Trends in Saves: Length and Distribution

In 2011, the average Save lasted 1.006 IP, an all-time low. This figure has been falling since the mid-1980s:

Year	IP/Save
1961	1.54
1971	1.63
1980	1.71
1982	1.78 (high)
1984	1.70
1986	1.60
1988	1.47
1990	1.43
1992	1.25
1994	1.19
1996	1.16
1998	1.15
2000	1.15
2002	1.09
2004	1.09
2006	1.05
2008	1.03
2010	1.013
2011	1.006 (low)

The number of Long Saves (more than 1 IP) plunged to a record low of 82, or 2.7 per team. In 1984, teams averaged 23 Long Saves. The 2011 figure is a 30% drop from just the year before (3.9 per team), and a 62% fall from 2004 (7.1 per team).

Saves of exactly 1 IP accounted for a record 85% of all Saves. That figure was 65% in 2000, 50% in 1992, and 25% in 1987.

For the first time ever, Long Saves were outnumbered by Short Saves (less than 1 IP). This is entirely due to the drop in Long Saves. Short Saves are actually less common now than they were 10+ years ago, further reflecting the hegemony of the "closer always starts the 9th if there's a Save chance" philosophy.

Finally, Saves continue to be ever more concentrated among designated closers. Here is the percent of all MLB saves that were accounted for by the top T in Saves (where T = the number of MLB teams that year):

Year	% Concentration
1961	49.9%
1966	50.5%
1971	56.2%
1976	53.9%
1980	57.2%
1986	60.0%
1991	62.8%
1996	74.6%
2001	76.2%
2006	76.5%
2011	78.6%