points above their projection–not much better.

2010 FEIN Projected Standings

zachfein — Fri, 26 Mar 2010 02:11:54 +0000

Sunday, I released my MLB projections based on a computer model. The individual projections can then be added up by team to create projected standings.

Before I add up any projections, I first adjust any player’s projected playing time that may be off. As I mentioned before, the projected playing time only looks at a player’s past playing time and adjusts for age. So a player such as Jose Reyes, who had had 700 plate appearances in his four seasons as a starter before missing 126 games last year, may have a lower projection than in reality. I also look for players like Rangers OF Julio Borbon, who will receive a starting role for the first time in his major league career.

Once each player is assigned their proper role and playing time forecast, I add up the projected stats for each team. If a team comes up over 4120 outs (AB minus H plus CS) or 1440 IP, I prorate their stats down to those levels; if a team falls short of those benchmarks, I assign replacement-level production for the remaining outs.

I use a Base Runs equation to calculate projected runs scored for offense, and for the pitching staff I simply look at runs allowed. Win percentage is found using Pythagenpat . Then I adjust each team’s win percentage by their schedule to come up with a final tally.

Here are the results:

American League
Team	W	L	RS/G	RA/G
AL East
New York Yankees	93	69	5.22	4.61
Tampa Bay Rays	92	70	4.84	4.29
Boston Red Sox	89	73	4.94	4.52
Baltimore Orioles	70	92	4.57	5.30
Toronto Blue Jays	70	92	4.36	5.03

AL Central
Minnesota Twins	86	76	4.98	4.82
Chicago White Sox	83	79	4.45	4.47
Cleveland Indians	78	84	4.66	4.92
Detroit Tigers	78	84	4.48	4.78
Kansas City Royals	74	88	4.30	4.80

AL West
Texas Rangers	81	81	4.80	4.97
Seattle Mariners	79	83	4.20	4.37
Los Angeles Angels	79	83	4.58	4.80
Oakland A’s	78	84	4.29	4.58

National League
Team	W	L	RS/G	RA/G
NL East
Atlanta Braves	89	73	4.86	4.34
Philadelphia Phillies	86	76	4.91	4.49
Florida Marlins	79	83	4.80	4.79
New York Mets	78	84	4.56	4.64
Washington Nationals	77	85	4.66	4.83

NL Central
St. Louis Cardinals	91	71	4.87	4.28
Chicago Cubs	85	77	4.70	4.44
Milwaukee Brewers	78	84	4.77	4.86
Cincinnati Reds	78	84	4.61	4.72
Houston Astros	76	86	4.55	4.76
Pittsburgh Pirates	75	87	4.54	4.84

NL West
Los Angeles Dodgers	87	75	4.78	4.33
Arizona Diamondbacks	84	78	4.94	4.62
Colorado Rockies	82	80	4.95	4.82
San Francisco Giants	79	83	4.37	4.38
San Diego Padres	77	85	4.34	4.48

It is important to remember that these are mean projections. Let’s give the Yankees a basic win spread:

85 wins: 15% of the time
90 wins: 30%
95 wins: 35%
100 wins: 20%

The weighted average of those four scenarios is 93 wins, the projection for New York. But 35% of the time (numbers for illustration only), their forecast would be off seven or eight games. So, there’s a large chance that the Rays would beat the Yankees for the AL East title, given their projections are the “true talent” levels (their average total if they played the season 1,000 times).

This is also why the win totals seem biased towards the middle. Most times that a team wins 65 or 95 games, their “true talent” is around five to ten wins closer toward .500 than they showed. If we project New York for 93 wins, there’s still about a 16% chance they win 102 (or so) or more games.

2010 FEIN Projections: Hitters and Pitchers

zachfein — Mon, 22 Mar 2010 02:47:51 +0000

In every aspect, these projections are better than last year’s. Why?

1. Custom weights. Each stat is weighted based on error tests from 1970 onward; for instance, BABIP for hitters is weighted at 0.88, while strikeout rate, which is more stable year-to-year than BABIP, is weighted at 0.49 (i.e., 2009 has a weight of 1, 2008 has a weight of 0.49, etc.). I use the past four years’ stats for hitters and three years for pitchers.

2. Personal characteristics factor into the projections. Weight, height, handedness, and speed score (for batters) are used to find custom aging curves for each stat (stolen bases peak before homers or walks) as well as to serve as the mean to which a player is regressed—heavier hitters, for example, hit more home runs, so they will be regressed to a higher average than a lighter hitter.

3. For pitchers, batted ball data—ground balls, line drives, and fly balls—are included in BABIP and HR projections. I project GB/LD/FB, and regress BABIP and HR rates to those numbers. A fly ball pitcher will allow more home runs than a ground ball pitcher, and the projections take this into account. Pitcher BABIP is further defense-adjusted based on Jeff Zimmerman’s UZR projections.

4. Saves are based on both subjective and objective data, in roughly the same process as last year. Some differences: The four closers were ranked based on multiple sources, not using projected ERA; the equation to find the percentage of team saves is different than last year, based on a larger sample size; and I used KFFL’s job security rankings in their closer depth chart to adjust the closer’s saves.

Projected saves are playing time-independent. Billy Wagner, Atlanta’s new closer, is projected to have 35 saves in 45 innings. The latter is based on his injury-plagued last two seasons, while the former is solely based on his ERA and Atlanta’s win total.

I gave Joe Nathan zero saves but left him and his forecast in the spreadsheet below, but note that his 2010 season is over after Tommy John surgery.

These are just some of the changes that I applied to last year’s algorithm. The forecasts, however, still do not take minor league numbers into account. As a result, there is no forecast for any 2010 rookies, and projections for players with little major league playing time may not be accurate. Nevertheless, I believe these projections are some of the best you’ll find out there, up there with CHONE and ZiPS.

Below you’ll find a Google spreadsheet with projections for any player who had at least one PA or faced on batter in 2009 or played in 2008 and will see playing time this season (such as Jim Edmonds or Jeff Francis).

Position is the player’s positional eligibility in a standard ESPN league (20 games to qualify at a position for hitters, five starts to qualify as SP, or eight relief appearances to qualify as RP); pitchers with a "P" did not play last year. Note that, for pitchers, their position listed is not necessarily the role I projected them to be. Kelvim Escobar is listed as a SP under ESPN’s guidelines, but I projected him to be a reliever as a Met.

Fantasy $ are found using LastPlayerPicked.com’s awesome Price Guide , which lets you upload your own projections to generate fantasy prices in any league. I used a 12-team league with one C, 1B, 2B, and 3B, three OF, and one UTIL for hitters and five SP, two RP, and two P slots for pitchers to find the dollar values.

Playing time is based on past playing time and age and does not take depth charts or freak injuries into account. As such, the PA and IP projections will not be that accurate, but the rate stats (such as AVG or HR/PA and the like) are unaffected. Feel free to substitute your own playing time forecasts into the projections below.

View the hitter and pitcher projections here

Download as an Excel file here

NCAA Tournament Trends

zachfein — Tue, 16 Mar 2010 18:58:28 +0000

Last year at this time I ran a series of posts about NCAA tournament trends—early-round upsets, sweet sixteen and elite eight matchups, and final four seeds. I won’t update it this year, but the main points still stand. I urge you to take 10 minutes out of your day to read the posts to help you fill out your bracket.

NFL Projections Accuracy: Is a Computer Better than Humans?

zachfein — Mon, 01 Feb 2010 03:45:56 +0000

How accurate were my objective FEIN projections ? You may think that a completely automated football forecasting system won’t be as accurate as human projections such as CBS Sportsline’s , FF Today’s , or ESPN’s . Let’s see how each set of projections stacked up in accuracy for the 2009 season.

I looked at quarterbacks with 200 pass attempts, running backs with 100 rushes, wide receivers with 40 catches, and tight ends with 30 catches this season (selecting only those with a projection from each system). There were 33 QBs, 47 RBs, 54 WRs, and 25 TEs who met these criteria.

I first adjusted each system’s projections up or down to make the average projection equal to actual average of the players in the sample. (For instance, if ESPN had projected 4.5 yards per carry and the average was 4.2, I subtracted three-tenths of a yard from each player’s projection.)

Then I compared each player’s adjusted projection to his actual number. I squared the difference between the two and weighted that number based on his actual number of attempts or catches, and took the square root of the average for each system.

Here are the results.

Note: Projections were assessed on a per-attempt or per-catch basis. Projecting 1,000 rushing yards on 250 attempts is the same as projecting 400 yards on 100 attempts.

All numbers below are prorated to the average number of attempts or catches for the sample: 438 attempts for QBs; 199 rush attempts and 29 receptions for RBs; 65 receptions for WRs; and 57 receptions for TEs. The rank among the four systems is in parentheses.

Stat	CBS	FF Today	ESPN	FEIN
Comp	17.2 (2)	17.1 (1)	19.1 (4)	17.5 (3)
Pass Yds	366 (2)	338 (1)	386 (4)	384 (3)
Pass TD	5.10 (4)	4.22 (3)	4.16 (2)	4.01 (1)
INT	5.88 (2)	6.26 (4)	6.14 (3)	5.61 (1)
QB FanPts	46.8 (4)	42.9 (2)	44.4 (3)	42.6 (1)

Rush Yds	127 (3)	155 (4)	118 (1)	121 (2)
Rush TD	3.16 (1)	3.33 (2)	3.62 (4)	3.36 (3)
RB Rec Yds	51 (2)	54 (3)	63 (4)	49 (1)
RB Rec TD	1.28 (3)	1.1 (1)	1.60 (4)	1.24 (2)
RB FanPts	44.5 (1)	47.5 (3)	49.5 (4)	44.6 (2)

WR Rec Yds	123 (2)	133 (3)	163 (4)	117 (1)
WR Rec TD	2.29 (1)	2.35 (2)	2.61 (4)	2.46 (3)
WR FanPts	26.0 (1)	27.4 (3)	31.9 (4)	26.4 (2)

TE Rec Yd	76 (1)	88 (3)	97 (4)	81 (2)
TE Rec TD	2.29 (1)	2.73 (3)	2.73 (4)	2.49 (2)
TE FanPts	21.4 (1)	25.2 (3)	26.1 (4)	23.1 (2)

Avg. FanPts	34.7 (2)	35.7 (3)	38.0 (4)	34.2 (1)

CBS Sportsline won three out of the four positions, but on average, my projections were the best overall, with an average error of 34.2 fantasy points. Nevertheless, it’s only one year, and the sample size is small. But it just goes to show that objective forecasts can be as accurate as subjective forecasts.

Week 16 NFL Ratings and Week 17 Predictions

zachfein — Wed, 30 Dec 2009 00:02:30 +0000

The Week 16 NFL ratings are below. Ratings are calculated based on an Elo backbone using both margin of victory as well as just the result of the team’s game. Final ratings, represented on a points scale—a 21 rating is three points better than an 18 rating on a neutral field—are an average of these two individual scores.

The Elo method rewards teams for performing better than expected and punishes teams for playing worse than expected. For instance, a team that is 15-point favorites and only wins by 10 will see a decline in their rating.

These ratings do not take into account teams who rest players, such as the Colts, who lost 4.74 points on their rating after losing to the Jets by 14 points. As well, Week 17 predictions are based on full-strength teams—the Patriots are three-point favorites by these ratings yet eight-point underdogs to Houston based on Vegas spreads because bookies believe the Pats will rest their starters.

Both New England and Cincinnati, however, are fighting for the No. 3 seed and will, I believe, play the whole game with starters in.

Click on table headers to sort.

Rank	Team	Rating	W-L	Pts	Jump
1	San Diego Chargers	31.78	33.71	29.86	+3.7
2	Indianapolis Colts	28.99	32.98	25.01	-4.74
3	New England Patriots	28.99	26.8	31.18	+2.02
4	Philadelphia Eagles	28.85	29.82	27.88	-1.08
5	Dallas Cowboys	27.5	26.66	28.35	+1.52
6	Carolina Panthers	26.64	22.53	30.75	+2.13
7	New Orleans Saints	26.38	30.05	22.71	-2.72
8	Green Bay Packers	25.83	25.26	26.39	+1.47
9	Atlanta Falcons	25.08	22.34	27.83	+2.21
10	New York Jets	24.79	22.19	27.38	+4.53
11	Arizona Cardinals	24.11	24.9	23.31	+1.72
12	Houston Texans	23.79	22.93	24.64	+1.47
13	Tennessee Titans	23.75	25.61	21.9	-2.64
14	Baltimore Ravens	23.23	21.64	24.83	-3.59
15	Cincinnati Bengals	23.23	24.59	21.87	+0.98
16	San Francisco 49ers	22.67	20.74	24.6	+0.88
17	Pittsburgh Steelers	21.6	21.75	21.46	+3.74
18	Minnesota Vikings	21.43	23.74	19.12	-2.98
19	New York Giants	20.87	20.53	21.2	-4.06
20	Denver Broncos	20.55	19.37	21.72	+0.4
21	Miami Dolphins	20.14	19.77	20.52	-2.1
22	Cleveland Browns	20.07	17.91	22.23	+3.45
23	Chicago Bears	18.99	18.9	19.08	+4.09
24	Tampa Bay Buccaneers	18.71	16.52	20.9	+3.66
25	Jacksonville Jaguars	17.18	18.21	16.14	-3
26	Washington Redskins	15.48	14.73	16.24	-1.16
27	Buffalo Bills	14.88	15.17	14.6	-3.4
28	Oakland Raiders	14.55	16.04	13.06	-4.31
29	Kansas City Chiefs	12.91	11.96	13.86	-0.99
30	Seattle Seahawks	11.09	13.84	8.33	+0.37
31	Detroit Lions	10.35	10.09	10.61	-1.11
32	St. Louis Rams	9.59	9.07	10.11	-1.23

Last week’s picks went 9-for-16; through eight weeks of predictions, the model currently stands at a 77-46, a 63 percent pick rate. Here are this week’s forecasts, calculated as the difference in ratings plus a home field adjustment; through 16 weeks, home teams have won by an average of 2.2 points.

Predicted spread is in terms of the home team. The Colts are 11.9-point favorites (with starters in for the whole game), for instance. Vegas spreads are also shown for comparison.

Home	Road	Spread	Vegas
Buffalo Bills	Indianapolis Colts	+11.9	N/A
Carolina Panthers	New Orleans Saints	-2.4	N/A
Cleveland Browns	Jacksonville Jaguars	-5.1	-1.5
Arizona Cardinals	Green Bay Packers	-0.5	-3
Dallas Cowboys	Philadelphia Eagles	-0.8	-3
Denver Broncos	Kansas City Chiefs	-9.8	-13
Detroit Lions	Chicago Bears	+6.5	+3
Houston Texans	New England Patriots	+3	-8
Miami Dolphins	Pittsburgh Steelers	-0.7	PK
Minnesota Vikings	New York Giants	-2.8	-8.5
New York Jets	Cincinnati Bengals	-3.7	-10
Oakland Raiders	Baltimore Ravens	+6.5	+10.5
St. Louis Rams	San Francisco 49ers	+10.9	7
San Diego Chargers	Washington Redskins	-18.5	-4
Seattle Seahawks	Tennessee Titans	+10.5	+4.5
Tampa Bay Buccaneers	Atlanta Falcons	+4.2	+3

Week 15 NFL Ratings and Week 16 Predictions

zachfein — Tue, 22 Dec 2009 21:22:01 +0000

The Week 15 NFL ratings are below. Ratings are calculated based on an Elo backbone using both margin of victory as well as just the result of the team’s game. Final ratings, represented on a points scale—a 21 rating is three points better than an 18 rating on a neutral field—are an average of these two individual scores.

The Elo method rewards teams for performing better than expected and punishes teams for playing worse than expected. For instance, a team that is 15-point favorites and only wins by 10 will see a decline in their rating.

Five teams that were predicted to win by seven or more points lost last week, and, as such, the ratings see a major shakeup this week. Twelve teams see a three-point swing in their rating from last week, and the Saints fall below No. 2 for the first time this season.

Click on table headers to sort.

Rank	Team	Rating	W-L	Pts	Jump
1	Indianapolis Colts	33.73	35.49	31.97	+0.26
2	Philadelphia Eagles	29.93	28.09	31.76	+4.35
3	New Orleans Saints	29.1	33.2	25.01	-4.4
4	San Diego Chargers	28.08	31.82	24.35	-0.23
5	New England Patriots	26.97	25.05	28.89	+1.44
6	Baltimore Ravens	26.83	24.43	29.22	+2.98
7	Tennessee Titans	26.39	26.34	26.44	+1.35
8	Dallas Cowboys	25.98	25.79	26.16	+3.77
9	New York Giants	24.93	22.56	27.29	+4.44
10	Carolina Panthers	24.51	20.54	28.48	+6.49
11	Minnesota Vikings	24.41	27.21	21.62	-6.06
12	Green Bay Packers	24.36	24.12	24.6	-3.56
13	Atlanta Falcons	22.87	20.62	25.12	+3.82
14	Arizona Cardinals	22.38	24.38	20.39	-1.35
15	Houston Texans	22.32	21.16	23.48	-0.46
16	Cincinnati Bengals	22.25	23.73	20.77	-1.12
17	Miami Dolphins	22.24	21.51	22.97	+0.08
18	San Francisco 49ers	21.79	19.72	23.85	-2.47
19	New York Jets	20.26	19.07	21.45	-2.62
20	Jacksonville Jaguars	20.17	20.03	20.32	+1.72
21	Denver Broncos	20.14	21.02	19.27	-2.77
22	Oakland Raiders	18.86	18.99	18.74	+3.01
23	Buffalo Bills	18.29	17.19	19.39	-0.53
24	Pittsburgh Steelers	17.86	19.11	16.61	+2.02
25	Washington Redskins	16.64	15.61	17.67	-3.71
26	Cleveland Browns	16.62	15.2	18.03	+1.14
27	Tampa Bay Buccaneers	15.05	13.1	17	+3.75
28	Chicago Bears	14.9	15.8	14	-3.06
29	Kansas City Chiefs	13.9	13.31	14.48	-0.46
30	Detroit Lions	11.46	11.58	11.34	+0.42
31	St. Louis Rams	10.82	10.08	11.57	+0.32
32	Seattle Seahawks	10.72	15.47	5.96	-6

Last week’s picks went 8-for-16; through seven weeks of predictions, the model currently stands at a 68-39, a 64 percent pick rate. Here are this week’s forecasts, calculated as the difference in ratings plus a home field adjustment; through 15 weeks, home teams have won by an average of 2.1 points.

Predicted spread is in terms of the home team. The Titans are 0.4-point favorites, for instance. Vegas spreads are also shown for comparison.

Home	Road	Spread	Vegas
Tennessee Titans	San Diego Chargers	-0.4	-3
Atlanta Falcons	Buffalo Bills	-6.6	-8.5
Cincinnati Bengals	Kansas City Chiefs	-10.4	-14
Cleveland Browns	Oakland Raiders	+0.2	-3.5
Indianapolis Colts	New York Jets	-15.5	-5.5
Arizona Cardinals	St. Louis Rams	-13.6	-14
Green Bay Packers	Seattle Seahawks	-15.7	-14
Miami Dolphins	Houston Texans	-2.0	-3
New Orleans Saints	Tampa Bay Buccaneers	-16.1	-14
New England Patriots	Jacksonville Jaguars	-8.9	-7.5
New York Giants	Carolina Panthers	-2.5	-7
Philadelphia Eagles	Denver Broncos	-11.8	-7
Pittsburgh Steelers	Baltimore Ravens	+6.9	-2.5
San Francisco 49ers	Detroit Lions	-12.4	-12
Washington Redskins	Dallas Cowboys	+7.3	+7
Chicago Bears	Minnesota Vikings	+7.5	+7

Week 14 NFL Ratings and Week 15 Predictions

zachfein — Tue, 15 Dec 2009 23:40:15 +0000

The Week 14 NFL ratings are below. Ratings are calculated based on an Elo backbone using both margin of victory as well as just the result of the team’s game. Final ratings, represented on a points scale—a 21 rating is three points better than an 18 rating on a neutral field—are an average of these two individual scores.

The Elo method rewards teams for performing better than expected and punishes teams for playing worse than expected. For instance, a team that is 15-point favorites and only wins by 10 will see a decline in their rating.

Click on table headers to sort.

Rank	Team	Rating	W-L	Pts	Jump
1	New Orleans Saints	33.51	35.32	31.7	-1.75
2	Indianapolis Colts	33.48	34.28	32.68	+0.24
3	Minnesota Vikings	30.47	30.1	30.84	+2.88
4	San Diego Chargers	28.32	29.79	26.84	+0.05
5	Green Bay Packers	27.92	26.97	28.87	-0.33
6	Philadelphia Eagles	25.58	26.32	24.84	+0.01
7	New England Patriots	25.53	23.95	27.11	+1.85
8	Tennessee Titans	25.04	23.98	26.1	+0.55
9	San Francisco 49ers	24.26	21.11	27.41	+4.15
10	Baltimore Ravens	23.85	22.87	24.83	+1.39
11	Arizona Cardinals	23.73	23.79	23.67	-4.94
12	Cincinnati Bengals	23.37	25.18	21.56	-3.19
13	Denver Broncos	22.92	23.35	22.48	-1.49
14	New York Jets	22.87	20.9	24.84	+0.24
15	Houston Texans	22.78	20.65	24.9	+3.87
16	Dallas Cowboys	22.2	23.03	21.37	+0.21
17	Miami Dolphins	22.16	23.21	21.11	+0.91
18	New York Giants	20.49	21.54	19.44	-0.09
19	Washington Redskins	20.35	16.64	24.05	+1.75
20	Atlanta Falcons	19.06	18.88	19.23	+1.36
21	Buffalo Bills	18.81	18.15	19.48	+0.78
22	Jacksonville Jaguars	18.46	20.68	16.23	-1.37
23	Carolina Panthers	18.02	17.45	18.59	-1.9
24	Chicago Bears	17.96	17.41	18.51	-0.1
25	Seattle Seahawks	16.72	18.17	15.28	-4.11
26	Oakland Raiders	15.85	16.66	15.04	-2.09
27	Pittsburgh Steelers	15.85	16.26	15.44	-3.71
28	Cleveland Browns	15.48	13.46	17.49	+4.24
29	Kansas City Chiefs	14.35	15.27	13.43	-0.85
30	Tampa Bay Buccaneers	11.3	10.62	11.98	-0.31
31	Detroit Lions	11.04	12.19	9.9	-1.2
32	St. Louis Rams	10.5	10.73	10.28	-0.54

Last week’s picks went 12-for-16; through six weeks of predictions, the model currently stands at a 60-31, a 66 percent pick rate. Here are this week’s forecasts, calculated as the difference in ratings plus a home field adjustment; through 14 weeks, home teams have won by an average of 2.3 points.

Predicted spread is in terms of the home team. The Jaguars are underdogs by 12.4 points, for instance. Vegas spreads are also shown for comparison.

Home	Road	Spread	Vegas
Jacksonville Jaguars	Indianapolis Colts	+12.7	+3
New Orleans Saints	Dallas Cowboys	-13.6	-7
Buffalo Bills	New England Patriots	+4.4	+7
Carolina Panthers	Minnesota Vikings	+10.1	+8
Denver Broncos	Oakland Raiders	-9.4	-14
Detroit Lions	Arizona Cardinals	+10.4	+11
Kansas City Chiefs	Cleveland Browns	-1.2	-2
New York Jets	Atlanta Falcons	-6.2	N/A
Tennessee Titans	Miami Dolphins	-5.2	-3
Philadelphia Eagles	San Francisco 49ers	-3.7	-9
Pittsburgh Steelers	Green Bay Packers	+9.7	-1
St. Louis Rams	Houston Texans	+9.9	+10
Baltimore Ravens	Chicago Bears	-8.2	-10.5
San Diego Chargers	Cincinnati Bengals	-7.3	-6.5
Seattle Seahawks	Tampa Bay Buccaneers	-7.8	-7
Washington Redskins	New York Giants	-2.2	+3

The Colts’ and Saints’ Chances of 16-0

zachfein — Tue, 08 Dec 2009 23:44:52 +0000

Through 13 weeks of football, both the Indianapolis Colts and New Orleans Saints are 12-0 with no huge upset special looming in the final four weeks of the season. Although both of their toughest games (Colts vs Broncos and Saints vs Cowboys) are at home, the chances of an undefeated season aren’t 100 percent or even 90 percent.

If the Colts have a 90 percent chance to win each of their remaining games, their probability of a 16-0 season is .90 x .90 x .90 x .90, or 65.6 percent. Using the Week 13 NFL ratings, we can estimate the win probability for each of their remaining games and, thus, their likelihood of an undefeated season.

For instance, the Colts (33.2 rating) play the Broncos (24.4) at home this week. That gives the Colts a 66.4 percent chance to win, using a Pythagorean formula that accounts for home field advantage. Here’s a list of the Colts’ remaining games and their predicted win probability:

Broncos	0.66
@ Jags	0.84
Jets	0.71
@ Bills	0.87

Multiplying each of these probabilities together yields a 34.7 percent chance of an undefeated season. Indianapolis’s remaining opponents have an average rating of 21.2—just under league average.

Here’s the same for the Saints:

@ Falcons	0.90
Cowboys	        0.76
Buccaneers	0.94
@ Panthers	0.86

The Saints’ easier schedule (an average opponent rating of 17.8, which ranks just below the Raiders’ 17.94 rating at No. 27) gives them a 55.2 percent shot at 16-0.