Carlos Correa’s memorable night in the field (a run-stopping throw home) and at the plate (a walk-off home run) was magnificent —
From a run-saving throw in the 6th to the walk-off homer in extras, Carlos Correa was CLUTCH in Game 2 pic.twitter.com/etn0VVcLFy— SportsCenter (@SportsCenter) October 14, 2019
— but it also came in Game 2 of the Championship Series. Important. But not most important . . . right? The Baseball Gauge doesn’t think so.
That site has a running Win Probability Added leaderboard, and even after last night’s heroics, Correa’s walk-off (his defensive play is not included) only gave him .200 WPA. Better said as: “Carlos Correa increased his team’s chances of winning the game by 20%.”
But the math can also be compared to historical data and computed with context via cWPA (Championship Win Probability Added): “Measures how much a player increased or decreased their team’s championship win probability. In that case, Correa’s walk-off win that evened the ALCS against the Yankees to 1-1 gave the Astros a 2.9% boost in winning the World Series. That’s not insignificant, but it’s not the best.
We already knew that Madison Bumgarner’s World Series relief appearance in 2014 was the clutchiest, ballsiest, bestest postseason performance in recorded human history, but the snooty math geeks who’ve hijacked baseball agree.
For the 2014 World Series, Bumgarner alone had a 1.2 WPA. Which means that combined for the series, he gave the Giants a 120% chance to win in his three appearances. He alone was worth .869 cWPA in 2014 — Madison Bumgarner increased the Giants’ overall win probability in the series by 87%. That’s just 2014!
If we go back to the beginning of the century, no player has done more to boost his team’s odds of winning the World Series than Madison Bumgarner. His 1.154 cWPA is 46% greater than the next guy — Lance Berkman (.694). On just raw WPA, his 2.0 leads all players, too.
That doesn’t mean the Giants’ probability of winning a World Series game was 200% when Madison Bumgarner was in it. WPA is a cumulative stat. In any given moment, the Giants have a certain probability of winning the game, and how Bumgarner — or any Giant — performed, positively or negatively, added to the value. There are a lot of discrete events in a given game, and one play alone doesn’t necessarily carry tremendous WPA value. That’s what the aLI (average Leverage Index) number is for. Leverage index measures the importance of a single play. “1” is average. “2” is twice as important, etc.
But since the turn of the century, no pitcher has done more to help his team win a championship than Madison Bumgarner.
There are some surprising Giants on this list, too. Jeremy Affeldt probably doesn’t get enough credit for his role in the championship era. There’s J.T. Snow’s 3-run home run and getting thrown out at home plate to end the 2003 NLDs. Barry Bonds’ 2002 World Series is in there as well.
The top 25 of WPA heroes for the Giants since 2000 is a fairly obvious list:
I didn’t expect to see Edgardo Alfonzo, Matt Herges, and Ty Blach on this. It’s the bottom 25 who surprise a little bit:
I fear the Belt Wars will only get worse from here. Belt’s -0.5 translates to a -0.061 cWPA, meaning his collective performance in the postseason — despite literally winning the 18-inning game — lowered the Giants’ probability of winning their championships by 6%.
But if you really want to get into it: Buster Posey just missed the cut with -0.1 WPA. That’s with an extra year of postseason play over Belt. And when you translate his -0.1 WPA and extra postseason run, it translates to a -0.082 cWPA. That means his collective performance in the postseason — despite literally this happening
— lowered the Giants’ probability of winning their championships by 8%.
Does that mean Buster Posey and Brandon Belt aren’t “Clutch”? Ultimately, who cares? Clutch is not a skill. Clutch is a result. Oh, it’s a result with value — we can say that a performance was clutch — but it’s not predictive, and that’s why hunting for clutch is a pointless endeavor. But it’s a great way to describe what did happen.
Madison Bumgarner gave up a three-run home run to Jake Arrieta in the 2016 NLDS, after all. But when we talk about baseball players, we usually talk about what they’ve done. What a player has done is the one thing none of the fans or owners can take away, after all. What players have done is what other players get compared to, and while nobody’s comparing Carlos Correa’s night to Madison Bumgarner’s 2014 World Series run, at some point, a postseason ace will emerge and he will get the comp.
Sabermetrician Tom Tango (who wrote The Book on analytics) got stuck on the legend of Clayton Kershaw as it relates to his postseason performance, and in a scenario that appeared to follow this Wendy Thurm tweet —
Maybe if Kershaw wasn't as good or wasn't as nice, he'd be Madison Bumgarner https://t.co/6JUH8Ct4Lw— Tangotiger (@tangotiger) October 10, 2019
He set out to compare Bumgarner and Kershaw:
Bumgarner is a bit over 1 year younger than Kershaw. If you line up their careers by age, Kershaw has been 2.4 to 4.7 WAR higher than Bumgarner in each year from age 21 through 29. This year at age 30, Bumgarner was 0.8 WAR behind Kershaw in 2018.— Tangotiger (@tangotiger) October 12, 2019
The above question encapsulates everything that you believe in Clutch Pitching. How you answer this question tells you how much you think that the clutch-gene is actionable— Tangotiger (@tangotiger) October 12, 2019
So, there’s a poll where 538 people say they’d rather start Clayton Kershaw in a World Series game today over Madison Bumgarner. That probably makes sense! Even with Kershaw’s diminished velocity, he still allowed a well below league average hard hit rate (33%) to Bumgarner’s top-10 worst 41.5%.
On the other hand, Clayton Kershaw has been spectacularly awful in the postseason, to the point that Tango went ahead and tried to weight postseason performance a bit more to see how the two would compare to see if there’s a better answer than just “feeling”:
I'd give the playoffs a certain amount of weight relative to the regular season. I don't know if it's 2X or 5X or 10X, but it's more than 1X. Let's say it's 5X. So, we end up with: pic.twitter.com/EqiLD2JtD3— Tangotiger (@tangotiger) October 12, 2019
But that’s looking at their entire careers and attempting to arrive at a number that does fall into the realm of predictability. In this case, the number (weighted runs allowed per 9 innings), still suggests that probability-wise, Clayton Kershaw is the safer bet by virtue of it being a push (again, consider this year’s hard hit rates).
And yet, the same system that gives us these calculations also gives us an actual, literal Clutch stat, simply defined by FanGraphs as:
Clutch measures how well a player performed in high leverage situations. It’s calculated as such:
In the words of David Appelman, this calculation measures, “…how much better or worse a player does in high leverage situations than he would have done in a context neutral environment.” It also compares a player against himself, so a player who hits .300 in high leverage situations when he’s an overall .300 hitter is not considered clutch.
Here’s where I’ll thread one final needle to prove the point so that you can argue with your A’s and Dodgers fan friends about Madison Bumgarner’s historic contribution to the Giants and the sport of baseball.
Mariano Rivera comes in as the player with the most Win Probability Added (+11.5) and cWPA (1.792), but by that Clutch stat — which, again, only measures what happened — Madison Bumgarner is +0.02 versus Rivera’s -0.01. I don’t know if that can all be attributed to 2014 Game 7 versus 2001 Game 7, but figure that because Rivera pitched in more high leverage situations as closer that he would have a much smaller margin of error.
In Bumgarner’s case, he succeeded in the average leverage situations and outperformed the greater leverage situations. His performance was just that extra 2%. Since 1903, that’s the top of all postseason baseball, but tied with Yadier Molina . . . seriously.
However! Yadier Molina’s postseason career doesn’t stand out beyond this one slim category. He has a -0.3 WPA for a his career, and a +0.475 cWPA. But for crushing the Mets in 2006 (.348/.423/.652 — 0.52 WPA for the series) and the Rangers in 2011 (.333/.414/.417 — +0.28 WPA) in high leverage spots, he gained the clutch numbers.
Bumgarner’s raw 2.3 WPA is just 17th all-time, but just the second-fewest playoff years (4) in the top 25, behind Eric Hosmer (just 2). Hosmer has as much WPA added (2.3) in five fewer games than Bumgarner, but just a +0.261 cWPA. Bumgarner’s +1.230 cWPA is second all-time behind Mariano Rivera, but if we just want to boil it down to Clutch, then Bumgarner (+.02) tops Rivera (-0.01).
The math explanation is that Mariano Rivera was so good that, eventually, the numbers he had amassed started working against him when he wasn’t quite as dominant. For Madison Bumgarner, it means he was outstanding when he had to be and pitched better than he had before in similar situations. That’s clutch. And that’s why he was the best.