SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all PFR content, so you can quickly and easily find the content you want.

Also, our existing PFR blog rss feed will be redirected to the new site's feed.

Pro-Football-Reference.com » Sports Reference

For more from Chase and Jason, check out their work at Football Perspective and The Big Lead.

Parity

Posted by Doug on August 15, 2006

In the comments to Friday's post, some discussion about parity broke out, so I thought I would run some quick numbers on parity through the years.

It seems to me that the word has at least two distinct meanings. First, it can mean that in a given season there are few dominant teams and that almost any team can beat any other on a given week. But the word also seems to have acquired some season-to-season connotations. That is, even if there are plenty of 3-13 teams and 14-2 teams, the teams that are good one year could easily be poor the next year, and vice versa. The 1998 Super Bowl Falcons, who were 7-9 the year before and 5-11 the year after, provide an example of this kind of parity.

For today let's focus on the first kind of parity. Are teams bunching up around .500? Last season, only five of 32 teams (15.6%) had records within one game of .500. That is the lowest percentage since 1975. The high water mark was 1983, when 16 of the league's 28 teams were within a game of .500. Here is the corresponding percentage for each year since the merger


% of teams within
YR a game of .500
===========================
1970 26.9
1971 30.8
1972 26.9
1973 19.2
1974 38.5
1975 15.4
1976 17.9
1977 25.0
1978 46.4
1979 35.7
1980 28.6
1981 42.9
1982 42.9
1983 57.1
1984 39.3
1985 32.1
1986 21.4
1987 32.1
1988 28.6
1989 39.3
1990 25.0
1991 25.0
1992 21.4
1993 42.9
1994 46.4
1995 50.0
1996 43.3
1997 33.3
1998 30.0
1999 41.9
2000 29.0
2001 32.3
2002 43.8
2003 18.8
2004 34.4
2005 15.6

I am not sure exactly when the NFL's reputation for excessive parity was born, but there are two periods that stand out: 1981--1984 and 1993--1996.

Here is another quick look at within-season any-given-Sunday type parity: what percentage of games are upsets, based on at-the-time records? If the home team has a better record and loses, I'll call that an upset. If the road team has a record which is at least two games better than the home team, and the road team loses, I'll call that an upset. To give you a feel for the kinds of games I'm talking about, here were last season's upsets:


Week 17: nwe (10- 5) loses to mia ( 8- 7) at home
Week 10: buf ( 3- 5) beats kan ( 5- 3) on the road
Week 10: nyg ( 6- 2) loses to min ( 3- 5) at home
Week 14: pit ( 7- 5) beats chi ( 9- 3) on the road
Week 12: oak ( 4- 6) loses to mia ( 3- 7) at home
Week 17: dal ( 9- 6) loses to stl ( 5-10) at home
Week 15: ind (13- 0) loses to sdg ( 8- 5) at home
Week 13: kan ( 7- 4) beats den ( 9- 2) on the road
Week 10: atl ( 6- 2) loses to gnb ( 1- 7) at home
Week 17: kan ( 9- 6) beats cin (11- 4) on the road
Week 14: nyj ( 2-10) beats oak ( 4- 8 ) on the road
Week 16: ari ( 4-10) beats phi ( 6- 8 ) on the road
Week 16: car (10- 4) loses to dal ( 8- 6) at home
Week 14: sdg ( 8- 4) loses to mia ( 5- 7) at home
Week 6: pit ( 3- 1) loses to jax ( 3- 2) at home
Week 16: bal ( 5- 9) beats min ( 8- 6) on the road
Week 11: stl ( 4- 5) loses to ari ( 2- 7) at home
Week 14: gnb ( 2-10) beats det ( 4- 8 ) on the road
Week 16: cin (11- 3) loses to buf ( 4-10) at home
Week 16: was ( 8- 6) beats nyg (10- 4) on the road
Week 8: hou ( 0- 6) beats cle ( 2- 4) on the road
Week 11: bal ( 2- 7) beats pit ( 7- 2) on the road
Week 17: gnb ( 3-12) beats sea (13- 2) on the road
Week 11: was ( 5- 4) loses to oak ( 3- 6) at home
Week 14: car ( 9- 3) loses to tam ( 8- 4) at home
Week 15: hou ( 1-12) beats ari ( 4- 9) on the road
Week 8: sfo ( 1- 5) beats tam ( 5- 1) on the road
Week 17: min ( 8- 7) beats chi (11- 4) on the road
Week 17: nyj ( 3-12) beats buf ( 5-10) on the road
Week 16: stl ( 5- 9) loses to sfo ( 2-12) at home
Week 7: cin ( 5- 1) loses to pit ( 3- 2) at home

I only counted games after week 6, and I probably ought to throw out week 17 but I didn't. Anyway, that's 31 upsets, and there were 128 games where an upset could have occurred, so that's an upset percentage of 24.2%. Again, that figure is nearly the lowest in post-merger NFL history:


Year Upset %
================
1970 28.9
1971 34.6
1972 23.2
1973 21.2
1974 37.6
1975 25.3
1976 21.3
1977 31.5
1978 38.0
1979 36.8
1980 37.7
1981 33.3
1982 34.3
1983 37.9
1984 30.0
1985 31.7
1986 30.3
1987 39.6
1988 37.6
1989 37.4
1990 33.1
1991 28.8
1992 30.2
1993 45.9
1994 35.1
1995 41.7
1996 39.2
1997 35.0
1998 29.4
1999 37.2
2000 35.1
2001 34.1
2002 36.1
2003 31.4
2004 33.6
2005 24.2

I have a few other thoughts for measuring within-season parity, but this post is long enough. I'll post more on this tomorrow, and then move on to between-season parity.

This entry was posted on Tuesday, August 15th, 2006 at 4:11 am and is filed under History. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.