WPA #8: What!? Another Pirate?

Moving right along in our look at the top ten offensive swings in win expectancy, we find ourselves at spot number eight, which happens to involve another member of the Pittsburgh Pirates. They might not know how to win, but apparently they know how to win big. The eighth biggest offensive play of the season took place on April 14, when the Pirates visited the Dodgers. Hiroki Kuroda took on Zach Duke, and both delivered somewhat stellar performances. Duke failed to strike out a batter but still recorded a quality start, and Kuroda had finished his first three starts with a 5.0 K/BB and 2.89 ERA. Still, Kuroda’s outing was not enough to end with a victory for the Dodgers. The game graph is below:


The Pirates took an early lead thanks to a two-run home run off the bat of Jose Bautista, but the Dodgers slowly crawled back into the game. Shockingly, their first run came on the heels of an Andruw Jones groundout, meaning that Andruw Jones did, in fact, do something productive. Kuroda then helped his own cause later in the game with an RBI double to knot the game at two.

In the sixth inning, Xavier Nady hit an RBI double of his own to propel the Pirates to a 3-2 lead. In the next half inning, James Loney and Blake DeWitt delivered RBI singles of their own to put the Dodgers in front for the first time at 4-3. When the ninth inning rolled around, the Dodgers led by the same score, and closer Takashi Saito entered the game. At this juncture, the Pirates had a 14.5% probability of winning the game. Doug Mientciewicz flew out to start the frame, reducing their expectancy to 7.8%. Jose Bautista followed with a single, before Saito struck out Luis Rivas. With two outs, and a runner on first base, the Pirates’ win expectancy was just 7.1%.

Adam LaRoche singled to centerfield, moving pinch-runner Brian Bixler up to second base. With two outs and runners on first and second, their win expectancy rose to just 13.2%. Nate McLouth then stepped up to the plate. He took ball one, but on the second pitch, launched a three-run dinger into orbit. The Pirates took the 6-4 lead, but unlike the previous two plays documented, his was not a walkoff hit. Their win expectancy rose by 79.6%, from 13.2% to 92.8%. In the bottom of the ninth, Matt Capps recorded a 1-2-3 save, capping (pun totally intended) quite the unexpected result in this ballgame.

Print This Post

Eric is an accountant and statistical analyst from Philadelphia. He also covers the Phillies at Phillies Nation and can be found here on Twitter.

Sort by:   newest | oldest | most voted
7 years 9 months ago

Regarding quality starts, I don’t like the old definition of that (7+IP, 3- ER), nor even the modernized one (6+ IP, 3- ER).

I like the PQS methodology that Ron Shandler’s Baseball Forecaster publication introduced and produces research on. It is a Quality Start built on sabermetric principles.

The methodology is very simple and easy to do by hand once you practice it: http://www.baseballhq.com/free/free03.shtml?src=hqf.

Here are the bare bones: 0 to 5 system. One point each for:

1) 6+ IP (automatic 0 for less than 5 IP)
2) IP >= H
3) K > IP – 2
4) K >= 2*BB
5) HR <= 1

Score of 0 or 1 is considered a DISaster start.
Score of 4 or 5 is considered a DOMinating start.

And the score is not dependent on the ER given up, kind of like FIP.

Thus Duke’s outing would be scored a “2” for 6 IP and no HR given up.
Kuroka’s outing would be scored a “3” for 6 IP, more than double K than BB, and one HR given up.

For Kuroda to get a DOM start, he would either have to pitch one more inning with no hits and at most one walk, or strike out four batters in the 7th or alternatively 5 more in the 7th and 8th, without giving up a HR and at most 2 walks.

7 years 9 months ago

What kind of ****ty system gives a 5 2/3 IP, 6 H, 3K, 2 BB, 2 HR outing a 0 while giving a near-identical 6 IP, 6 H, 5K, 2 BB, 1 HR outing a 5? From a linear weights standpoint, there’s only about 1.2 runs separating the two performances and yet one’s a gem and the other an unequivocal bomb. It’s too black and white. Where are the shades of gray?

Imagine calculating QB Rating like that: 1 point if completion percentage ≥ 65%, 1 point if yards per attempt ≥ 8, etc. What a completely useless tool that would be. Speaking of Ron Shandler…