THE BOOK--Playing The Percentages In Baseball

Friday, September 25, 2009

Poz revists 1980.  But, what does WPA say?

In talking about the Cy Young award given to Steve Stone ahead of Mike Norris, Poz concludes:

Steve won the award even though Norris:

– PItched 34 more innings.
– Gave up 10 fewer earned runs — his ERA was 2.53 to Steve’s 3.23.
– Struck our 31 more. Walked 18 fewer. Gave up fewer hits. Threw 15 more complete games. Allowed the fewest hits per nine in the league. And so on.
– Won only three fewer games despite playing for a significantly inferior team. His A’s won 83 games. Steve’s Orioles won 100.

So yes, I fully appreciate Steve Stone standing up for wins. He should actually be president of the “Wins” fan club. He should have “Wins Rule!” bumper stickers made. And I didn’t not realize until this very moment — because I do love Steve Stone, even now — that this whole thing is not at all about the potential injustice of Zack Greinke losing the Cy Young. I do think Zack will win.

No, this is about the injustice of Mike Norris not winning the Cy Young in 1980.

Steve Stone winning is like Dawson winning the 1987 MVP.  Wins, RBIs, etc.  The staples.  But, Fangraphs gives us a different view into things.  Namely WPA.  If you sort by WPA, two things jump out:
1. Mike Norris had a +6.2 WPA, far ahead of number two (Tommy John’s elbow at +3.5), and Stone is in sixth at +2.8.
2. That list is littered with relievers, with rookie Doug Corbett front-and-center at +7.6 wins (on 136 innings), and 2nd-yr player Quiz at +7.1 wins (on 128 innings).

Imagine, for a moment, you take a typical ace reliever today.  And you have both of them on the same team.  And you fuse them into one body.  That’s what Corbett did.

Now, as I talked about in the past, you can’t just do a straight-up comparison of starters to relievers with WPA and not make adjustments.  You have to make the adjustments, as I discussed with Gagne, Prior and Schmidt in 2003.

You have to also remember what you had in 1980 in Quiz: a reliever with an LI over 2 and pitching 128 innings.  That’s astounding!  Who can imagine so many high leverage opps available for one pitcher.  But, he got them.

Anyway, let’s follow the process laid-out there.  Corbett had a +7.6 WPA on a 1.76 LI, which is a +5.9 WPA on 1.38 LI.  Quiz is +7.1 in 2.14, or +5.2 LI on 1.57 LI.  Next, for replacement level, we add .030 wins per 9IP times his adjusted LI.  We give Corbett 0.8 wins and Quiz gets 0.7 wins.

Net/Net: Corbett’s wins (adjusted and baselined) is 6.7, and Quiz is 6.0.  Corbett’s got the slightly better case than the two, but you can justify either one.

What about Norris?  Well, we add .12 wins per 9 IP, and with Norris having 284 innings, that adds an extra 3.8 wins on top of his already +6.2 wins above average, for a total of 10 wins.  Steve Stone comes in at around 6.1 wins (adjusted and baselined).

Stone was as much a story as Corbett or Quiz.  And none of them was close to Mike Norris.  Not unless Mike Norris benefitted a tremendously fantastic amount from his fielders.  The A’s did have the lowest BABIP allowed in the league, and they did have Rickey, Dwayne Murphy, and Tony Armas (might be equivalent to the current Mariners outfield?).

I love how we can get a different view into the past.