THE BOOK--Playing The Percentages In Baseball

As soon as I saw him setup to throw to second my jaw just dropped.  What a bizarre decision.  All I can think of is that he thought he could get a DP or he forgot there was a runner on 2B. 

Any chance this quiets talk of the Angles “doing the little things” and “playing the game the right way,” etc?  Nah, didn’t think so…

Just a mental error by Izturis.  He was geared up for the DP and in the heat of the moment didn’t realize that there was no chance for a DP there. Should have but didn’t. I’m sure (I hope at least) that Scioscia will explain it to him after the game.

After seeing Vladdy flail away at everything - fastballs inside, breaking pitches away - even a fastball almost a foot outside, I’m starting to think that I might IBB Hunter for Vladdy with runners on 2 and 3 and 2 outs!

The other thing Itzuris had to take into account was the game conditions:  wet and cold.  Grips on balls, footing of fielders, etc. would all be sub-optimal.  If there were ever a situation where the throw to 2B was correct, this wasn’t it.  Take the sure out, don’t try to be a hero.

The throw to second without a possible DP would never be correct in that situation (where the throw was obviously awkward), wet weather or not.  It would only be correct if it were as easy as a throw to first, which clearly wasn’t the case.  Just a brain cramp by Izturis.  In defense of him, you are trained as a fielder to normally try and get the force at second with 1 out even if you can’t turn the DP, so as to keep another runner out of scoring position with 2 outs.  Then again, a good fielder adapts to the situation.  As well, he should have been reminding himself that in this situation, where the runner going to second means nothing, absent a DP, the throw goes to the easy base.

The win probabilities there are fascinating because I was thinking about those exact situations twice last night. The Yankees had runners on 1st & 3rd with two outs in the 9th & 12th innings. Both times the runner on 1st took 2nd on defensive indifference.

According to FanGraphs (click on my name for the link to last night’s play log) the win probabilities were:

1-3, 2 outs, tied: .645
-23, 2 outs, tied: .642

Which really doesn’t make sense to me. 2nd & 3rd in particular...why does FanGraphs have the win% at .642 while you have it at .674?

Joe Morgan on ESPN Radio absolutely killed Izturis for this play.  When Joe and Tango forcefully agree, that’s a pretty impressive alliance.

That said, I have a hard time being so harsh on a player for making a split-second decision during a fielding play.  Izturis has been a great defender this year.

The throw to second without a possible DP would never be correct in that situation (where the throw was obviously awkward), wet weather or not. 

The throw to second would never be correct in the wet and cold weather, regardless of the game situation either.  If the Angels had scored a run and therefore putting the runner in scoring position added to Yankees’ win probability, you still shouldn’t make the throw when the field conditions make the chance of what actually happened much greater.  You talk about how managers don’t have the chance to make good decisions, this is in fact something a manager can discuss with his fielders during the game.  It’s the coaching staff’s obligation to make sure the players understand that.  You don’t run wet, cold scenarios in spring training.  Of course, baseball is also not meant to be played in these conditions.

"After seeing Vladdy flail away at everything”

That’s nothing new.  The problem is he used to be able to hit those pitches.

FAngraphs uses Markov that I supplied.

The link I supplied in this thread is empirical (actual).  Will have sample size issues.

Since the only difference between 1 and 3 and 2 and 3 in that situation is the rare occurrence of a force at second but no play at first or an easier play at second than first (so more errors on the throw to first), the WE should be almost the same.  .674 and .661 is too great a difference I think.

“The throw to second would never be correct in the wet and cold weather, regardless of the game situation either.”

That is probably true as a general rule, but certainly it depends on the chance of error on the throw to second as opposed to the throw to first (as well as the chance of a DP, which in this case was zero).  There is a break even point of course.

MGL, when I use Markov, the win probability is almost a dead heat, as you’d suspect.  Indeed, the win probability gets flipped around!

That, however, is a quirk in how I do the Markov.  Anthony/5 spotted this quirk as well.  For an explanation, you can read it here:
http://www.insidethebook.com/ee/index.php/site/article/quirks_in_the_win_expectancy_tables/

Anyway, MGL is right that you haev to be careful.  The results are based on samples, and so, there is a sampling error.  And that the .016 win difference is probably not significant.

Playing 5 hours and 13 innings out in the rain could help cause a mental slip here or there. I thought they should have stopped the game earlier and picked it up the next day.

Tango is right that he should still know where he is going to be throwing the ball.

It’s shocking how many players haven’t rehearsed in their mind what they are going to do with the ball until it is hit to them.

Brian/13: right, exactly my point.

It’s not like this required “split-second thinking”.  You need that if you are unprepared for something that does not happen much.

A ground ball where the 2B has to run back in the 1b/2b hole?  Are you kidding me?  He should have been totally prepared.

And, his line after the game shows that he didn’t get it.

Of course, the real problem was that the second base umpire had decided to be a stickler about the fielder having to actually touch the base to record a force out, so Aybar couldn’t field the throw as far away from the bag as he liked.