Do we have to worry about vultures?

The following guest post was written by Brendan Bingham

With Felix Hernandez having won the AL Cy Young Award and Bert Blyleven’s recent election to the Hall of Fame, it seems that wins and losses have fallen out of favor as the primary measures of a pitcher’s success.  However, we remain a long way from ignoring these stats altogether.

Vulture is a term applied to a relief pitcher who collects victories, often not so much through his own pitching prowess, but opportunistically, thanks to the timely late-inning hitting of teammates.  There is an air of deprecation in the use of this term.  In one scenario, the relief pitcher enters the game with his team losing.  This vulture is in a relatively invulnerable position.  If he pitches poorly, perhaps his ERA suffers, but he cannot be charged with the loss.  But if his team scores enough runs to take the lead, he stands to be credited with the win, sometimes after only a short and unspectacular time on the mound.  In a second – perhaps more sinister – scenario, the relief pitcher enters the game with the lead, but fails to hold it.  At least this vulture is at risk of taking the loss.  However, with his team’s rallying to retake the lead, he “earns” the victory, an unsavory accomplishment worthy of its being named for the large carrion-eating bird.

How common are vulture wins?

With the expansion and increasing specialization of relief pitching, we’re hearing more and more about vultures.  Modern day 25-player rosters are typically composed of no less than 11 pitchers, more often 12, sometimes 13.  In contrast, having 10 pitchers on the roster was the norm in the 60s and 70s.  The 4-man starting rotation has become a thing of the past.  Starters are pitching fewer innings and very few complete games.  Relief pitching has been broken down into three or four distinct sub-disciplines, and relievers as a group are pitching more innings.

The table below shows the number of wins collected by starters and relievers since 2000.  Please note that “% Starter Wins” is not a traditional winning percentage.  For the purpose of this analysis, losses and losing pitchers are ignored.  For each game played, we’re simply asking – was the winning pitcher a starter or a reliever

Year	Starter Wins	Relief Wins	% Starter Wins
2000	1680	748	69.19
2001	1716	712	70.68
2002	1703	722	70.23
2003	1730	699	71.22
2004	1657	771	68.25
2005	1741	689	71.65
2006	1723	706	70.93
2007	1682	749	69.19
2008	1682	746	69.28
2009	1706	724	70.21
2010	1736	694	71.44

These percentages are remarkable in their lack of variation.  With only minor year-to-year changes, the win in about 70% of MLB games in the past decade was credited to the winning team’s starting pitcher.

It is reasonable to expect that the frequency of vulture wins has increased over time, given the transformation that has taken place in the bullpen, or at least that was my expectation going into this analysis.  However, a quick survey of every tenth year from 1920 to 1990 shows evidence of change, but not where I had expected it.  Indeed, wins by relief pitchers were once less common, but you have to go back to 1950 to see a noteworthy departure from the 70/30 split of recent years.  Before looking at these numbers, I had expected the percentage of wins by starters to have been higher in 1960, if not also in 1970.  Bear in mind that I am not including all years from these earlier decades; rather I am trusting that one year in ten is representative of the decade, always a dangerous assumption.

Year	Starter Wins	Relief Wins	% Starter Wins
1920	1062	166	86.48
1930	990	242	80.36
1940	958	270	78.01
1950	965	265	78.46
1960	856	376	69.48
1970	1386	557	71.33
1980	1533	568	72.97
1990	1509	596	71.69

Traditionally, pitchers were expected to finish what they started.  Relief pitchers were often called upon only when the game was out of reach.  That’s an admittedly simple-minded view of relief pitching, but perhaps close enough to accurate for the 20s, 30s and 40s.  While it is only in the past two decades that our recognition of relievers’ contributions has included election to Cooperstown, it was during the 1950s that some well-known relievers emerged.  Two examples are Clem Lebine, who had double-digit relief wins for the ’55 and ’56 Brooklyn Dodgers, and Elroy Face, who performed the same feat for the ’59 and ’60 Pirates.

How important are vulture wins to team success?

We need look no further than the past season to see that vulture victories are sometimes key to team success, but more often not.

2010 National League
Team	Total Wins	Starter Wins		Relief Wins	% Starter Wins
PHI	97	70		27	72.16
SF	92	61		31	66.30
ATL	91	59		32	64.84
CIN	91	57		34	62.64
SD	90	66		24	73.33
STL	86	68		18	79.07
COL	83	58		25	69.88
FLA	80	63		17	78.75
LAD	80	55		25	68.75
NYM	79	53		26	67.09
MIL	77	52		25	67.53
HOU	76	52		24	68.42
CHC	75	60		15	80.00
WAS	69	42		27	60.87
ARI	65	49		16	75.38
PIT	57	34		23	59.65
2010 American League
Team	Total Wins	Starter Wins		Relief Wins	% Starter Wins
TB	96	73		23	76.04
NYY	95	72		23	75.79
MIN	94	73		21	77.66
TEX	90	58		32	64.44
BOS	89	70		19	78.65
CHW	88	64		24	72.73
TOR	85	63		22	74.12
DET	81	53		28	65.43
OAK	81	64		17	79.01
LAA	80	62		18	77.50
CLE	69	51		18	73.91
KC	67	46		21	68.66
BAL	66	42		24	63.64
SEA	61	46		15	75.41

In the NL, Cincinnati’s league-leading relief wins were important to their making the post-season.  Based on starting pitcher wins, St Louis would have won the Central division.  The Giants also beat out the Padres due in part to their posting more relief victories.  Meanwhile, the Phillies, who finished the season with the most total wins, had 27 wins by relief pitchers, a good but not outstanding total.  Many teams, including the division-trailing Washington Nationals, posted vulture win totals similar to Philadelphia’s.

In the AL, the Texas Rangers benefited from posting a league-leading 32 relief wins.  They would have finished third in their division based on starter wins alone.  However, the other three post-season teams put up unspectacular numbers of vulture wins.  Moreover, the Baltimore Orioles, with the second-worst record in the league, had more relief wins than the division-leading Tampa Bay Rays or the wildcard Yankees.

In general, the number of wins by starters is a better predictor of team success than wins by relief pitcher.  Poor teams and good teams often have similar numbers of relief pitcher wins.

Bottom Line

Although modern-day relief pitching has become very specialized, its primary purpose is to preserve starting pitchers’ wins, not to generate wins from the bullpen.  Closers are the most highly respected (and best paid) relievers, and the key stat by which they are judged is saves, not wins.  Wins from the bullpen are great, but they’re unpredictable, and when a team generates lots of them, the accomplishment speaks more to the team’s clutch hitting than its pitching.  If you had to summarize the Texas Rangers’ pennant-winning season with one phrase, I suspect you would more likely choose “timely late-inning hitting” than “dominant relief pitching.”

Post Script

It was my interest in the distribution of starter and relief pitcher wins that prompted my prediction about Denny McLain (see Graham’s Jan 27^th post).  On the reasoning that 1) the percentage of wins by starters doesn’t change much between eras and 2) in any era, there is only one winning pitcher per ballgame, I asserted that McLain’s 30-win season should survive his being transported to another era, provided he winds up on a winning team and is given the opportunity to pitch as much as he did in 1968.  Sound reasoning, or so I thought, but stat converter says otherwise.  1968 remains a special year for pitchers, particularly Denny McLain.

This guest post was written by Brendan Bingham. Email him at Brendan@calibertherapeutics.com