.comment-link {margin-left:.6em;} <$BlogRSDUrl$>

Tuesday, May 02, 2006


Torre and Tie Games on the Roads 


球季開始有一次機會讓 Mo 上來,最後 Torre 讓 Scott Proctor 把球賽給敗掉了。那一次我在寫 blog 時本來想要狠狠批評一下,畢竟 Torre 違反了 statheads 的 "traditional wisdom" (statheads 可以有所謂的 traditional wisdom 了嗎?)。由於球季剛開始,那個主題就輕輕放下,並沒有對 Torre 提出非議。

在那之後一個星期,我對之前放過不批評一事感到非常後悔。我們看看 Mo 今年四月出場記錄:

4/09 @ LAA 1.0 IP Yankees won 10-1
4/11 v KC 1.0 IP Yankees won 9-7 (1st sv of the season)
4/15 @ MIN 1.1 IP Yankees lost 5-6 (1st blsv and 1st L of the season)
4/19 @ TOR 1.0 IP Yankees won 3-1 (2nd sv)
4/23 v BAL 1.0 IP Yankees won 7-1
4/26 v TB 2.0 IP Yankees lost 2-4 (2nd L)
4/27 v TB 1.0 IP Yankees won 4-1 (3rd sv)
4/30 v TOR 1.1 IP Yankees won 4-1 (4th sv)

9.2 局以四月來說其實不能說太少,去年 Mo 在整個冬天都沒有摸球的情況下,手臂雖然休息夠了,但是進入狀況特別的慢,以致於整個四月只有 6.2 局。前年他四月投了 13.1 局,2003 因為受傷只有 1 局。2002 有 11 局、2001 12.1 局、2000 12.0 局、1999 10.1 局。

如果身體狀況正常,Mo 應該要有 10 到 11 局的工作量。以這個數字來講,其實相差非常小。但是以他出場時機來看,我們可以看出今年這 9.2 局並沒有充分運用到 Mo。在那 9.2 局裡面有兩局是派 Jim Kaat 可能也可以全身而退的比賽中投出來的 (4/09 和 4/23),真正的 high leverage 局數嚴格說還要更少,像 4/27 和 4/30 這兩次也不算。

Mo 今年的健康情況可能是過去兩年最好的。既沒有缺乏休息也不至於因為休息太多而找不到感覺。結果在今年球季初打線超級不穩定的情況下直到 4/09 球隊第六場比賽才第一次出來。他今年出狀況的原因主要是因為 rusty,而原因是 Torre 把他放在 bullpen 板凳上放到 rusty。

昨天對 Red Sox 又是在客場的 tie game,然後 Torre 又在讓 Mo 上場之前把比賽輸掉。

這一次我實在沒辦法忍住不說話了。當比賽進入後段雙方平手時,manager 應該要從最好的投手開始用起。Aaron Small 撐了兩局已經很了不起了,接下來應該要把最好的投手拿出來,最起碼也應該在壘上有跑者時把 Mo 推出來而不是 Tanyon Sturtze。如果那時候是 Mo 而不是 Sturtze,接下來也可以直接讓 Mo 面對 Ortiz。這不是馬後砲,而是因為 Mo 的 cutter 使他面對左打比面對右打更有效一些。今年到目前為止雖然例外,不過這種小樣本能給我們的 splits 應該解釋能力有限,還是看他過往的成績比較有意義。

Yankees 球迷應該都沒有忘記 2003 WS 那隻 Jeff Weaver 被打出的全壘打。那一次跟這次是類似觀念的產物。Torre 一直等待打線表現同時希望奇蹟出現,他的 long man 可以幫他一直投下去。結果是球賽敗在 bullpen 裡面第五或第六好的投手,而他的最好的 bullpen 投手根本就沒有上場機會。

希望下一次 Mo 上場的機會是明天而不是下個星期。

Comments:
推Jim Kaat。Al Leiter現在也在YES了,可能也可以考慮換他上去...。
 
4/9和4/23的出賽是不是因為前面實在是個太久沒上場,所以出來活動筋骨?
 
除了讓 Mo 上來運動一下我們找不出任何理由需要在這種時候用 "closer"。
 
Torre is still the one to blame. Neither Sturtze nor Porctor is the second best pitcher in the pen. Farnsworth is the second best reliever. Mike Myers could be the second best choise if we are about to face left-handers, but Sturtze? Proctor? Come on.

They dont call him clueless Joe for nothing.
 
The New York Daily News did call Torre "Clueless Joe" for nothing. They have that on the cover before he did anything in New York. They just thought he's clueless.

Now he is praised as a genius and often does clueless things.

This is life.
 
The difference between home and away teams is too trivial to mention. I don't even bother to go there. Everyone knows how to spell sabermetrics knows that.

The game yesterday was not the highest probability move because Farnsworth faced better hitters than the ones Mo faced. This is the whole point of the modern closer usage pattern, which is such a cliche.
 
"Most of the managers don't send their closers in tie games when their teams are VISIT TEAMS. They start from the 2nd or 3rd best reliever instead of the closer. Don't forget the visit teams need their closers to close the bottom inning once they have a lead. "

Boston is probably not among "most" you mentioned (see 4/21 for example), and I am glad that Tito is not one of them, since it is nonsense. Assume that the runs each pitcher will give up in a game is dtermined (but we could not see it in advance). If you have a lead, and 3 innings (7-9) to finish the game, it doesn't affect the result of this game no matter how you use your relievers. You would like to save your best reliever last, since you wont increase your chance by pitching him ealier. As a result, managers would pitch him last, since he might be no factor to this game. The late tie game situation is totally different. You have a better chance to win the game if you pitch your relievers from the best to the worst. If your 2nd or 3rd best reliever can not potect the lead, you would lose the game ealier by pitching them first. Again, the only advantage to let the best reliever pitch last is to save him from this game. Therefore, unless your best reliever was used in the previous game, or you really want to give him some rest, it is plain stupid to pitch your 2nd or 3rd best reliever first.
 
I dont know why you think that I didnt get the topic. I mentioned games with lead only to explain why it is better to let the best reliever close games. It is not to increase the chance to win the game, just potentially to save your best reliever for future games.

Anyway, back to the late tie game situation, now I bet that you do not think it clearly. The two cases you mentioned:

Use closer first:
1st->2nd->3rd->...->(n-1)th->nth

Use 2nd best reliever first:
2nd->3rd->4th->...->nth->1st

You mentioned "If you lost on m-th reliever (m <= n), you lose in both ways." You want to think it again?

Suppose that your m-th reliever will give up some runs. You think that the result will be the same no matter he pitches in x-th inning or (x+1)-th inning? In a time game, if he gives up runs in x-th inning, your team lose immediately. However, if he gives up runs in (x+1)-th inning, there is always a hope that the home team gives up the same or more runs in (x+1)-th inning. Since the home team will probably have to use a worse pitcher to pitch in (x+1)-th inning, the chance of giving up runs is increasing as the game goes on. If not to rest players, the best strategy is to use your relievers from the best to the worst. It is plain and simple.
 
I was talking about the case that the visition score runs in (x+1)-th inning, and m-th reliver gives up runs in x-th or (x+1)-th inning. Not the case you mentioned.

Let's make the example as simple as it can be. Suppose that the visiting will not score runs untill (x+1)-th inning, and the m-th best reliever is the first to give up runs. Does it matter that he pitches in the x-th or (x+1)-th inning? If he pitches in the x-th, the team loses. If he pitches in the (x+1)-th, the team may win it at (x+1)-th or later.

I probably did not make my point clear, and I hope that it is clear this time. I was not to use "plain simple" to emphasize my point. I thought that it was plain simple, but probably I was wrong.
 
oops, I dont know what to say. When did I agree on the normal case that the first method has disadvantage? If winning is the only thing to concern (ignore whether to save pitchers for other games), both methods do the same in "normal" case. However, in some cases like I mentioned, the first method has some advantage. Therefore, overall, the first method is better.

Worry about the dropoff after 3rd or 4th? You can try m=2 or 3, and you still have those examples.

If the home team has the best reliver to pitch on 9th, 2nd to pitch on 10th, 3rd to pitch on 11th, ... the visiting team has a better chance to win this game with 2nd best to pitch on 9th, 3rd to pitch on 10th, and so on? I thought that most people would feel which is the better way to help the team win this game, without doing any analysis.

While pitching on 4/30, Rivera did not pitch on 4/28 and 4/29. If Torre tried to give him more rests, fine. However, I doubt that it is the case, especially Rivera didn't pitch much this season. In addition, I believe that people are not arguing whether Rivera should face no.8. It is when 2 on to face no.2 and no.3. Isn't this the case that you mentioned to use the best reliever to get out of troubles?
 
This closer issue has been widely discussed before in this blog and in PTT BBS Yankees board, so I don't want to spend much time on it.

There are two quick cuts I can make:

1. Torre managed ahead in the second game into the season and the consequence is that we didn't get to see Mo until the 6th game, that game was a blowout and usually doesn't need Mo. He didn't get his first save opportunity until the 7th game of the year, while on the same day Papelbon gathered his 4th save.

2. As the old saying goes, tomorrow may rain.

It doesn't pay off to manage too far ahead in the regular season than in the playoff.
 
To damonleadoff

Right from my first reply, I had my assumption before my statement. And I wrote it in almost every reply. You ignored that every single time, just to attact the statement. That's why we will never agree any single thing.

The case I mentioned may be rare, but I am not sure it is rare comparing to the case that you need a lot from your best reliever in the next games. This is not trivial, but this is not something I talked earlier.

BTW, anyone paid a little attention to the weather would know that the chance of having sox-yanks game on 5/2 is very little. The chance of having the game and needing Rivera is even smaller.
 
To damonleadoff,

My 1st reply: Therefore, unless your best reliever was used in the previous game, or you really want to give him some rest, it is plain stupid to pitch your 2nd or 3rd best reliever first.

My 2nd reply: If not to rest players, the best strategy is to use your relievers from the best to the worst. It is plain and simple.

My 4th reply: If winning is the only thing to concern (ignore whether to save pitchers for other games), both methods do the same in "normal" case. However, in some cases like I mentioned, the first method has some advantage. Therefore, overall, the first method is better.

You can keep claiming the 2nd method is better because of saving the closer. I am not going to argue it as it is something too complicated (hard to argue how to compare the chance to win and to save a pitcher for later use). If you still dont think that 1st method is better under my assumption, so be it. I dont think any more words from me would help.

As to the question whether to use Rivera at the beginning of 8, I probably wont as the manager, but I think that it is a close call. I do strongly believe that Rivera should be used to face Loretta.
 
Post a Comment


This page is powered by Blogger. Isn't yours?

Page visited since 1/28/04
Hit Counters
Ad: Bionicle
Listed on BlogShares