clock menu more-arrow no yes

Filed under:

By the numbers: David Bote, Daniel Descalso and the truth about clutch hitting

CHICAGO, ILLINOIS - APRIL 19: Daniel Descalso #3 of the Chicago Cubs celebrates after he scored on the triple by Anthony Rizzo #44 of the Chicago Cubs during the fourth inning of a game against the Arizona Diamondbacks at Wrigley Field on April 19, 2019 in Chicago, Illinois. (Photo by Nuccio DiNuzzo/Getty Images) ORG XMIT: 775308763

Baseball analysts don’t deny the existence of clutch hitting. David Bote’s game-winning RBI single in the Cubs’ 2-1 victory Sunday against the Diamondbacks was clutch. When Daniel Descalso singled in the seventh to drive in the first run in a 2-0 victory April 11 against the Pirates, that was clutch, too.

But while clutch hitting is real, clutch hitters have proved elusive. No real pattern has been found in terms of who comes through and who falters when the game is on the line, and there is no clutch-hitting component in WAR.

Fangraphs.com does have a stat labeled ‘‘Clutch‘’ on the win-probability portion of each player and leader page.

It’s based on win probability added, which measures how much each plate appearance adds or subtracts from the chance of winning, along with a leverage index that measures the importance of each game situation.

Most Clutch season scores fall between 1 and minus-1. Descalso and Bote both fared well in 2018. Descalso was third in the National League and fifth in the majors at 1.69 for the Diamondbacks, and Bote, in a partial season for the Cubs, was eighth in the NL and 20th in the majors at 1.35.

The Clutch formula is (WPA/pLI) – WPA/LI. WPA is win probability added, LI is the leverage index and pLI is the average leverage index for all game events. WPA/LI is computed for each plate appearance and totaled to yield a context-neutral value, while WPA/pLI uses a player’s average index.

Fangraphs says subtracting WPA/LI from WPA/pLI tells us ‘‘how much better or worse a player does in high-leverage situations than he would have done in a context-neutral environment.’’

In other words, the Clutch stat tells us whether a player performs better than his norm in high-leverage situations.

Descalso had a 3.08 WPA last season, meaning his plate appearances added just more than three wins to the Diamondbacks’ win expectancy.

Some players hit more often in high-leverage situations and get more opportunities to build WPA. WPA/LI, which normalizes WPA to opportunity, left Descalso 12th in the NL at 1.43. A WPA/pLI of 3.12 left him with his lofty 1.69 Clutch score.

A couple of cautions: Clutch compares a player to himself. A weak hitter can rank high in Clutch with mediocre performance in important situations, and a good hitter can get a negative score if he doesn’t reach his own standards at high-leverage time.

RELATED

Morrissey: Yelich’s start makes Bryant’s that much tougher to watch

Cubs just haven’t been the same since Dodgers beat them in 2017 NLCS

Last season, Mike Trout had a 4.61 WPA/pLI and a 6.91 WPA/LI. Both figures are higher than Descalso’s, and anyone would be thrilled to have him in a high-leverage plate appearance. But with such a high WPA/LI, his clutch was minus-2.30.

The other caution is that Clutch isn’t predictive. All players with lengthy careers have seasons on the negative side, and they don’t follow any particular pattern.

Clutch has value in describing a season, but it isn’t an indicator of overall value or of future performance. Like other attempts at measuring situational performance, it hasn’t been shown to enhance the likes of WAR.