JAWS and the 2020 bWAR Update, Part 3

No pitcher took it in the JAWS quite as hard as position players Ernie Lombardi and Josh Donaldson did via Baseball-Reference’s latest update to its version of WAR, which I’ve spent the better part of the past two weeks unpacking — at least when I wasn’t stocking my freezer and my pantry while reading the grim COVID-19 news. B-Ref’s latest influx of data resulted in alterations to five different areas of the metric that affected players as far back as 1904 and as recently as last season. Lombardi, a Hall of Fame catcher, lost a whopping 7.3 WAR due to the introduction of detailed play-by-play baserunning and caught stealing data from the 1930s and ’40s, while Donaldson lost 3.8 WAR due to a change in the way Defensive Runs Saved is calculated. By comparison, the largest swing for a pitcher, either positive or negative, was the 2.2 WAR gained by Hall of Famer Christy Mathewson.

B-Ref’s version of WAR is different from that of FanGraphs, particularly when it comes to pitching; it’s based on actual runs allowed, with adjustments for the qualify of the offenses faced and the defenses behind it, where the FanGraphs version is driven by the Fielding Independent Pitching categories as well as infield flies. As bWAR is the currency for JAWS, it’s of particular interest to me, even at a time when the Hall itself is closed due to the pandemic. I’ve grazed by the pitchers in my two recent updates, mentioning a few tidbits here and there while trying to avoid a typical Jaffe-length 3,000 word epic, but in this installment I’ll take a closer look at the those most affected. To review, here are the five areas where B-Ref’s WAR update has incorporated new (or recently unearthed) data, ordered for chronological effect:

• New Retrosheet Game Logs (1904-07)
• Caught Stealing Totals from Game Logs (1926-40)
• Baserunning and Double Plays from play-by-play data (1931-47)
• Defensive Runs Saved changes (2013-19)
• Park factor changes (2018)

So the big thing for history buffs, as the site itself noted last month, is the addition of four years worth of box scores that account for every game of the careers of Hall of Famers Ty Cobb and Walter Johnson. The latter’s major league debut, on August 2, 1907, happened to be against Cobb’s Tigers. B-Ref’s play-by-play data doesn’t go back quite so far (the earliest boundary is now 1918, though it’s incomplete), so it’s not apparent via the aforementioned link, but it turns out that the first hit Johnson surrendered was to Cobb, who was batting cleanup that day. It was one of six hits Detroit rapped out in the Big Train’s eight innings. Cobb, just 20 years old but en route to his first of 11 batting titles, came away quite impressed. In the aftermath of the game, he said, “We couldn’t touch him … every one of us knew we’d met the most powerful arm ever turned loose in a ball park.”

Years later, Cobb recalled, “The first time I faced him, I watched him take that easy windup. And then something went past me that made me flinch. The thing just hissed with danger,” which, damn, that’s just about the coolest description of anything on a baseball diamond, ever.

But I digress. The addition of the 1904-07 games allows for better estimates for the quality of opponent each pitcher faced in the newly-covered seasons. Via Alex Bonilla’s example in the blog entry linked above, we now know that Mathewson did a greater proportion of his 1907 work against the NL’s better offensive teams:

“Prior to this change, we used the league average (excluding his team) of 3.36 runs per nine innings as the expected quality of his opposition. However, with game-level data, we can see that Mathewson’s actual opponents averaged 3.55 runs per nine innings, showing that Mathewson was probably used strategically and started more games against better opponents.”

Via that adjustment, Mathewson’s 1907 WAR increased by 0.9, from 7.7 to 8.6, with his margin over the second-ranked total by one Ed Karger expanding from 1.0 WAR to 2.2. Matty’s gain is the largest single-season change in either direction for any pitcher in this update, and as noted, the 2.2-win bump his career pitching WAR received was also the largest. But where I came up with 11 Hall of Fame position players whose WARs swung by at least 2.5 in ether direction in this series’ first installment, only two enshrined pitchers moved by at least a full win, and by the time you get to the 10th-largest change from among that group, you’d barely notice:

Mathewson’s JAWS increased by 1.7, from 86.2 to 87.9, moving him past Lefty Grove and into the number six spot in the starting pitcher rankings. Interestingly enough, nine of the 10 pitchers above increased in value, likely because as with Mathewson, they were used more often against better teams and thus gained a bit of extra credit in adjusting for quality of opponent. Brown was the only pitcher besides Mathewson whose JAWS moved by a full point, which means that all of his gains were consolidated among his seven best seasons; Joss (0.7) was close. As I noted in the second installment, the career WAR standard for Hall pitchers dropped by 0.2 (from 73.5 to 73.3) but the peak standard rose by 0.1 (from 49.9 to an even 50.0), the only position where the two standards moved in opposite directions.

One of the largest JAWS drops is that of Wes Ferrell, a candidate on the 2013 and ’16 Pre-Integration Era Committee ballots who figures to have a shot at landing on the 2021 Early Baseball Era Committee ballot as well. A better player than his enshrined older brother, catcher Rick Ferrell, Wes was a righty who starred for the Indians (1927-33) and Red Sox (1934-37) before arm troubles turned the remainder of his career, which ran through 1941 (his age-33 season), into a slog. At his best, amid an historically high-scoring era, he was regarded as the equal of Grove, though the advanced stats solidify his place as the Junior Circuit’s true second banana. From 1930-35 he finished second to Grove in pitching WAR four times, and for the 1929-36 period, his 49.2 WAR ranked second among AL pitchers, albeit well behind Grove’s 66.1.

Ferrell made up some of that ground with his bat, as he was a better hitter than his brother, with a career line of .280/.351/.446 (100 OPS+) and 38 homers (Rick hit .281/.378/.363 for a 95 OPS+ with 28 homers). It’s the estimate of his offensive value that suffered via this WAR update; he lost seven runs in the double play avoidance category, dropping his batting and baserunning contribution from 11.8 WAR to 11.1 (9.3 of which came during the 1929-36 stretch). His career/peak/JAWS line — which includes all of his offensive contributions — dipped from 60.7/54.7/57.7 to 60.0/54.3/57.1. His JAWS ranking fell from 41st to 47th, still within one point of a very respectable group of enshrinees (Jim Palmer, Carl Hubbell, Hal Newhouser, Bob Feller, Roy Halladay, Juan Marichal, and Willis). His peak score’s ranking fell from 25th to 28th, but he’s still over four wins above that standard. As I said when I examined his candidacy at greater length in 2016, I remain open-minded about his inclusion, as I’m loathe to dismiss a high-peak/short-career player who merely deprived us of an undignified denouement. I expect we’ll have another chance to examine his candidacy in November.

Leaving the ancient pitchers behind to focus on the contemporary ones, their values were affected both by the adjustment of 2018 park factors, which use three years of data (2017-19, in that case), and by the update to team Defensive Runs Saved totals, because one element of their pitching WAR uses an estimate of the quality of the defense behind them. As Bonilla explained:

Team total DRS changed by as much as 46 runs for a given team and season – the 2019 Dodgers defense improved from 75 DRS to 121 DRS by non-pitchers under the new system. Once applied to a specific pitcher, however, the changes to WAR are much smaller in magnitude than the changes to individual fielders. The most extreme example is Hyun-Jin Ryu, who pitched 182.2 innings in front of the 2019 Dodgers defense. Considering the Dodgers defense to be 46 runs better across the entire season, and considering that Ryu was the pitcher for 13.52% of the Dodgers’ balls in play in 2019, we adjust our expected runs allowed for Ryu by 6.2 runs for the season. After following the rest of the steps in our pitching WAR calculation, the end result is a drop of 0.3 WAR for the season. All other changes to pitching WAR from this change to team defense are smaller than Ryu’s 0.3 WAR drop in 2019.

Between those two factors, here’s the list of the recent pitchers whose WARs swung by the largest amounts in either the positive or negative direction:

I used a 10-WAR cutoff to clear out some clutter; seven other players had swings of at least 0.6 in either direction, including José Ureña, whose gain of 0.9 WAR basically doubled his career total. As you might figure, several constellations of teammates, such as the Indians’ Bauer, Carrasco, and Kluber, and the Royals’ Duffy, Kennedy, Guthrie, Herrera (and some guys who didn’t make the 10-WAR cut), are among the most affected.

As for the pitchers relevant to a Hall of Fame discussion, here are the active leaders in JAWS, and how their scores changed:

Until the final month of the season, Greinke stood as the active leader among pitchers in JAWS, but he was overtaken by his new teammate, who allowed just 11 runs in his last 53.1 innings, with a 72-to-8 strikeout-to-walk ratio. Where the pair were just a hair apart at the time, they’ve now got a bit of daylight between them. Further down, Sabathia, who retired at season’s end, has taken multiple hits in B-Ref’s updates; in 2018, his 2011 WAR was revised downward from a career-high 7.5 to 6.4 — a double whammy because that season was part of his peak score.

Verlander aside, most of these pitchers lost ground, and from the standpoint of building their cases for Cooperstown, none of them will be helped by a shortened (or, gulp, completely canceled) 2020 season, though this does hit some pitchers harder than others. Given his total of 2,622 strikeouts and an average of 200 for the past three seasons, Greinke was on course to reach 3,000 in late 2021 (incidentally, his final year on his six-year, $206.5 million deal), but now he’ll more likely have to chase down the milestone in 2022. Scherzer (2,692 strikeouts) had an outside shot of reaching the milestone this year — he would have needed to exceed his previous career high, set in 2018, by eight — but now he’s looking at mid-2021. Not a huge difference for either pitcher, most likely, though we’ve all seen outstanding hurlers on the wrong side of 35 show up one spring having lost their mojo, sometimes for good; Palmer, Tom Seaver, Pedro Martinez, and Roy Halladay come to mind.

This concludes our tour through the bWAR update, a series that probably wouldn’t have been possible without the delay to the start of the 2020 season, but one that represents the first of several Hall of Fame-related projects I’ll keep busy with while waiting — and hoping — for Opening Day. In the meantime, hang in there and stay safe, folks.