THE BOOK--Playing The Percentages In Baseball

"It imparts no meaningful information...”

Come on, “no” meaningful information?  Would it have been so difficult for him to be completely honest and say, “Little,” or “Very little?”

In case you haven’t noticed, I hate when smart people say dumb or incorrect things just to make a point or because they are sloppy. I will occasionally do the same and if anyone points it out to me, I will usually admit my mistake (like calling Ichiro perhaps the greatest “hitter” of all time).  Most of these guys when you call them out will either ignore it or make up some excuse or poor rationalization.

I don’t like the premise of the article which is basically that the win used to mean a lot and now it doesn’t.

The core reason is that the responsibility for a team win is now spread out to the point where a starting pitcher typically does not bear enough of the credit or blame for the outcome of the game to pin the entire thing on him, as we do when we give any pitcher a win or a loss. We can think about the credit for a team win by looking at what I call “The Pie of Responsibility,” because we’re all familiar with pie charts, and besides, pie is awesome.

And…

When Cy Young pitched, a win truly meant something. That’s no longer the case.”

His premise is that in the (very) old days starters pitched the whole game most of the time.  O.K.  Well, that is one reason why pitchers nowadays get lots of no-decisions. Because the bullpen comes in and blows a lead or helps turn a loss into a win.

It’s not like they always award a win or loss to the starting pitcher, no matter what.

And this:

The Pie of Responsibility

Offense:

Defense:

Bullpen:

Randomness: 

For the purposes of this thought exercise, we’ll just say that the starting pitcher’s share of “The Pie of Responsibility” for a team win in any particular game is less than 30 percent.

So in the old days, there was no offense, defense, and randomness?

He doesn’t say that, but he does imply that there is a large difference between the starting pitcher’s responsibility for a win or less in the old days and now.

The only difference I see is the bullpen and, as I said, that is mitigated by the fact that the pitcher may not get a decision.

So I don’t really see the point of the article, other than pitcher wins don’t tell us all that much about how they pitched.  And the sky is blue and my grass is plastic…

---"So I don’t really see the point of the article, other than pitcher wins don’t tell us all that much about how they pitched.  And the sky is blue and my grass is plastic…”

Well, that *was* the point of the article. It may be ho hum news to saberists, but he seems to be directing it more towards Cy Young voters, who may have to deal with the issue with Felix vs CC in a month…

I liked the article.

I’m curious what you smart people think of the breakdown Law provided. Is this about right?

http://assets.sbnation.com/assets/447197/keith_law_s_pie_of_responsibility.jpg
(As created by Sky Kalkman)

Is that too much offense? Too little defense? Not enough randomness? Or does that sound about right, as a general estimate?

I had a proposal in another thread that if we had to do it over again, the win should have been awarded to the player who had the biggest individual role in his team winning, not necesarily the pitcher.

Sort of like a MVP for the game, but done by the official scorer using some sort of objective criteria and not through polling sportswriters.

Nonetheless, even with my proposal one person would have been awarded the win, even though lots of people would have contributed to the win, including non-player employees of the team.  If you do something this arbitrary, then it will be arbitrary.

But over the course of a career the arbitrariness should cancel out.

"Well, that *was* the point of the article. It may be ho hum news to saberists, but he seems to be directing it more towards Cy Young voters, who may have to deal with the issue with Felix vs CC in a month…”

Had he left out the thesis (which was a major part of the article, I think) that the win was fine in the old days, but not now, I would have no problem whatsoever with the article.

I think that “randomness” is mostly timing.  Our team wins 5-3, but when the 5 were scored and when the 3 were scored can determine whether the starter gets the win, a reliever, or whether there is a blown save and a reliever win.  So with a complete game I think most of the randomness disappears.

To allocate I think of it this way; the average starter in a win goes about 6 2/3 innings.  Subtract 1 for defense.  In close games the relievers are pitching at higher leverage so should get more credit.  I subtract another 1/3 inning.  Maybe timing issues should lead to another small subtraction.  Divide by 2 for the offense.  and we have 25 to 30%.  Of course somebody gets the Win.  If it is a reliever it is mostly timing, so not remotely close to 25%.