The special teams guys have gotten just a little bit of love this summer, while we are in the dog days and waiting for pre-season to begin. Chase got it started last month with some discussion of the best return games, and a two-part follow up on Josh Cribbs and his 2007 season. Doug followed that up by introducing some new field goal kicker data, and asked whether Neil Rackers' 2005 was better than Cribbs' 2007 season. So, I figured the punters needed a little print as well.
Below is a list of Pro Bowl Punters for each year since the merger, going in reverse chronological order. I've listed the team with the PFR abbreviations for each franchise, and for those that have relocated at various times, I also list which location the team was playing if applicable (this will come up in the discussion later). For each pro bowl punter, I also list two other relevant pieces of information for the purpose of my discussion. First, I list the team gross punting average (though I figure most, if not all of the punts were made by the individual in question--it was just easier to compile this way), with the league rank in punting average for that season in parentheses. Next, I list the team rank in a stat that Doug developed for use in his approximate value work, and which was first discussed here, called OPPED (Offensive Points Per Estimated Drive). OPPED gives us a decent estimate of the caliber of offense that the particular pro bowl punter was playing with, ranging from Reggie Roby with the 1984 Miami Dolphins who went 14-2 (#1 in OPPED) to Todd Sauerbrun with the 1-15 2001 Carolina Panthers (#31 in OPPED) and all points in between.
Year Name Team Gross Avg. OPPED RK ============================================================= 2007 Lechler rai-oak 49.1 (1) 23 2007 Lee sfo 47.3 (2) 32 2006 Moorman buf 43.6 (12-t) 25 2006 McBriar dal 48.2 (1) 3 2005 Moorman buf 45.7 (1-t) 19 2005 Bidwell tam 45.6 (3) 20 2004 Lechler rai-oak 46.7 (1) 16 2004 Berger nor 43.6 (4) 17 2003 Hentrich oti-ten 43.9 (3) 6 2003 Sauerbrun car 42.8 (7) 18 2002 Hanson jac 44.2 (2) 18 2002 Sauerbrun car 45.1 (1) 30 2001 Lechler rai-oak 45.6 (2) 3 2001 Sauerbrun car 47.0 (1) 31 2000 Bennett sdg 46.2 (1) 28 2000 Player crd-ari 44.2 (8) 26 1999 Tupa nyj 45.0 (4) 13 1999 Berger min 45.4 (2) 5 1998 Hentrich oti-ten 47.2 (1) 10 1998 Turk was 43.5 (13-t) 18 1997 Barker jac 44.9 (5) 2 1997 Turk was 44.6 (6) 19 1996 Gardocki clt-ind 45.7 (2) 14 1996 Turk was 45.1 (4) 6 1995 Bennett sdg 44.7 (2) 18 1995 Feagles crd-ari 43.8 (4-t) 29 1994 Tuten sea 42.9 (7-t) 23 1994 Roby was 44.4 (2) 25 1993 Montgomery oti-hou 45.3 (1) 5 1993 Camarillo crd-ari 43.7 (7) 9 1992 Stark clt-ind 44.8 (2) 26 1992 Camarillo crd-ari 42.8 (10) 22 1991 Gossett rai-la 44.2 (4) 13 1991 Camarillo crd-ari 44.7 (3) 27 1990 Stark clt-ind 42.8 (5-t) 20 1990 Landeta nyg 44.1 (2) 8 1989 Roby mia 41.7 (7) 10 1989 Camarillo crd-ari 43.6 (1) 26 1988 Horan den 43.8 (1) 13 1988 Arnold det 42.4 (5) 28 1987 Mojsiejenko sdg 42.0 (1) 27 1987 Arnold det 41.8 (2) 21 1986 Stark clt-ind 44.7 (2) 27 1986 Landeta nyg 44.8 (1) 6 1985 Stark clt-ind 44.7 (1) 11 1985 Hatcher ram-la 42.6 (6) 23 1984 Roby mia 44.7 (2) 1 1984 Hansen nor 43.1 (4) 18 1983 Camarillo nwe 44.6 (2) 20 1983 Birdsong crd-stl 41.5 (9-t) 15 1982 Prestridge den 45.0 (1) 25 1982 Jennings nyg 42.8 (6) 19 1981 McInally cin 44.8 (1) 2 1981 Skladany det 43.5 (2) 3 1980 Guy rai-oak 43.6 (3) 16 1980 Jennings nyg 44.8 (1) 26 1979 Grupp kan 43.1 (1) 26 1979 Jennings nyg 42.7 (2) 27 1978 Guy rai-oak 41.7 (4) 13 1978 Jennings nyg 42.1 (2) 21 1977 Guy rai-oak 43.3 (1) 2 1977 James atl 41.2 (3) 25 1976 Guy rai-oak 41.6 (3) 2 1976 James atl 42.1 (2) 26 1975 Guy rai-oak 43.8 (1) 6 1975 James atl 41.5 (5) 21 1974 Guy rai-oak 42.2 (1) 2 1974 Wittum sfo 40.8 (5) 20 1973 Guy rai-oak 45.3 (2) 12 1973 Wittum sfo 43.7 (4) 17 1972 Wilson kan 44.8 (1) 17 1972 Chapple ram-la 44.2 (2) 11 1971 Wilson kan 44.8 (1) 9 1971 Widby dal 41.6 (8) 1 1970 Wilson kan 44.9 (2) 15 1970 Green chi 40.8 (14) 19 =============================================================
Another piece of information that I didn't list is the record of these teams that had Pro Bowl punters. The overall winning percentage for the 76 Pro Bowl punter teams since 1970 is only 0.490 (569-592-7). Take out Ray Guy's seven pro bowl seasons with the Oakland Raiders between 1973 and 1980, and the winning percentage for the remaining 69 seasons plummets to 0.464.
How is it that the best players at a position can collectively contribute so much that their teams have a losing record over a large span of time? Here are some possibilities:
a) The contribution of punters is negligible in determining the overall quality of a team;
b) The difference between the top punters and an average punter is fairly small;
c) The Pro Bowl punters, as a collectively group, were unlucky enough to be on worse than average teams, and so these teams would have been even worse but for the outstanding contribution of the punter;
d) The Pro Bowl selections are not always based on the proper criteria, and selectors are not properly placing punting statistics in context.
Quick, I'll give you one guess as to which one I think is the biggest factor.
I don't think it is (a), though I'm certainly not going to try to convince you that a punter contributes as much as other positions like quarterback, left tackle, running back, or defensive end. Setting aside how we measure performance for a minute, don't you think that if a punter truly did his job better than others at his position, over the course of time and with a large enough sample size, those exceptional players would have contributed to winning football? (b) probably has some validity, compared to positions like quarterback where the difference between a star and a replacement level talent is pronounced and leads to a large difference in win expectation. I just threw in (c) but don't believe that to be the case here. No, I think the big problem is (d).
If you look at the list above, it is fairly apparent that gross punting average is highly correlated with pro bowl appearances for punters. Over 80% of the Pro Bowl punters ranked in the top 5 in gross punting average in the season they made the pro bowl. The problem with relying on gross punting average is that it is context-dependent, and lends itself to some of the same Simpson's paradox issues that Doug pointed out in the kicker post last month.
I'll use Darren Bennett from 2000 and Todd Sauerbrun from 2001 as one type of example. At the time, these two had the highest gross averages of any punter selected to the Pro Bowl, very impressive league leading averages of 46.2 and 47.0 respectively. Of course, both of these men played on teams that went 1-15, and had very poor offenses. Thus, these punters were punting in situations conducive to higher gross averages--namely punting from deep in their own end a lot, rather than having to sacrifice distance for accuracy by punting near or across midfield. I would expect that punters on bad offenses that punt a lot, with the majority of punts occurring inside their own 40, would have a higher gross average than punters on average offensive teams. Yet, it seems like this may not be fully accounted for in pro bowl selection.
While the Pro Bowl punters tend to play on below average offenses (59.2% played on teams below the average in OPPED), there is another subset of Pro Bowl Punters at the other extreme, those who played on elite offenses and post high gross averages. The potential problem here is sample size of punts, and thus the flukish effect of a few long punts on gross average. While as an overall group, we might expect punters on top offenses to have lower gross averages, because of the smaller sample size, some might jump up in average and appear on the leader board. I'll submit that Matt McBriar of the 2006 Dallas Cowboys wouldn't have averaged 48.2 yards per punt if he punted twenty more times that season, and it would have tended back toward the league median. That said, I'm a little less skeptical of Pro Bowl punters on good offenses, because of the decreased likelihood that they are building those high averages on punts from inside their own 20, where they can boom it without hesitation.
There is a larger issue that I think the selectors are failing to take into account, and it is also tied to the gross punting average as a driving factor. Weather. Six franchises have not had a single pro bowl punter since the merger--Green Bay, Cleveland, Pittsburgh, Philadelphia, Baltimore (Ravens), and Houston (Texans). Houston will be entering its seventh season. The others all have something in common-they play outdoors in cold weather. Throw in that Chicago hasn't had a Pro Bowl punter since the 1970 season, and Kansas City since 1972, and we might suspect that there is a little-recognized weather bias when it comes to Pro Bowl selections.
Using the same in-season average temperature data I had used for this guest post on home field advantage and climate, as well as the December wind chill data borrowed from the website Advanced NFL Stats for this post on Playoff Home Field Advantage, I divided the team/seasons into five categories.
THE DOME TEAMS (DOME) are pretty self-explanatory: they consist of the seasons that Minnesota, Detroit, New Orleans, Atlanta, Saint Louis (Rams), Seattle, Indianapolis, and Houston (Oilers) played in domes.
THE WARM WEATHER TEAMS (WARM) are all of those with an average football season (September through December) temperature over 70 and an average December wind chill temperature over 60: Los Angeles (Rams and Raiders), San Diego, Arizona, Houston (Texans), New Orleans (pre-dome), Tampa Bay, Jacksonville, Miami.
THE MODERATE WEATHER TEAMS (MOD) are all of those cities with an average football season temperature over 60, and an average December wind chill above 40: Dallas, San Fransisco, Oakland, Tennessee, Carolina, Atlanta (pre-dome), Washington, Baltimore
DENVER (DEN) consists of, well, Denver, since it is in its own time zone and elevation and is considered a cold weather venue.
SEATTLE (SEA) consists, of, you guessed it, Seattle playing outdoors since 2001. It doesn't feel right to include it with Denver because of the elevation differences. Seattle, along with Green Bay, is actually the coolest NFL venue for the first two months of the season, but the winter months are not as harsh as the cold weather teams in the midwest and east coast, so the weather patterns are distinct.
CENTRAL AND EAST COLD WEATHER TEAMS (COLD) consists of the remaining outdoor teams: New England, New York (Giants and Jets), Philadelphia, Kansas City, Saint Louis (Cardinals), Minnesota (pre-dome), Green Bay, Chicago, Detroit (pre-dome), Cleveland, Pittsburgh, Buffalo, Cincinnati.
So, I figured up the percentage of total seasons by a team in each category since 1970, to determine the average expected number of pro bowlers from each group of teams. The Central and East Cold Weather Teams, for example, comprise 27.3% of all seasons since 1970. The next chart shows the actual number of pro bowlers for each group, the expected number of pro bowlers, and the net difference.
Category Actual Expected Difference ======================================================== DOME 13 13.8 -0.8 WARM 17 13.0 +4.0 MOD 27 14.7 +12.3 DEN 2 2.6 -0.6 SEA 0 0.5 -0.5 COLD 17 31.3 -14.3 ========================================================
Denver's elevation factor is fairly well-known, and as it turns out, seems to be accounted for in Pro Bowl selection. Dome teams have, interestingly, produced roughly the expected number of pro bowlers. It is the outdoor warm and moderate teams that have benefited to the detriment of the cold weather teams. Even excluding Ray Guy, the moderate weather teams have dominated Pro Bowl selections more than we would expect if weather played no role.
The cold weather teams have produced almost half as many Pro Bowlers as would be expected based on the number of teams. This number would be even worse but for the contribution of one particular outdoor venue-the Meadowlands. Seven of the seventeen cold weather punters played for the Jets or Giants.
I'll close with some random Pro Bowl punter notes:
-- Only three punters have won a Super Bowl and played in a Pro Bowl in the same season: Ron Widby (Dallas, 1971), Ray Guy (Oakland, 1976 & 1980), and Sean Landeta (New York Giants, 1986 & 1990). Three others have played in a Pro Bowl the same season they lost a Super Bowl: Pat McInally (Cincinnati, 1981), Reggie Roby (Miami, 1984), and Todd Sauerbrun (Carolina, 2003).
--Pat McInally is one of the more interesting punters I came across. In 1980, the 6'6" McInally actually started 7 games at wide receiver, and had 18 receptions that year, including a 59 yard touchdown catch against the Browns. How many punters today could catch a 50 yard touchdown pass?
This entry was posted on Thursday, July 24th, 2008 at 6:02 am and is filed under General. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.