Playoffs Or Bust?: Molitor's Murky Future

Believe what you want. Learn to look it up.

Actually, it is you who is wrong here. The leverage index for the plate appearance in question (Boshers vs RHB Jackson) was 1.30. Not 0.33. Citing a different plate appearance is misleading at best.

For the chance for Cleveland to win the game.

Which is the also the chance for Minnesota to lose the game, which is what we should be trying to avoid. By WPA, we had a 19% chance to win the game when Jackson was coming up. Not great, but not the kind of chance you disregard in the opening game of a 3 game home series versus your division rival in a pennant race. Especially not in regard to bullpen decisions after an off day with a likely cancellation the next day too.

It is a fair criticism, but probably does need to be balanced with the fact that the move that was criticized was remedied today yet had the same result.

So the standard you are applying here is that a move has to have a virtual 100% success rate for a manager to be criticized for not making it? I don't get it.

Everybody here is fine with bringing in Boshers to face the LHB, even though it has far less than 100% success rate. Why do we suddenly get so stringent when analyzing the follow up move?

It was the 7th inning. We needed 7 more outs, win or lose. Boshers was never likely to finish the game. He hasn't gotten 7 outs in a game since his season debut, when he came in during the 3rd inning of a 14-3 loss to Detroit. He hasn't faced 9+ batters in a season except when he came in during the FIRST inning of a 9-0 loss to the White Sox.

Molitor already decided to pull Busenitz with 2 out and nobody on to play matchups, and nothing had really changed with our win expectancy since then, but the leverage had gone way up. Molitor was playing matchups, until suddenly, he wasn't? I don't fault Molitor based on the results so much as the internal inconsistencies of his logic.

My response is that it is defensible to try and steal a couple more outs in that situation. Molitor is playing with such a short deck of good relievers that he has try whatever he can to get through situations like that where he (wisely) won't use his handful of actual mlb quality relievers.

 

He also knows he has Gibson the next day who often gives short starts and thus might have to burn more relievers the next day.

 

Molitor is very aggressive when he has leads trying to get wins, takes a few more risks when trailing, which is very defensible in the grand scheme of the bullpen structure he has and upcoming starters he had lined up.

Everybody and his dog knew the game on Wednesday was a likely postponement.

And if you are squeezing outs from RPs, why pull Busenitz with 2 outs and nobody on? Let him get out of the inning and then see what our bats can do to inform your next pitching move. Busenitz was already pretty much unavailable for Wednesday anyway.

And the limited deck of options actually means you have less to gain from saving marginal pitchers for later. Was Pressly really going to be any better or more reliable with 2-3 days rest? All of these guys are optionable too if you really need a fresh arm.

It obviously isn't what cost us the game Tuesday, but it is the kind of move sequence I just don't get with Molitor.

I assume he pulled Busenitz to keep him available for the next day. He's managing the whole season, he had something like 9 games in 8 days staring him in the face.

I understand the scenarios you are pointing out, I just see a manager trying to steal some outs when trailing late in a game. I don't especially agree with the calculations and tradeoffs you have suggested on usage. I do agree he didn't use the most optimal pitching matchup, but I would still argue it was a calculated risk to preserve usage.

Again: the next day's game was fairly unlikely.  (I should know, I was hoping to attend Wednesday night, so I was following the weather pretty closely, and there was a major storm rolling in.)  And even with the uncertainty inherent in any weather forecast, the rest of the 8 MAN pen was completely rested. following the off day on Monday.  Keeping Busenitz available was hardly necessary.

Even so, whether Busenitz gets 5 or 6 outs on Tuesday, or throws 21 or ~26 pitches, doesn't seem like a particularly meaningful distinction to that end anyway.  Busenitz has only thrown on consecutive days once all season, and it was after only 15 pitches the day before. And that was in a stretch where the Twins had 21 consecutive days without a day off.  Half of Busenitz's MLB appearances this season have seen him throw more than the 21 pitches he threw Tuesday.  When Molitor pulled Busenitz, it was pretty clear he was playing matchups to keep the Tuesday game close, not trying to stretch his pen.

And two batters latter, it was pretty clear he abandoned matchups, even though the game was still just as close.  I don't think the pen options were good enough to warrant that sudden change in strategy.

A 19% chance of winning is ow leverage.

Not really. It was 21% when Molitor bothered to go for the platoon match up with 2 outs and nobody on. Is that something you do in "low leverage"?

Tigers had basically the same ~20% chance of winning at the end of their 8th and start of their 9th down 1 run last Saturday. At that point, you are a "bloop and a blast" from a tie or a win. Note the Tigers used their best reliever to pitch the 9th to preserve that ~20% win probability, clearly not playing it like it was low leverage.

Low leverage would be when the Tigers were down 4 earlier in that game, or the Twins were down 5 after the Jackson HR Tuesday, about a 4% chance of winning. You hope for a comeback at that point, but there is not much you can do. Still plenty you can do around 20%, though.

If you look around the league, I doubt you will see a lot of strategic priorities changing just because of ~20% win probability. You don't have to go all "elimination game" in your reliever usage, but you sure as heck don't shrug and leave your scuffling LOOGY out to dry vs RHB with a rested pen in reserve.

Statistics from those who spent a lot of time working on them say it was a low leverage situation.

Again, no, you were looking at the wrong plate appearance. 0.33 was 2 outs, nobody on. When Boshers was allowed to face Jackson with 2 on, that was 1.30 leverage. That's not low, it's above average, fairly close to the average game entering leverage index of Pressly, Duffey, and Belisle.

As the next day pointed out, switching from Boshers did not provide a better result. Molitor did as you said. No change in the results.

There is no such thing as guaranteed results. Doesn't mean you ignore the process.

And if Pressly came in and gave up a HR to Jackson in the game Tuesday instead of Boshers, then maybe Molitor tries a different RHP in the Thursday game with better results.

The process of using one tiny bit of information to make a decision is a flawed one.

 

I am not using "one tiny bit of information" here. I have considered not only Bosher's platoon splits, but also his overall quality (he's a fringe MLBer), the batter's platoon splits (neutral career), the size and composition of the Twins pen, how rested the pen was, the likelihood of a game the next day, etc. Throw in that we had 3 offensive innings left, at home, Salazar was dominating but probably only had 1 inning left, and Andrew Miller was still on the DL which meant their pen wasn't at full strength. And that we already cut Busenitz's appearance somewhat short, so we already invested something in this game. Even if we lost, the outcome of that PA could have had more informational value too -- can a RHP do the job? That seems like a more valuable question to work toward answering than, can Boshers wriggle out of his own jam that I would never trust him to do in a more important game anyway?

A lot of factors informed my conclusion that our win probability at that point wasn't worth risking to squeeze one out from Buddy Boshers. Feel free to disagree, but please don't mischaracterize my argument and evidence.