Risk Aversion

I will try not to make this a frequent occurrence, as most people familiar with me know what I am about to say, but I just have to complain once again about Charlie Manuel‘s use of his starting rotation and his lack of use of the bullpen. Roy Oswalt made his third start since being activated from the disabled list in early August. His first start back was unspectacular and he didn’t have his normal velocity against the San Francisco Giants, who scattered 12 hits over six innings. Oswalt bounced back on August 13 with seven solid frames against the Washington Nationals while his velocity returned to normal.

Last night, however, Oswalt had arguably his best start of the season. Over eight innings, he struck out nine Nationals, beating his previous season-high of seven against the San Diego Padres on April 21. While there is some skepticism about the accuracy of the radar guns at Nationals Park, Oswalt was consistently around 91-94 MPH, averaging over 92 MPH. If there were any concerns about Oswalt last night, they were quickly laid to rest with his dominance of the Washington lineup. It was a pretty easy game. While Oswalt was breezing through eight innings, the Phillies took advantage of the mediocre John Lannan and the Nationals’ unreliable defense, tossing up five runs through six innings.

The issue came going into the bottom of the eighth inning. Oswalt, three starts removed from the DL, was at 96 pitches and the Phillies were 98 percent favorites in the game (and virtual locks for the playoffs), according to FanGraphs. While the bullpen was short a couple arms (Ryan Madson and David Herndon), recent call-up Michael Schwimer was certainly available, having spent a few days in the ‘pen with no actual work. Instead, Manuel sent Oswalt out to start the eighth.

Oswalt continued to dominate, striking out the first two Nationals he faced, Ian Desmond on five pitches and Rick Ankiel on nine. Ryan Zimmerman kept the inning alive by singling on the fourth pitch he saw. Oswalt had accrued an additional 18 pitches since the start of the inning, putting him at 114. Manuel strode to the mound, but as has been the case so often this year, did not remove his starter. He had a light-hearted chat with his pitcher, then jogged back towards the dugout.

Even if there was a legitimate reason to let Oswalt start the eighth, there certainly wasn’t one to let him finish off the inning. It has little to do with the specific pitch count, other than that it was “high”. Every pitch carries more risk than the preceding pitch. Even if we can’t quantify exactly how much, we know this to be the case. Additionally, when assessing the risk and reward of each scenario, we are really looking for an expected value; basically, what we stand to gain and how often we can expect to win versus what we stand to lose and how often we can expect to lose. Again, even if we can’t quantify every aspect, we can at least give ourselves a good idea of where we stand.

After the Zimmerman single, the Phillies were 99 percent favorites to win the game, up one percent from the start of the inning. Let’s make a list of potential rewards for letting Oswalt finish the inning:

I can’t think of one. The Phillies stood to gain one or two percentage points in win probability (virtual locks) and zero change in their playoff standings, since they are virtual locks and were playing a 60-63 team with no playoff hopes whatsoever.

How about a list of things that could have gone wrong?

  • Oswalt gets injured

What would they stand to lose in that scenario?

  • Their post-season starting rotation is short a pitcher, either putting more strain on Roy Halladay, Cliff Lee, and Cole Hamels, or requiring the Phillies to use Vance Worley in his stead. While there are arguments to be made that a swap from Oswalt to Worley wouldn’t be negative, it is certainly not a situation the Phillies would like to deal with, given their current situation.

Even if the probability of injury is small, say five percent, it still vastly outweighs the next-to-nothing the Phillies gain from using Oswalt for a meaningless regular season game in mid-August.

As I’ve said in previous articles, I fully recognize that Manuel is extremely skilled at managing a clubhouse. A strong argument could be made that he is the best manager in that regard in Major League Baseball. Perhaps he better earns the trust and support of his clubhouse by taking risks such as leaving his starters in for an inning or two too long. That’s certainly possible.

Even if that were true, though, is it really necessary to risk the team’s playoff future and a player’s short- and long-term health for that? That’s about as close to a legitimate argument as I’ve heard support the superfluous taxing of the starters’ arms, and it still comes up very short, in my estimation.

Finally, what is the point of going with a 12-man pitching staff when you aren’t going to use them all? Schwimer was promoted on Wednesday, and he has yet to pitch in a game. The eighth inning of last night’s game was a perfect opportunity to get the rookie’s feet wet: it was a low leverage spot and the Phillies would have greatly appreciated a solid inning (or two) of work with Madson and Herndon unavailable. Yet the rookie sat on the pine again while Oswalt tossed pitches 97 through 116 and the Phillies’ odds of winning increased by one whole percentage point. With Ross Gload playing injured (often needed to be pinch-run for when he gets on base) and only one other infielder on the roster with Placido Polanco injured, it would seem better to carry 11 pitchers and add another back-up infielder, perhaps Pete Orr.

The Phillies are on pace for 106 wins. They’re super awesome and we should all be extremely grateful for what we get to watch on a daily basis. However, it will all be for naught if the Phillies falter in the post-season because one of their starters suffered an injury in the eighth inning of a meaningless game in August because Manuel didn’t want to make his guy feel less manly, or because he wanted to chase a CG SHO. It’s just not worth the risk.

Ryan Sommers put it beautifully in his article on August 17:

Even once the playoffs arrive, for all the hair-pulling, hand-wringing, and second-guessing that will occur, the Phillies are essentially throwing the best roster in the league at a swirling cyclone of small sample variance and unpredictable machinations and hoping that they’re spit out on the other side as World Champions.

Let’s hope that “the best roster in the league” they throw at that swirling cyclone of small sample variance includes each every one of the four aces they worked to hard to acquire and keep around.

Leave a Reply



  1. Phillie697

    August 22, 2011 10:52 AM

    Oswalt’s track record is just too hard to ignore. I’m willing to bet he’s on the post-season rotation and not Worley. I would trust a healthy Oswalt with a post-season start over Worley ANYDAY.

  2. KH

    August 22, 2011 11:07 AM

    I agree with Bill on this one because Oswalt has an injury history.

  3. Jonny

    August 22, 2011 02:21 PM

    I LOVE Uncle Charlie. And I’m with Bill on this. Oswalt probably should not have pitched the 8th even if it did turn out good.

  4. hk

    August 22, 2011 08:28 PM

    What’s Charlie going to do tonight? Lee’s at 105 pitches with a shutout through 7 and an insurmountable lead. Will he let him through ~20 pitches to get the CG SO? If so, will Bill have another article on the site by 3AM?

Next ArticleGraph of the Intermittent Time Period