Hope y'all had a happy 4th. And say what you will about Juan, but he sure has had a knack for clutch hits this season. Anywho, Azruavatar already touched a bit on the bullpen's performance Monday and Solanus talked about win probability Tuesday. Today my plan is to mesh both worlds and talk about Tony La Russa and how well he's managed his bullpens over the years. La Russa is credited for popularizing the monstrous 7 deep freak of a bullpen you see today, each with it's different "specialists". You may or may not be a fan of it, for what it's worth I thought it would be fun to take a look at TLR's bullpen usage and see how well he's done at leveraging his relievers.
There are a number of way to skin this cat statistically, but being the Fangraph fanatic that I am, I think the best way to tackle the subject is by looking at three important stats, and those are Win Probability Added, Batting Runs Above Average (BRAA) and Leverage Index. Quick primer before we dive in: I think by now most of you are familiar with win probability added, or WPA. If not, I'll just direct you here. WPA is especially important when evaluating relievers because it gives them direct credit for how they help/hurt their team's percentage of winning a particular ball game. Batting runs above average is a bit of a different animal but is easy to understand. It's based on Tangotiger's run expectancy tables and play-by-play data. Depending on the base/out state, the batting team is expected to score X amount of runs. BRAA tracks the difference between the Run Expectancy at the start of the play and the end of the play and attributes that difference to the correct player. This is important for relievers because WPA is based on how crucial the game situation is, and that is measured by the leverage index. But some of these relievers pitch well but are relegated to less important situations because their manager just doesn't trust them for some reason. They may be good at preventing runs, but they are preventing them in fairly meaningless situations. I figure if they are good at preventing runs in less important situations, odds are they'll do the same job in crucial ones. Get it? This is what BRAA can tell us and this what we'll be looking at. I think we'll see some definite patterns and be able to judge what TLR is doing, as with Fangraphs we have 5 and half seasons worth of data to look at. Nuff 'splaining, into the research:
20 relievers were used, they threw a total of 529 innings with a 3.11 ERA. They posted a cumulative 5.16 WPA, or were good for about 10 and a half wins.
Who did he trust? (OK, a l'il more explaining needed-When I ask who did he trust, I mean what relievers got the ball in higher then average leverage situations then the rest of the bullpen. I consider Izzy is a given, until 2006...I'll get to that later.) Steve Kline (.98 WPA/13.08 BRAA, and I'll be posting it this way throughout), Rick White (.69/7.97), Gene Stechschulte (-.22/-3.87).
Who did he not trust? (When I say who did he not trust, that would be well below the average LI of the rest of the bullpen.) Mike Timlin (.77/7.78), Mike Crudale (.48/10), Luther Hackman (-.41)/-.80), Mike Matthews (-.62/-3.07).
Overall grade: I'd have to rate this bullpen and it's usage a A-. This pen got the job done. My only complaint was that Timlin wasn't used in higher leverage roles, which he would go on to excel at for the Red Sox. Perhaps he would've had later down the road for the Cards had he not had been traded. Mike Crudale was deserving of more important jobs as well.
22 relievers were used, they threw 460 innings and put up a 4.79 ERA. Their WPA was -3.03, or - 6 wins.
Who did he trust? Cal Eldred (-.29/1.67), Steve Kline (-.25/5.84), Dustin Hermanson (.01/-2.99), Jeff Fassero (-1.52/-11.31).
Who didn't he trust? Kiko Calero (-.39/3.25), Esteban Yan (-1.21/-13.67).
Overall grade: D. Apologies for any flashbacks I might have given some of you, I think we all still have a little post traumatic stress disorder from watching this horrific bunch. The failure wasn't all on La Russa, Izzy was on the DL for much of the season and just look at the bums he had to work with. Still, his persistence with Fassero and the fact that they didn't send Yan packing sooner then later doesn't help his case. A D grade might be harsh given the fact that you can't squeeze blood out of a turnip, but thankfully the message was received in the front office and they got some much needed help later that winter.
A total of 15 relievers were used, they pitched 457.1 inning and had a 3.01 ERA. They added 7.67 of WPA, or about 15 and a half wins.
Who did he trust? Steve Kline (.49/11.65), Julian Tavarez (.87/8.03), Ray King (2.09/21.08).
Who did he not trust? Kiko Calero (1.49/13.94), Cal Eldred (.40/-1.14) Dan Haren (.-21/4.52).
Overall grade: A+, performance wise but there were a few things that could have been done better. First of all, Al Reyes was lights out almost all season for Memphis but couldn't crack the roster until September. Secondly, Crazy Hoolie and the "Pine Tar Cap of the Apocalypse" was great and all, but Calero was clearly the better pitcher. Lastly, Ray King was nails all season but Steve Kline was used in more crucial situations. Had the bullpen been reshuffled a little bit, they may have won 107, 108 games. I guess TLR just had to settle for 105. Wimp. OK, gnat straining time is over. Oh, how I do miss those glory days.
16 relievers were used, they threw 397.2 innings and put up a 4.25 WPA, or 8.5 wins.
Who did he trust? Julian Tavarez (1.18/5.29), Al Reyes (2.15/18.12), Ray King (-1.91/-5.21).
Who did he not trust? Cal Eldred (.57/7.53), Brad Thompson (1.14/8.94), Randy Flores (.38/-2.24).
Overall grade: B+. Another great bullpen results wise, but on the negative the Burger King got broiled and La Russa still kept trotting him out there. Bad decision. Also, not showing much creativity, Al Reyes was the best reliever on the staff and yet wasn't trusted as much as Tavarez. (In case you haven't heard, Reyes very quietly has established himself as one of the games better closers in baseball for the Rays this season.) Brad Thompson also was very good, but as a rookie he was eased into crucial situations like an old man into a hot bath, as his average LI was a measly .74.
13 relievers were used and they put up a 4.06 ERA, and had a 1.24 WPA, or about 2.5 wins.
Who did he trust? Wainwright (1.70/15.44), Looper (.55/9.08), and Flores. (.29/-6.19)
Who didn't he trust? At least during regular season, Josh Kinney (.47/7.20). His average leverage index per batter faced was a pretty minuscule .47. He also didn't show much faith in Wonderbrad (.14/10.14), Jorge Sosa (-.31/2.17), Josh Hancock (.11/-2.80) Tyler Johnson (-.73/-4.66).
Overall grade: C-, but TLR figured it out in the post season, so huzzah! La Russa gets a big demerit here because of the whole Baron von Isringhausen debacle. Izzy's LI of 2.34 led the league, but he was a -.84 in WPA. It took 10 blown saves to get in through their skulls that Jason was hurt and much earlier Wainwright could have been easily installed as the closer while Looper, Kinney and Thompson would have held the fort down. Had that happened, the Tomahawk chop-page might not have been necessary. As for TLR's bullpen usage in the playoffs, A+.
16 relievers have been used over 259.2 innings and have logged a 3.99 ERA and a WPA of 4.14, or about 8 wins. Their average LI is only .87, meaning they haven't had tons of big situations.
Who has he trusted? Ryan Franklin (1.71/13.50), Randy Flores (-.65/-9.36), Kip Wells (.30/2.12), Troy Percival (.16/1.67), Troy Cate (.51/3.91)
Who hasn't he trusted? Russ Springer (.55/4.02), Tyler Johnson -.46/.450, Brad Thompson (.47/.36), Todd Wellemeyer (.24/1.22), Kelvin Jimenez (.04/-11.22)Overall grade: INC. So far, so good for Ryan Franklin and Izzy. Russ Springer has been fairly effective but has mostly pitched in low leverage situations. Kelvin Jimenez was a disaster just in mop up duty and has rightly been banished to AAA. Flores has been abysmal and Troy Cate may ascend to the top LOOGy soon. The addition of Percival is a welcome blessing so far too, hopefully he can stay healthy.
Last but not least: How well has Izzy been leveraged? Well, looking at NL closers with 20 + saves over the last 5 seasons, Izzy has finished 8th, 14th, 10th, 7th, 1st in average leverage index per game. This season he's 10th amongst NL closers. It's funny that the season he leads the league in LI was also his worst season. Overall this tells me he's been pitched in some jams (some self inflicted), but he's been used more strictly as a 9th inning man, not really a "fireman".
Overall there are some definite patterns to be found here. For one, TLR tends to do a little too much hand holding with the youngsters and some of the newbies. Al Reyes, Kiko Calero, Adam Wainwright, Brad Thompson, Josh Kinney, etc. all had to gradually work their way up into higher leverage situations, and at times that was to the team's detriment. I can't blame TLR for wanting kids to prove themselves, but not while hanging on to languishing vets because of their rep and pay scale. He didn't do that with Wainwright, however, giving him more important roles then the ex-Met closer Looper. Another thing that bugs me about TLR is the over reliance on LOOGys, especially over the last 3 seasons. These guys have been a killer, particularly Flores this season and the last. My final gripe is that he seldom will call on his relief ace when the game is truly on the line, more often then not he will hold onto Izzy strictly for a "save" situation. On the positive side they have taken some guys who were considered fungible and have turned them into some useful parts. TLR has done a decent job developing young relievers as well as milking the best out of previous unknowns and scrapheaps like Cal Eldred, Al Reyes, Kiko Calero, Josh Kinney, and Ryan Franklin to name a few.
I think the results speak well for themselves. You may not like the La Russian monstrosity, but more often then not, it's helped the team win. It would be nice to see him think outside of the box more often. There's a 2006 World Series flag flying that helps prove that doing such can get results, too. (And when I say think outside of the box, I DON'T mean converting relievers into starters! Sorry, just venting some frustration over the last two games. I'm pretty sick of that type of out of the box thinking.)