Building A Bullpen - Part IV

The save is a statistic that has been disproved as far as any real predictive value or even descriptive value. The way closers are utilized primarily in the 9th inning is often a misrepresentation of when the leveraging index research would tell managers to use closers (i.e. their best reliever). I have my differences with the leveraging and win expectancy but we've all watched as managers have let a middle reliever face the heart of an opponents batting order in the 8th inning and then the closer to face lesser hitters in the 9th. There's an intuitive aspect to that that just doesn't seem to make sense.

Tony LaRussa obviously ascribes to that school of thought, which considers the mentality required for the 9th inning significantly different from the 8th inning and thus requiring a unique reliever. This isn't an unusual train of thought nor one that I'm particularly inclined to berate LaRussa for. I don't like it but the insular nature of baseball makes it hard for new ideas to really penetrate the business at times. At least we haven't been subjected to (very many) seasons like Joe Borowski for the Tribe in '07 or Todd Jones for the Tigers in '06. Knowing that players like Rafeal Betancourt and Joel Zumaya were pretty clearly the superior pitchers in those seasons compared to their closing counterparts is frustrating.

That was the case for the Cardinals in 2006 as we all watched (and often cursed) the blown saves from Isringhausen. It's safe to say that very few, if any, of us realized the real extent that his hip was prohibiting him from pitching. That season Adam Wainwright, who had been put in the pen in a misguided effort to wring value from Sidney Ponson, put together a fine collection of relief appearances before turning truly dominant in the postseason. Watching Isringhausen's dramatic improvement in 2007 makes for a slightly different situation to the Tigers and Indians who simply put inferior pitchers in as "closers".

The Tigers situation in 2006 was closer that I would have guessed looking at peripherals. Zumaya has a pretty sizable advantage when considering WPA and WPA per plate appearance. Plus, just ask yourself who you'd rather have finishing games in 2006: Joel Zumaya or Todd Jones? Without scrutinizing the numbers any closer, my gut reaction says Zumaya was the better pitcher. The difference between Borowski and Betancourt, however, is so extreme that it makes one wonder what Eric Wedge was thinking. You can read about WPA and gmLI at Fangraphs -- the short version is that WPA is a counting stat for wins added where a win equals .5 and gmLI is the average leverage index for a reliever when they entered the game.

In any case, the idea of the closer being confined to the 9th inning seems flawed and the notion of saves as a worthwhile statistic seems a bit archaic. I certainly don't want to pick a pitcher based on their year to year saves. A quick look at the saves leaders from 1987-2007 shows 130 different players who recorded 20 or more saves in a season. Of those 130, only 35 managed to have at least three consecutive years of 30+ saves and only 20 had at least 4 consecutive years of 30+ saves.

Continuing down the road of saves led me to look at the career leaders list. That makes me wonder if there isn't some nugget of truth to be found in the save statistic -- count something for long enough and it starts to look important! I thought it would be interesting to look at the active leaders left (Trevor Hoffman - 1st, Mariano Riveira - 3rd and Billy Wagner - 7th) as well as our own incumbent closer who ranks 24th all time. I rifled through some numbers and thought that this graph was pretty interesting.

Each players K/BB is plotted versus their age through 2006 in seasons where they accumulated saves. While Isringhausen doesn't look all that special compared to the greats, there's some pretty wild relief seasons in there. They weren't necessarily the years where the reliever had the most saves but the peripherals are outstanding. There's some rounding in the table, which is why it doesn't exactly match the graph.

Following this to it's logical conclusion, if a reliever is truly a top echelon performer, than chances are at some point they'll get a chance to close games and garner some saves. If they have continued success, they'll probably be given every opportunity to retain their "closer" title even past the point of being the best reliever on their team. Those pitchers who have climbed up the leader board in career saves combine being the best reliever with having a chance at getting those saves. They probably have been the best relievers in terms of reliability and performance over an extended time period.