Sunday, September 30, 2007



Atlanta +3 vs. Houston
Arizona +6.5 vs. Pittsburgh

San Diego -11.5 vs KC
Philly -3 at NYG


St Louis +13 at DALLAS
Baltimore -4 at CLEVELAND
NY Jets -3.5 at BUFFALO
MINNESOTA +3 vs Green Bay
DETROIT +3 vs Chicago

CINCY +7.5 vs New England
Seattle -2 at SAN FRAN

Tampa Bay +3 at CAROLINA
MIAMI -3.5 vs Oakland
Denver +10 at INDY

Sunday, September 23, 2007


Quick Picks this week:

Buffalo +16.5

NY Giants +4.5

Minnesota +3
Jacksonville +3.5
Cincinnati +3


Arizona +7.5

Sunday, September 16, 2007


Last week:

Level of picks:
5: 0-0
4: 0-0
3: 2-1-1
2: 2-5
1: 1-2-1

Last week, I didn't like all that much, and should have stayed away from the over/under "guesses". This week, I do like more games, including making one my first 5-rated pick.

1. ARIZONA +2 1/2 vs. Seattle (5). Arizona's D actually looked quite good for 58 minutes of the game against San Fransisco. Seattle's offense is banged up. Throw in the "2nd year in a new stadium" home field advantage boost and the divisional climate advantage, and I like Arizona outright here.
2. Washington +7 at PHILADELPHIA (4).
3. Kansas City +13 at CHICAGO (4).
4. Houston +7 at CAROLINA (3).
5. MIAMI +4 vs. Dallas (3).
6. Oakland +10 at DENVER (3)
7. Buffalo +10 at PITTSBURGH (2)
8. Cincinnati -8 at CLEVELAND (2)
9. ARI/SEA UNDER 42.5 (2).
10. New York Jets +10 at BALTIMORE (2).
11. San Fransisco +3 at ST. LOUIS (2).
12. TAMPA +4 1/2 vs. New Orleans (1).
13. KC/CHI UNDER 34.5 (1).
14. NY GIANTS -3 vs. Green Bay (1).
15. NEW ENGLAND -4 vs. San Diego (1).
16. NO/TB OVER 41.5. (1)
17. JAC/ATL OVER 34.5 (1)
18. HOU/CAR OVER 38.5 (1).

Friday, September 7, 2007


Just to make a record, to see how good or bad I do this year in predicting games, I am going to post about 15-20 picks per week here. Maybe some commentary, maybe not. I am going to assign a confidence rating of 1-5 on each pick, with most being lower. In other words, random guesses.

For week one, for the most part, no real strong feelings. There are three divisional games between outdoor opponents who have similar climates and are from the same time zone. Over the last 10 years, the home team in these matchups is 115-148-5 ATS (44%). That trend is slightly stronger in the second matchup between teams rather than the first. The game also goes "under" 54% of the time (55% of the time in the first matchup). So, I will be making some picks based on that.

Here is my first attempt at picks in 2007:

  1. HOUSTON -3 vs. Kansas City. (3) Chiefs O-line is a mess, offense has not had opportunity to play together. I don't see the 2007 Chiefs as being a poor road team, and it gets started in week 1.
  2. NE/NYJ UNDER 41. (3). See above on same climate rivals.
  3. New England -6 at NY JETS. (3). Ditto.
  4. WASHINGTON -3 vs. Miami (3).
  5. Atlanta +3 at MINNESOTA (2).
  6. Pittsburgh -4 1/2 at CLEVELAND (2).
  7. PIT/CLE UNDER 37.5 (2).
  8. BAL/CIN UNDER 40.5 (2).
  9. Tampa Bay +6 at SEATTLE (2).
  10. Baltimore +2 1/2 at CINCINNATI (2).
  11. BUFFALO +3 vs. Denver (2). This is the only West to East cross country in week one.
  12. CHI/SD OVER 42.5 (1).
  13. Arizona +3 at SAN FRAN (1).
  14. ATL/MIN OVER 36 (1).
  15. GREEN BAY +3 vs. Philadelphia (1).

Monday, April 23, 2007

Temperature Differences and Home Field Advantage

Back in October 2006, the pro-football-reference blog posted a two part guest post I prepared on the effects of travel distance and climate here, and

This is a refinement of that information, plus the info with all games, not just divisional.

I tweaked the methodology on climate slightly. Previously, I used the average temperature for the months of September through December for each outdoor city, and the difference between the averages for two cities. The only slight problem with this is that not all cities with similar average temperatures have similar climates. To try to capture this better, in this post, I used the average monthly difference between the cities in question, then re-sorted the data.

An example of this is Seattle and Chicago. These cities are within 1 degree in average temperature over the whole football season. However, Seattle is the coolest city in September, while Chicago is much warmer, but by the end of the season, the numbers have reversed. My new methodology was to look at the absolute difference in degrees for each month, and average them. The other refinement was a slight tweak to the dome average temperature, I used 72 degrees (the data supported this, as like the outdoor vs outdoor games, the home field advantage between dome and outdoor was weakest against teams between 65 and 76 degrees).

I have included dome teams, but for now, I am excluding Denver. Denver is a case study on its own. Here is the data for all other divisional games, using the new climate methodology.

within 5 degrees difference, monthly average: 289-259-1 (0.527)
5.1 - 10.0 degrees difference, monthly average: 303-267-1 (0.532)
10.1-15.0 degrees difference, monthly average: 136-98-0 (0.581)
15.1-20.0 degrees difference, monthly average: 199-126-0 (0.612)
20.1-25.0 degrees difference, monthly average: 144-85-0 (0.629)
25.1 + degrees difference, monthly average: 72-47-0 (0.605)

Again, the effect of climate differences in divisional games is real and fairly strong. However, it is at the similar temperatures, not the extreme ones, where the difference is noticeable. Once divisional opponents are outside a 10 degree difference, the home field advantage increases dramatically, and beyond 15 degrees, it is fairly constant.

However, the effect is not the same with non-divisional games. Here is the overall data for all games played between 1986-2005 (except involving Denver) using the same criteria.

within 5 degrees difference, monthly average: 575-453-4 (0.559)
5.1 - 10.0 degrees difference, monthly average: 653-498-1 (0.567)
10.1-15.0 degrees difference, monthly average: 450-318-0 (0.586)
15.1-20.0 degrees difference, monthly average: 529-364-1 (0.592)
20.1-25.0 degrees difference, monthly average: 227-158-0 (0.590)
25.1 + degrees difference, monthly average: 119-77-0 (0.607)

So, there is a split between divisional games and non-divisional games. The overall effect still appears when we combine both. Why is this? Other than a random split? Two guesses:

First, the NFL may schedule divisional contests at a higher frequency both early and more importantly, later in the season (they do have to get 2 in vs each divisional opponent), and these are the same times when weather differences will be more extreme.

Second, maybe there is also a familiarity component to reducing home field advantage, so that visitors from similiar climates, who are also familiar with the specific venue, are better off than visitors from similiar climates, but unfamiliar with the specific city/stadium.

I did look at all cases since the merger where a team stayed in the same city it had been in for the previous 5 years, but moved into a new (but same climate) venue, to see what effect it had on home field against divisional opponents from a similiar climate. This would include: Buffalo (1973), NY Giants (1976), NY Jets (1984), Washington (1997), Cincinnati (2000), Pittsburgh (2001), New England (2002), Detroit (2002) and Philadelphia (2003). The sample size is not large enough to say anything with statistical certainty, but it was interesting that the home field advantage against similiar divisional opponents spiked in year 2 following the move, and disappeared thereafter.

These teams went 12-9-1 at home against their "climate" rivals in year 1 after the move, and an incredible 18-4 at home in year 2. They won about the same number at home and on road in the years leading up to the move, and reverted to this pattern in years 3-6 following the move. So, there may be something to the familiarity thing. I may also try to look at conference rivals who played at the same venue in more than 2 consecutive seasons, division rivals after relocation, and new divisional rivals (such as with realignment in 2002).

Finally, here are the point spread records since 2002, of the home team in divisional games, sorted by temperature difference.

Within 5 degrees on average: 59-75-2 (.441)
5.1 to 10 degrees on average: 56-76-4 (.429)
10.1 to 15 degrees on average: 48-48-4 (.500)
15.1 to 20 degrees on average: 25-15-0 (.625)
20.1 degrees or more on average: 30-30-0 (.500)

Friday, April 13, 2007


Some random NFL kicker items from the 2006 season:

First, did the NFL juice up the ball last year for kickers? I looked at the touchback stats for all kickers who kicked off at least 50 times in both 2005 and 2006. Here are the results:

2005: 179 touchbacks out of 1,847 kickoffs (9.7%)
2006: 250 touchbacks out of 1,824 kickoffs (13.7%)

Of the 25 kickers qualifying under my criteria, 13 of them showed significant improvement in the touchback rate, 10 were roughly the same, and 2 declined (Rackers and Janikowski). This also does not take into account other veteran kickers who kicked off last year after not kicking off the year before, who also showed a significant increase in touchbacks over their most recent kickoff performances of prior seasons.

In line with that, kickers are aging like fine wine. If they were baseball players, they would probably be getting accused of performance enhancement. I don't know what all the reasons are for kickers in the NFL, and I am sure there are multiple factors. Here are the numbers for the 11 kickers who were age 33 or older at the start of last season (Andersen, Carney, Stover, Elam, Kasay, Vanderjagt, Wilkins, Nedney, Mare, and Vinatieri).

40-49 yard field goals
2006: 76 of 96 (79.2%)
rest of career: 874 of 1270 (68.8%)

50+ yard field goals
2006: 16 of 30 (53.3%)
rest of career: 226 of 430 (52.6%)

Kickoffs (excluding Elam and Andersen)
2006: 82 of 601 (13.6%)
rest of career: 919 of 6689 (13.7%)

Monday, April 2, 2007


How much does the change in time zone effect the home field advantage?

To attempt to come up with an approximate answer to this question, I used the home team wins and losses in all games since 1986, sorted by climate and distance. I eliminated games involving dome teams (for now) because I wanted to isolate the effect without the noise coming from the advantage of outdoor teams vs. dome teams. I also (for now) eliminated games involving Denver.

Denver clearly has a strong home field advantage. Denver plays all of its games against opponents from different time zones, and some of the advantage is likely due to this. But Denver also plays opponents from different elevations and weather patterns as well. So, for now, we will look at how time zone changes affect other teams, to get a sense of how much of Denver's advantage is due to this effect.

Here is the data for games between outdoor teams from the same time zone versus one time zone difference, versus two or more time zone differences.

  • Same Time Zone -- 759-603-2 (.557)
  • 1 Time Zone Diff -- 329-219-3 (.600)
  • 2+ Time Zone Diff-- 451-317-1 (.587)

The home field advantage is weaker when two outdoor teams are from the same time zone. However, while the advantage increases with a change in time zones, it does not appear to continue to increase with additional time zone changes. When we weight the time zone changes by cross-referencing against climate changes, the effect of playing an opponent from a different time zone versus one from the same time zone is roughly an increase in expected winning percentage of +0.035, but the effect of additional changes in time zone is negligible.

In games involving Denver and another outdoor team, the home team is 167-87 (.657) since 1986. In all other outdoor vs outdoor games since 1986, the home team is 1539-1139-6 (.575). Thus, Denver has had an advantage of +.082 above a normal outdoor vs. outdoor situation. Roughly +0.047 of this, then, is due to either random chance due to the smaller sample size, or to legitimate factors other than time zone changes, such as elevation changes and environmental changes between Denver and the opponent.

This research also supports that the effect of cross-country travel is overstated. Changing from the Eastern to the Pacific Time Zone has been no less a disadvantage than changing from the Eastern to Central Time Zones.

Conversely, the effect that coming from a similar environment and the same time zone has on reducing home field advantage is understated. In a future post, I will look at this further by looking at "against the spread" data.

Friday, March 23, 2007

Teams Drafting Running Backs or Wide Receivers Early in the Draft

This post follows these two: and

Following the same format, here is the data for the Running Back and Receiver Teams.

Year N-1: 44 . . . 0.325 . . . 5.2 . . . 1 . . . 0
Year N : 44 . . . 0.477 . . . 7.6 . . . 19. . . 6
Year N+1: 43 . . . 0.490 . . . 7.9 . . . 14 . . . 5
Year N+2: 40 . . . 0.469 . . . 7.5 . . . 14 . . . 5
Year N+3: 40 . . . 0.491 . . . 7.9 . . . 16 . . . 2
Year N+4: 40 . . . 0.439 . . . 7.0 . . . 9 . . . 1


Year N-1: 37 . . . 0.382 . . . 6.1 . . . 4 . . . 2
Year N : 37 . . . 0.488 . . . 7.8 . . . 11. . . 5
Year N+1: 37 . . . 0.481 . . . 7.7 . . . 13 . . . 2
Year N+2: 33 . . . 0.482 . . . 7.7 . . . 12 . . . 5
Year N+3: 30 . . . 0.473 . . . 7.6 . . . 8 . . . 1
Year N+4: 28 . . . 0.550 . . . 8.8 . . . 10. . . 4

Friday, March 9, 2007

Teams drafting Quarterbacks or Offensive Linemen Early in the Draft

This is a continuation of this post, To start, here are the results of looking at teams drafting QB's and OL's in the first 12 picks of the NFL draft, since 1978.

Year N-1 represents the year before the player was drafted. We would expect the records this year to be poor, because the team would not be drafting in the top 12, barring a trade. Year N represents the rookie year immediately following the draft. Year N+1, N+2, N+3 and N+4 are the next 4 seasons following the rookie year of the drafted position in question.

The numbers presented are as follows for each year: the first number is the number of qualifying teams who drafted a player in the top 12 at that position; the second number is the team combined winning percentage for that year; the third number is the average number of regular season wins for the teams in question, assuming a 16 game schedule; the fourth number is the number of subject teams who qualified for the playoffs in the year in question; and the fifth number is the number of subject teams who advanced to the conference championship game or Super Bowl.

So, if you look at the Teams drafting QB's in Year N, that tells you that there were 41 teams who have drafted a QB in the top 12 since 1978, they combined for a 0.371 winning percentage in the first season following the draft (5.9 wins on average), and that 4 out of 41 made the playoffs, and 1 out of 41 advanced to at least the conference championship (Pittsburgh 2004).

One minor problem did present itself. There were 4 expansion teams who had no record the year before the player was drafted. However, to account for the general quality of these teams, I treated them as a "2-14 team" the year before the draft in question, which I think is a good approximation of the talent level on the expansion roster at the time the position in question was drafted.

Here are the numbers:

YEAR N-1: 40 . . . 0.286 . . . 4.6 . . . 3 . . . 3
Year N: 40 . . . 0.379 . . . 6.1 . . . 5 . . . 1
Year N+1: 37 . . . 0.470 . . . 7.5 . . . 11 . . . 4
Year N+2: 36 . . . 0.486 . . . 7.8 . . . 12 . . . 2
Year N+3: 33 . . . 0.485 . . . 7.8 . . . 12 . . . 7
Year N+4: 31 . . . 0.474 . . . 7.6 . . . 8 . . . 4

YEAR N-1: 49 . . . 0.362 . . . 5.8 . . . 2 . . . 0
Year N: 49 . . . 0.402 . . . 6.4 . . . 4 . . . 1
Year N+1: 48 . . . 0.404 . . . 6.5 . . . 7 . . . 3
Year N+2: 47 . . . 0.455 . . . 7.3 . . . 13 . . . 6
Year N+3: 47 . . . 0.472 . . . 7.6 . . . 17 . . . 3
Year N+4: 46 . . . 0.513 . . . 8.2 . . . 16 . . . 6

For baseline comparison, the chances of a randomly selected team making the playoffs over this entire time period is about 39%, and the chances of a team making the conference championship is about 14%.

Teams drafting QB's were worse on average the year before the draft in question. The OL teams won slightly more games in the rookie season, but the QB teams showed more improvement. The QB teams outperformed the OL teams in years 2-4 in terms of winning percentage, percentage of teams making playoffs, and percentage of teams advancing to the conference championship. The OL teams were better in terms of winning percentage in year 5, and slightly better or roughly equal in terms of playoff teams and championship teams that year.

The conventional wisdom is that teams that draft QB's are in for a rebuilding project. This appears true for the first year, but after that, it is actually the teams drafting OL that have taken longer to build to respectability and competing for championships.

Next up will be the RB and WR teams.

Friday, March 2, 2007

Building a Team the Right Way Early in the Draft

I am going to predict that, leading up to the NFL draft, you are going to hear many draftniks and fans say that (insert bad team name from last year here) should draft OT Joe Thomas and not take a QB or RB early in the draft, because you build the team the right way by going for offensive line, and some variation on QB's being more risky, and RB's being fungible.

Off and on over the next month and half leading up to the draft, I am going to try to look at the history of the draft to see whether this actually holds water. My guess, based on a quick review of the early first round draft picks over the last 25-30 years, is that it does not. In my opinion, a big part of the rationale for this is the same situation noted by Doug Drinen in his post "Reputation and Information", re-posted here Except, it is the opposite side of the coin, since we don't have many meaningful stats on offensive linemen, the bad ones may not be perceived as bad as other positions where we have more statistics to measure.

Who is the bigger bust, Ryan Leaf or Robert Gallery? No-brainer, you say? I was at the KC-SD game at Arrowhead in 1998 when Leaf went something like 1 for 15. He was definitely a disaster. But living in an AFC West market, I have seen more than enough of Gallery as well. His performance in the season opener against San Diego was every bit as much a disaster, as he served as an express highway to the quarterback all game long. Could the Raiders horrible performance over the last few years, and inability to keep a qb standing for more than 2 seconds this year, have anything to do with drafting Gallery with their high pick a few years back?

Well, those are but two isolated examples. I am going to try to incorporate all the evidence, great pick, so-so, or bust. How to do it is the problem. As Drinen's article points out, it is not going to be possible to compare positions by raw statistics, because some QB/RB have lots of data, while others have little direct data. And I don't think Games Played or even Games Started is a viable option because of the differences in position, and how they are utilized. If a quarterback and offensive lineman are equally bad, the lineman will probably start more games. Take Mike Williams at Buffalo for example. If the Tackle is bad, he can move into the guard spot to see if he works there. He would have to be worse than roughly 5 other players to get benched. He may still get benched, but out of necessity and depth, he is probably going to start more games than a similar QB.

So, that leaves us with something that I think is the best we have at this point: wins and losses, and how the team performed in the years after a pick of a certain position was made. It is by no means perfect. One player alone does not determine a team's future. Nor is a team's success directly tied to the quality of one pick. If we have enough data points, some of that stuff will even out. Acknowledging it is not perfect, I am going to try to look at each position by the following method:

Look at all top 12 picks in the NFL draft since 1978 (the year of passing liberalization and expansion to 16 games), look at the record the year before the pick, and gauge the team's success over the next 5 seasons, by win-losses, percentage of team's making playoffs, and percentage advancing to at least a conference championship game.

I will not have the time to look at each individual's career with the drafting team. So, if the Packers start getting good at the end of the 5 year's after Tony Mandarich is drafted, even though he is out of the league, so be it. I guess I am not looking at how good the player drafted was directly, but rather, what the effect of drafting a certain position has on the franchise over the next five years.

Anyway, I will try to start with QB, OL, and RB, and move on to the others.

Lineup Placement

In the Royals first spring training game yesterday, former first overall pick and top prospect Alex Gordon was in the lineup in the 7th spot. The commentators on sports radio whb 810 stated that this is probably where he was going to start the season. I don't know if they were basing it on inside information from the manager, or if it was speculation. This does, however, pass for conventional wisdom. Start the rookie out lower in the lineup so that there is no pressure on him.

I'm just not sure if it is a valid point. Presumably, if Alex Gordon is routinely hitting in the 7th spot, it is because the "when healthy" regular lineup will be 1) CF DeJesus; 2) 2B Grudzielanek; 3) RF Teahen; 4) DH Sweeney; 5) LF Emil Brown; and 6) 1B Ryan Shealy. Now, I have no problem with the first 4 in the batting order, with the big assumption that Sweeney is healthy. But if Gordon is as good as all evidence is indicating, based on his minor league numbers, college career, draft status, and ranking in publications like Baseball America, then he is a better hitter than both Emil Brown and Ryan Shealy . . . right now.

The Royals are clearly not the Yankees, and do not have a lineup full of proven stars. If he is the better hitter, then I think he should be in the 5 spot from the start. It will be moot in a few months when he moves that direction, but I do not see the benefit of starting him off in between the two aforementioned hitters, and the strikeout machines that are John Buck and Angel Berroa.

Tuesday, February 20, 2007

A Yard is a Yard

A few weeks ago, Doug Drinen posted a study on his blog starting here- that looked at evidence of irrational coaching behavior, by looking at the difference between run/pass distribution on 2nd and 10, depending on whether the previous unsuccessful play was a run or a pass.

That post got me thinking about other potential examples of irrational coaching behavior, and the first circumstance that came to mind were 4th and 1 situations after a team was "stuffed" on the previous 3rd down play and failed to get the first down. My impression is that coaches get a little more conservative in this situation than, say, if they get stuffed on first down and then get to 4th and 1 after a pass. I think there is probably some mentality of not wanting to fail back to back that keeps coaches from going for it in this situation as often as they should.

Before I get to the numbers, some other thoughts. A run that would be classified as a 1-yard run if it occurred on 2nd and 6 would be classified as a 0-yard gain if it occurred on 3rd and 1, even if it gained a full yard or close to it, if it failed to make it past the first down marker. Thus, I would expect that 4th and 1's resulting from 0-yard rushing gains on 3rd down would be, on average, closer to the first down marker than other 4th and 1's. How much closer I cannot say, but it seems intuitive to me that some percentage of these 0-yard gains actually moved the line of scrimmage closer to a first down.

On the other hand, I would expect 4th and 1's resulting from officially recorded positive gains to be more randomly distributed, ranging from 4th and inches to almost 4th and 1 and 1/2 yards. Thus, if all else was equal, and there was no bias against going for it after being stopped with little to no gain on the previous play (and if the relative distance of a 4th and 1 was a determining factor), I would expect that the attempt rate and success rate on 4th and 1 following failure on 3rd and short to be, at least to some degree, greater than following 3rd and medium or long plays.

Now to the actual numbers from the 2006 seasons, with a couple of caveats. I looked at all 4th and 1 opportunities between the opponent's 40 yard line and the opponent's goal line, with the thought that this was the area where teams are actually willing to go for it. I excluded plays occurring in the last minute of each half. I counted it as an opportunity even if a team took a delay of game penalty, with the thought that this was a conscious decision "not to go for it" by the team in question. I did not count it if there was another pre-snap penalty (false start, e.g.). If a team attempted to go for it, but was called for an accepted offensive penalty during the play, I counted it as an attempt, but did not count it against the success rate. The thinking here is that it is neither a success nor a failure in the same sense that conversion vs. failed attempt is, because the team could then punt or attempt a field goal. However, it was an "attempt" because the intent of the coach was clearly to go for it.

Here are the numbers, broken down by field position. The first number is the total number of opportunities, the second number is the attempt percentage (attempts/opportunities) and the third number is the success percentage (conversions/total attempts (minus above caveat)).

4th and goal from the 1

  • 1 yard or less gained on 3rd down 23/ 0.696 /0.500
  • 2 to 4 yards gained on 3rd down 8 /0.625 /0.800
  • 5 or more yards gained on 3rd down 6 /0.500 /1.000

4th and 1, 10 yard line or in (excluding goal to go)

  • 1 yard or less gained on 3rd down 17 /0.412 /0.833
  • 2 to 4 yards gained on 3rd down 6 /0.500 /0.667
  • 5 or more yards gained on 3rd down 16 /0.438 /0.571

4th and 1, 11-20 yard line

  • 1 yard or less gained on 3rd down 16 /0.125 /0.500
  • 2 to 4 yards gained on 3rd down 9 /0.333 /0.667
  • 5 or more yards gained on 3rd down 10 /0.500 /0.600

4th and 1, 21-30 yard line

  • 1 yard or less gained on 3rd down 18 /0.500 /0.875
  • 2 to 4 yards gained on 3rd down 8 /0.625 /0.800
  • 5 or more yards gained on 3rd down 18 /0.611 /0.636

4th and 1, 31-40 yard line

  • 1 yard or less gained on 3rd down 24 /0.708 /0.824
  • 2 to 4 yards gained on 3rd down 11 /0.818 /0.556
  • 5 or more yards gained on 3rd down 18 /0.944 /0.688

4th and 1, all between opponent's 2 and 40 yard line

  • 1 yard or less gained on 3rd down 75 /0.467 /0.818
  • 2 to 4 yards gained on 3rd down 34 /0.588 /0.650
  • 5 or more yards gained on 3rd down 62 /0.645 /0.641

Now, I cherry-picked out the goal line situations, where the majority result from short to no gains on the previous down. Here, I think the rationale is different. In my subjective opinion, if there is something NFL coaches as a group love, it is to pin the opponent on the goal line. An overinflated sense of being able to pin the opponent deep must be what drives some coaches to punt on 4th and 1 from the opponent 36, for example. And so here, success is immediate gratification, and failure leaves the other team snapping in their own end zone.

As for the remainder of 4th and 1 situations, the coaches for teams that gained little yardage on 3rd down are going for it less frequently, but being far more successful when they do, suggesting that the 4th and 1 attempts may be closer on average. More data is necessary, but it appears that in 2006, coaches who were stuffed on 3rd down were more conservative on fourth down than their peers.

FYI, here are the teams with the most times facing 4th and 1 following a no gain play on 3rd and 1:

Green Bay - 5

Dallas - 4

Detroit - 4

Oakland - 4

Washington - 4

Monday, February 12, 2007

Does Saunders create bad defense?

This was posted on the KC Chiefs website by Bob Gretz on January 19, 2007, in an article about defensive rankings:

"WASHINGTON’S DEMISE: Could it be that what the Chiefs proved for the better
part of five years and what the Redskins showed in 2006 is that it’s
to have a good defense when Al Saunders is your team’s offensive

There are plenty of people in the NFL, and around Arrowhead
who don’t think it’s a coincidence that when Saunders took over
the offensive game plans for the Redskins, their defense turned in a
bad season. . . . "

I hope he is joking. Al Saunders, offensive coordinator of one of the best offenses in the league, was apparently responsible for the Chiefs horrible defense. It had nothing to do with the GM spending early picks on Ryan Sims, or Eddie Freeman, or William Bartee. It had nothing to do with playing Eric Warfield and Dexter McCleon. Or having no edge pass rushers like the current tandem of Jared Allen and Tamba Hali during 2001-2003. Oh, and Saunders apparently made Adam Archuleta suddenly become terrible, depleted the Redskins depth due to continuous trading of draft picks in seasons he was not there, and injured their starters through his poor game planning.

Amazing. I would try to rebut this with actual statistics, but what is the point. Clearly, Gretz is the master of correlation versus causation.

Saturday, February 10, 2007

Home Field Advantage and Passing

This is a follow up to the previous post, in which I looked at home and road won-loss records of all teams finishing between 6-10 and 10-6 for the years 1995-2006.

Here are the pass offense stats:

  • below 6.0 - 18 teams; .576 at home; .299 on road
  • 6.1-6.3 ypa - 34 teams; .563 at home; .419 on road
  • 6.4-6.6 ypa - 43 teams; .583 at home; .407 on road
  • 6.7 -6.9 ypa- 35 teams; .577 at home; .423 on road
  • 7.0-7.2 ypa - 29 teams; .591 at home; .425 on road
  • 7.3-7.5 ypa - 25 teams; .605 at home; .455 on road
  • 7.6-7.8 ypa - 14 teams; .580 at home; .455 on road
  • 7.9 and up - 12 teams; .667 at home; .469 on road

And here are the pass defense stats

  • below 6.0 ypa - 22 teams; .591 at home; .438 on road
  • 6.1-6.3 ypa - 35 teams; .636 at home; .443 on road
  • 6.4-6.6 ypa - 46 teams; .598 at home; .428 on road
  • 6.7 -6.9 ypa- 48 teams; .560 at home; .422 on road
  • 7.0-7.2 ypa - 34 teams; .581 at home; .414 on road
  • 7.3-7.5 ypa - 14 teams; .589 at home; .321 on road
  • 7.6-7.8 ypa - 4 teams; .563 at home; .406 on road
  • 7.9 and up - 7 teams; .464 at home; .339 on road

Teams improve their overall record as pass defense improves, but the difference between home and road performance stays fairly constant. The same is generally true of the pass offense as well. However, the 18 teams that were "poor" offensive passing teams (ypa 6.0 or lower) were horrible on the road while being decent at home.

As a result, I looked at teams who were below average in passing, while also being good at stopping the run. Here are the teams from the bottom passing offense groups of 6.6 ypa and below, sorted by their rushing defense numbers:

  • 3.5 or below - 13 teams; .615 at home; .365 on road
  • 3.6 - 3.8 ypc -30 teams; .594 at home; .350 on road
  • 3.9 - 4.1 ypc -24 teams; .583 at home; .443 on road
  • 4.2 - 4.4 ypc -18 teams; .528 at home; .396 on road
  • 4.5 and up - 9 teams; .542 at home; .417 on road

To summarize, the top 2 groups on this list are the below average passing offense, but above average run defense teams. Those 43 teams had a +.246 difference in their home winning percentage versus the road winning percentage, which equates to almost 2 more wins at home on average over the course of a season.

It appears that these types of teams (bad offense, good rush defense) have a greater road disadvantage because of the offense, but a greater home advantage because of the rush defense, creating a greater home/road differential.

Home Field Advantage and Rushing

Using the database, I looked at all teams that finished between 6-10 and 10-6 for the years 1995-2006. Why those records? It's partially arbitrary-- I could have just as easily narrowed it to 9-7/7-9, or expanded it to 11-5/5-11. But my goal was to look at the middle class of the NFL, teams that generally have some strengths but also some flaws. I felt this dividing line would accomplish that.

I then compared some of the standard rate stats, such as rush yards per attempt and pass yards per attempt, to see what team characteristics, if any, contributed to a greater difference between home and road winning percentages. This post will focus on the rushing stats, both offensive and defensive.

210 total teams finished between 6-10 and 10-6 during the 12 seasons reviewed, an average of over 17 per season--so slightly more than half the teams in the league on average. The entire group averaged winning .586 at home and .417 on the road, for a +.169 difference between home and road winning percentages. This would equate to +1.36 more home wins than road wins over the course of a 16 game schedule.

Here are the numbers divided by offensive rush yards per carry:
  • 3.5 and below - 32 teams; .582 at home; .398 away
    3.6 to 3.8 ypc - 34 teams; .590 at home; .408 away
    3.9 to 4.1 ypc - 63 teams; .567 at home; .423 away
    4.2 to 4.4 ypc - 40 teams; .591 at home; .417 away
    4.5 and above - 41 teams; .611 at home; .431 away

Here are the numbers divided by defensive rush yards allowed per carry:

  • 3.5 and below - 26 teams; .611 at home; .399 away
  • 3.6 to 3.8 ypc - 47 teams; .625 at home; .380 away
  • 3.9 to 4.1 ypc - 57 teams; .605 at home; .425 away
  • 4.2 to 4.4 ypc - 41 teams; .537 at home; .419 away
  • 4.5 and above - 39 teams; .548 at home; .460 away

There is basically no indication that simply running the ball better on offense, without any other info, either increases home field advantage or reduces road disadvantage.

Unlike the offensive numbers, the defensive numbers appear to show a difference between how the "middle class" performs at home vs. on the road, depending on whether the team is good at run defense or not. Teams that are not particularly good at stopping the run show less difference in how they perform at home vs. on the road.

If the strength of rush defense does increase home field advantage, there is a potential explanation. If a team is better at stopping the run, it is conceivable that such a team would be somewhat more likely to place its opponent into more 3rd and long situations. This might translate to a bigger advantage at home, where the offense is subject to crowd noise, than on the road, where the home crowd would presumably be quiet to aid the offense. Of course, we also need to keep in mind that this is looking at some of the variables. A team that is poor at stopping the run, yet finishes in the middle, likely has strengths in other areas to compensate. A team that is good at stopping the run, but does not finish in the upper tier, also likely has some other flaws that may be combining to create road disadvantage as well.

Next time, I will play with the passing numbers, overall offensive numbers and overall defensive numbers.