Best Pitching Performances #5-#1

This morning, our topic of discussion involved the bottom half of the top ten pitching performances of 2008, as determined by single-game WPA. As mentioned then, no pitcher accrued an individual game WPA above +1.0 this season, but there were still some absolutely fantastic outings. For posterity’s sake, numbers ten through six were:

10) Bronson Arroyo, 8/26 @ Hou:   0.660 WPA, 9 IP, 5 H, 1 ER, 2 BB, 3 K
9)  Roy Oswalt,     9/6 @ Col:    0.676 WPA, 9 IP, 1 H, 0 ER, 2 BB, 6 K
8)  James Shields,  5/9 vs. LAA:  0.685 WPA, 9 IP, 1 H, 0 ER, 0 BB, 8 K
7)  Jeff Karstens,  8/6 @ Ari:    0.695 WPA, 9 IP, 2 H, 0 ER, 2 BB, 4 K
6)  Matt Cain,      7/24 vs. Was: 0.707 WPA, 9 IP, 4 H, 0 ER, 0 BB, 4 K

And here are the top five:

#5: Cliff Lee, 5/12 vs. Toronto
Cliff Lee had a remarkable season. After having to fight for his job in Spring Training, Lee went onto win the 2008 AL Cy Young Award. Without attempting to stir any discussions about Lee/Halladay, his top performance of the year, via WPA, occurred on May 12, against Doc’s team. Lee pitched a complete game shutout, scattering seven hits and two walks over nine innings, striking out five in the process. At the end of the day, Lee was putting the finishing touches on an incredible streak to start the season, earning a 0.715 WPA.

#4: Josh Banks, 5/25 vs. Cincinnati
Jeff Karstens seemed a tad out of place on this list, but at least there are plenty of people who have heard of him. Banks, however, is not well-known, and did not have a very solid 2008 season, yet he somehow managed to harness everything he has into the fourth best performance of the season. For those that do not remember, this 5/25 Padres/Reds game was the one that went 18 innings. Banks pitched six fantastic relief innings, surrendering five hits and no runs to go with two walks and four strikeouts. His work earned him a 0.718 WPA.

#3: Jesse Carlson, 4/16 vs. Texas
Keeping with the theme of relievers earning high WPA marks, Jesse Carlson of the Blue Jays found himself in quite the predicament against the Rangers early in the season. BJ Ryan had blown the save in the ninth inning, and the Rangers were again threatening in the tenth. Brian Wolfe allowed the first three batters to reach base safely, and was lifted in favor of Carlson. Jesse entered into a bases loaded, no outs, situation, and managed to get out of it, recording a “Houdini” in the process. He would pitch two more scoreless innings, limiting the baserunners to one hit and two walks, while striking out four. The Rangers would win the game, but Carlson recorded a 0.721 WPA for his stellar work.

#2: Ben Sheets, 9/6 vs. San Diego
Ben Sheets has always been the guy with the ridiculous “stuff” and potential to be fantastic if he could stay healthy. We got to see a lot of him this year, and he didn’t disappoint, but none of his games were better than the one on September 6. Against the Padres, home at Miller Park, Sheets tossed a five-hit, complete game shutout, walking one and fanning seven. His 0.729 WPA for the day placed him second on our list, though quite the distant second behind the best performance of the season.

#1: CC Sabathia, 6/10 vs. Minnesota
How could a list like this not have Sabathia? An eventual teammate of second-place Sheets, Sabathia’s game on June 10 actually occurred before he was sent to Milwaukee. Back when he was a member of the Indians, CC tossed a five-hit, complete game shutout, with no walks and five strikeouts. His single-game WPA, the best of any game for a pitcher this year, was 0.775, significantly better than everyone else on this list.

Nobody may have produced single-game WPAs above +1.0, but it is tough to imagine, after seeing these games, what someone would have to do to accomplish such a feat.

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.

12 Responses to “Best Pitching Performances #5-#1”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Joe says:

    Although I’m fairly clued up on most advanced metris – I’m not fully knowledgable on the exact workings of WPA. I know its a score based on individual events within a game but maybe it would be helpful to explain how the order of these performances work out in a basic sense as opposed to a statistical one? For example there were two no-hitters this year and neither are mentioned here. Also there doesn’t seem to be a huge amount of correlation between baserunners allowed and position on this list. I can see how the relief efforts are different cases but to an untrained eye the Shields game looks like the best performance, without wishing to seem like I’m trying to judge performance just on box scores I’m must be missing something?

    Vote -1 Vote +1

  2. Eric Seidman says:

    Joe, the reason that some of the games seem better than others is that WPA is not WPA/LI, in the sense that WPA counts certain situations as being worth more than others. For instance, a two-run double in the ninth inning has the potential to be worth loads more than a two-run double in the first inning. Pitchers create their own situations, so keeping the situational importance in the metric is a better way of judging them than to remove it, which is what WPA/LI does.

    Jon Lester’s no-hitter had a WPA of 0.252, not even close to this list, primarily because the game was never close. The Red Sox won 7-0, and after a while, the game was vastly out of reach. There may not be a correlation between baserunners allowed and position on the list, but WHEN these baserunners are allowed, and how the pitcher got out of the situation is of tantamount importance. Via game score, this order might shift, but in terms of adding wins to your team, via shifts in win expectancy, this is the order.

    Vote -1 Vote +1

  3. qqqqqqqq says:

    Does WPA ever consider pennant runs? I’m wondering because Johan Santana pitched a game on the second to last day of the season where the Mets HAD to win it to stay alive. And he pitched flawlessly for a WPA of 0.564. Was the pennant race considered in the calculation of the WPA?

    Vote -1 Vote +1

  4. Isaac says:

    Does WPA give an edge to pitchers that put themselves in sticky situations only to get themselves out of those said situations? If so thats a bit ridiculous. The fact that Zambrano and Lester missed this list because they were more than flawless is very, very strange. Am I wrong in that assumption?

    Vote -1 Vote +1

  5. SleepyCA says:

    I think the answer is, pitch a lot of innings in a 0-0 ball game and don’t get pulled in the middle of an inning with men on base. Alternatively, I guess, a reliever could come into a situation such as a tied game with the bases loaded and no outs in the bottom of the ninth, get out of it, then pitch “x” shutout innings.

    I don’t know where you’d go to search single game WPA’s, but using the first set of criteria, I’ve found a couple of historical >1.0 WPA games. frank tanana threw 13 shutout innings on september 22, 1975, and ended up with a WPA of 1.132, and Tommy John had a game in 1983 where he threw 13 shutout innings and came away with 1.217 WPA. I only checked two of the games on that PI query, and both succeeded, so I’m sure a number of other pitchers have done it, as well.

    OTOH, I’d bet this is something we never again see, at least by the first set of criteria. Mark Mulder’s 2005 10IP CGSHO was only .832 WPA, and that is about the limit i see any modern pitcher throwing.

    (hope those links worked ok- i hate not having a preview button here)

    Vote -1 Vote +1

  6. SleepyCA says:

    Also hate that the links aren’t underlined. They are there, though, just hover the mouse over the text until you find them if you are interested.

    Vote -1 Vote +1

  7. PaulDeReno says:

    I would say that a list like this is near worthless.

    If Tim Linsecum were to begin pitching the 3rd with the score tied at 0, walked the bases loaded, and then gave up a line-drive DP with a forceout at home, and a 395 foot fly ball that stays in the yard in dead center, the effect on WPA is the same as if he struck out the side in 9 pitches. It goes from the probability of winning with the score tied when you’re on the field to that same situation in your next turn at bat. Regardless of what went on that inning, same result on the scoreboard = same WPA.

    Of course, it’s not totally meaningless, because if a pitcher is bad (results-wise), he loses WPA and has to get bailed out by the offense. Of course, if he’s good, the offense has to stink it up for him to register a gaudy WPA. Assume a scoreless tie. If a pitcher pitches a 1-2-3 inning, his team has an extra turn at bat, so their chances of winning are at that point slightly higher. If the offense blows it, the two teams return to equal footing, and the pitcher has to regain the lost WPA advantage of the extra inning. Repeat the process 8 times, and you have the maximum WPA any pitcher can have in a 9-inning game regardless of who wins provided the starter does not give up the winning run.

    Vote -1 Vote +1

  8. SleepyCA says:

    Here’s an interesting one: Lindy McDaniel earned a win and 1.08 WPA pitching 13 innings in relief in 1973. Nice.

    Vote -1 Vote +1

  9. PaulDeReno says:

    Thus, the highest possible WPA for a pitcher in a 9 inning game is a 1-0 complete game shutout with the pitcher on the visiting team and when the pitcher’s team scores in the top of the 9th. The pitcher will gain time increments that his team gives up again until the 9th inning, when it finally capitalizes. Yet, the other team in the 9th will have a fighting chance, and the pitcher would have to finish them off. If the pitcher were on the home team, he would not get that crucial “finish him” component; the batter would have taken care of that for him. Also, if his team scored before the 9th, his contribution in WPA would be proportionally smaller, as the batter would have consumed much of the pitcher’s time credits. In a tie game, the WPA returns to around .5 at the end of every inning, because they each have the same (infinite) opportunities to score. In a 1-0 game, the leading team creeps up in WPA at the end of every inning because the 27-out barrier helps the leading team.

    Vote -1 Vote +1

  10. TangoTiger says:

    Paul captures it perfectly here. WPA captures the performance relative to the timing. Every time the pitcher can get out of an inning not allowing a run, then he keeps gaining WPA, starting with 0.05 wins in the first inning, and going up in the ninth inning. The amount gained is directly proportional to the Leverage Index at the start of his inning. Thus, he gains the most when the score is close. I would not say necessarily that it’s when it’s a 0-0 game. It might be a 1-0 game. I’m not sure.

    WPA/LI (aka Situational Wins) is perfect for hitters, since it values each PA as “1”, which is really what we want to do, to make a fair comparison.

    However, what we *really* want is something in-between WPA/LI and WPA for a pitcher. A pitcher doesn’t just come into a PA like a batter, but he creates the environment. And so, each PA must be worth more the more runners are on base. We really need to multiply the WPA/LI by the LI for the base/out state. A bases empty situation has an LI of 0.7, while one with the bases loaded is say 3.0 times. That’s what you really want to do: WPA/LI * boLI.

    For a season, a pitcher’s boLI will come in pretty close to 1. Even if it doesn’t, the boLI of all the top pitchers will be close to each other so that they will all be similarly biased.

    For a single game however, it might make a difference.

    Vote -1 Vote +1

  11. TangoTiger says:

    Click on my name for a couple of relevant links. It’s not a breezy read, so be forewarned.

    Vote -1 Vote +1

  12. A.M. says:

    How did Jack Morris’ game 7 victory in 1991 rate?

    Vote -1 Vote +1