THE BOOK--Playing The Percentages In Baseball

A couple innings later, similar play from Ibanez to Ruiz, but the throw was high so Ruiz had to jump for it. The runner was safe… but if the throw was 1-2 feet lower (aka more accurate), it would’ve easily been an out.

It got me thinking...ERA should be a defensive stat for every player. It could be more like… R / Defensive Opportunities to home plate = Def ERA. So for example… say 10 runs scored during 25 opportunities the SS had a chance to throw him out… his ERA would then be calculated from that. I bet this would stop a lot of players from throwing to 1st base for the out when they could throw home.... but that’s besides the point.

I’m sure the formula would need tweaking… I only gave it about 2 minutes of thought so far. I like the idea though—Defensive ERA. Plays like these shouldn’t be credited to the pitcher in reality.

---"Cliff Lee didn’t do anything good on that play.
**************

I disagree. To me, it looked like the pitch was a pitcher’s strike right on the inside border of the zone, probably right where Lee wanted to throw it. Good pitch, bad result.

After 116 votes:

72% say Lee should get no credit for the out
28% say Lee should get credit for the out

It’s interesting.  If, let’s say, there was a 28% chance of an out occurring, then this numbers make perfect sense.  Simply put: that Lee allowed that hit implied that the runner might try for home and there was some chance of an out occurring.  So, Cliff Lee set into motion a chain of events.  If there was a 28% chance of an out on that play, then that’s what the readers of this blog want to credit for him.

So, if we extend this to all possibilities:
a. Lee allows the hit, and the runner stays put at 3B.  If that happens, 29% of the time, that’s what the readers of this blog think he should get credit for.

b. Lee allows the hit, and the runner tries to score.  That attempt will occur 71% of the time.  Of those times that he tries to score, he makes an out 28/71 or 39% chance of being thrown out.

And of those times he actually does score, the batter will get to 2B 11/42 or 26% of the time.

Therefore, the readers here, as a consensus, follows the PZR approach: Cliff Lee’s responsibility ends once the ball hits the ground.  At that point, you freeze frame the play, and establish the odds of a run score, the batter getting to 2B, and an out being made on the runner or the batter.

Mathematically, we have this:
http://www.tangotiger.net/re24.html
starting point:
2b, 3b, 2 outs = 0.626 runs

ending point:
2 runs and 0 outs, batter at 2B 11.2%
: 2 + 0.348 = 2.348

2 runs and 0 outs, batter at 1B 31.0%
: 2 + 0.245 = 2.245

1 run and 0 outs, runner stays at 3B 29.3%
: 1 + 0.530 = 1.530

1 run and 1 out to end the inning 28.4%
: 1 + 0.000 = 1.000

That’s a weighting ending point of 1.691 runs.

Subtract the starting point, and Cliff Lee’s responsibility here is 1.065 runs (above average).

Of that, we all agree that the runner on 3B was his.  That runner scoring was worth 0.722 runs. 

That means that the rest, meaning the runner on 2B and the batter, was worth +0.343 runs added.

***

Post #1 talks about PZR a bit more:
http://www.tangotiger.net/archives/stud0191.shtml#1001

A simpler way to describe it is that you look at the best-case scenario (runner thrown out to end the inning, only 1 runs scores, so an RE of 1.000) and the worst case scenario (both runners score, batter reaches 2B, so an RE of 2.348), and split the difference ( average RE of 1.674) and give that to Lee.

So, Lee started with RE of 0.626 and ended with 1.674, so his run impact is +1.048.

This is probably the best way to think about it, absent figuring out all the possibilities in-between.

#2/dave --

I’m also curious about the implicit assumption that because Lee threw (what looked like) a ball, he did nothing “good.”

Let’s start with the fact that he threw a pitch out of the strike zone that enticed the batter to swing anyway. Given that hitters generally hit for much higher averages in the center of the strike zone, and for decreasing averages as you move away from the center of the strike zone, shouldn’t we consider his pitch to be “successful” in that it induced a swing on pitch in a low-average location?

In other words, if he throws that pich, and 8 times out of 10, the batter swings through it, what’s the justification for saying he did nothing good on the 2 times when the hitter manages to make contact in fair territory?

It seems to me that the batter made great contact on that pitch.  That maybe the batter was expecting that pitch inside.

If you were to remove the plate from the equation, and just look at the pitcher and the batter, that looked to me like the batter got the pitch he wanted.  He didn’t look like he got “busted” inside or anything like that.

I’d like to see some other stills of that batter whoever it was, and see if he steps into other pitches, inside and outside, and how far away he stands from the plate.

Setting that aside, the discussion is what is the point at which you stop responsibility for the pitcher.

I didn’t see Greinke’s HR allowed yesterday, but if let’s say that given the location and sequencing of those pitches would result in a total of 0.73 HR, but he allows 3 HR, do we credit him with allowing 0.73 or 3?

That is, at what point does responsibility stop?  Just prior to the batter contacting?  Or just after?  Or by the end of the play?

Realistically, he shouldn’t get credit for runs or outs, he should get credit for giving up a line drive. That it happened to be a line drive with 2 runners in scoring position is whatever.

If we’re going to assign runs which are just flukes of sequencing and when the randomness of a ball in play occurs, might as well just go with it and always assign full credit.

The throw from Ibanez, with an average defending LF, may have prevented a run 30% of the time, why not credit him with 1.3 runs if we are going to play this game?

"That it happened to be a line drive with 2 runners in scoring position is whatever. “

This is not necessarily true.  This presumes that batters and pitchers don’t approach the base-out state (or game state) any differently.

This would be impossible.  For example, the impact of a K with men on 3b and less than 2 outs is obvious.  A pitcher will do whatever it takes to get that K and the batter will do whatever it takes to avoid that K. 

There are other scenarios where the pitcher will pitch disproportionately on the edges and disproportionately down the middle.  This is even more apparent by count.

So, I would not just say a “whatever” to the game condition as if it’s some random parameter.

I’d like to make a more simplistic addition to the more thoughtful posts above.

We already allow official scorers to assign errors to plays.  By implication, the official scorers credit runs, since runs that aren’t caused by position player errors are assigned to the pitchers.  So why not do the same with outs?

In the above example, there is a run and an out.  The run is assigned to Lee, as runs not due to errors are assigned to the pitchers.  The out could be assigned to either Lee, Ibanez, or Ruiz, since we are not allowing fractions.  In this case the rule should state that the out gets credited to the fielder who has done the most to create the out, particularly if he has done more than make a routine play.  In this case the out should be credited to Ibanez, though you can make an argument for crediting to Ruiz.

As outs are credited, there will be alot of debate on individual calls, but over the course of a season and the course of a player’s career the mistakes should cancel out.  Likewise with players (not) getting fractional credit, deserved or undeserved, that will cancel out over the course of the season.

Eventually you can start assigning outs and runs to individual players on the offense.

Tango:

Lee doesn’t have anything to do with the out at home. It’s Ibanez’ throw and Ruiz tag and holding of the ball that “produces” the out, so Lee should n’t be credited with the out if there is any (in this case, there was). He shouldn’t be penalized if the batter advances to second because that’s also a repercussion of the fielder’s choice.

To me Lee’s responsability ends with: Runner on 3rd scores, runner on 2nd advances to 3rd, batter stays at 1st. Anything that happens after that should be credited (positively and negatively) to the fielders and runners trying to advance extra bases.

Alv/11:

No, you can’t make it stop at 100% like that.  What if it was a weakly hit ball such that the runner was already halfway between 3B and home?  Ibanez still (unwisely) throws home, but the runner is easily safe even with a perfect throw.

That is, nothing the defense could have done to stop the two runs from scoring.  So, that’s why in that case, Lee would get both runs.

In this case, there was a high chance that the runner would score, so Lee needs to get a high percentage of the second runner as well.  It’s only a perfect throw from Ibanez and a great block by Ruiz that allows Lee to get 0 runs.

So, you need to step back from your position of being 100% sure.  It is ONLY because I ask people to choose one of the 4 that you can take a 100% position on one of the 4. 

But the reality is that each of the 4 answers is correct, to some degree.

Colin brought up the point of positioning for Ibanez. 

While for the sake of argument, this illustration presumed that positioning was not on the table.  After all, I’m just trying to find a decent example to have a discussion on crediting.

That said, I could have included a 5th option:
0 runs, 1 out (Ibanez should have been better positioned for the flyout)

***

By the way, for those people who selected, which is a healthy 26%:
1 run, 1 out

Do you realize you are giving zero credit to the Phillies fielders?