Do we need to reevaluate the JAWS standard for starting pitchers?

As many of you reading this likely know, Jay Jaffe is the progenitor of the Jaffe WAR Score system, better known as JAWS, the greatest name for an advanced metric ever (I will hear no counterarguments to this). It takes a player’s career Baseball Reference WAR and averages it with combined WAR from his seven best seasons (WAR7), and those seasons do not necessarily have to be consecutive. In other words, it derives a score of a player’s Hall of Fame worthiness based on his peak and longevity.

Finally, it sorts the scores by position so that you can compare apples to apples. It is extremely useful in evaluating careers, but even Jaffe himself will tell you that it is just a starting point. It does not factor in playoff performance, awards, historical significance, etc. Nobody is saying that JAWS or WAR is the the be-all and end-all for Hall of Fame decisions.

If you were to go to a position’s JAWS page, you would find a list of all players at that position sorted by JAWS with a break at the average JAWS score. It shows the average WAR and WAR7 scores, too, and you can sort by those numbers if you would like, but the default presentation is by JAWS.

Let’s take a look at the JAWS scores for starting pitchers. The first thing that might stand out to you is that there are A LOT of starters in the Hall of Fame. That is because there are more starting pitchers than any other position. Throughout history, teams have had a minimum of four or five starting pitchers, but obviously they do not have that many for each and every other position.

Projecting the 2019 Baseball Hall of Fame vote

Photo by Otto Greule Jr/Getty Images

Thanks to Ryan Thibodaux’s tracker, we can use statistical probability methods to project the final vote totals for our favorite players.

There are 72 starting pitchers in the Hall of Fame. Shortstops and right fielders are at a distant second at 25. There are more than 32 pitchers above the JAWS average alone, though a few of them are not in the Hall for no good reason. Because there are so many starters in the Hall, I believe that a starting pitcher should not be inducted unless he really raises the bar. Now I don’t know how popular that opinion is, but as we will soon see, it raises a problem with JAWS for pitchers nonetheless.

Clayton Kershaw is considered by many to be a Hall of Famer already. If he retired right now, he would very likely get in. He might not get in on the first chance because of those who overvalue longevity, but he would get in. I am not going to write about why here — that would be an entire article — but you are free to check out his B-Ref page. I bring this up because despite the fact that Kershaw was the best pitcher in baseball during much of his career, his JAWS score is underwhelming. His 57.1 JAWS ranks just 44th all time, and well below the JAWS average of 61.8. His career WAR of 64.6 ranks 46th, and his 49.6 WAR7 ranks 45th.

(JAWS counts a pitcher’s hitting. That is why their WAR on the JAWS page might not line up with the pitching WAR on their individual player pages.)

Of course, Kershaw is still going, so he still has time to boost his case, at least for those who care more about how long you play than how well you play. However, one person for whom you can’t say that is the late Roy Halladay. He is on the ballot for the first time, and he appears to be a lock to get in. Devan Fink did a great job detailing why. As Fink detailed in that article, Halladay’s 57.5 JAWS ranks below the JAWS average, and coincidentally just ahead of Kershaw. He is tied with Hall of Famer Juan Marichal. Also as with Kershaw, his WAR7 is roughly at the Hall average, too.

The problem here is one that you might have heard before: innings pitched. WAR is, in effect, a cumulative stat. The more you play, the higher it goes up, as long as the player is performing above replacement level. Halladay pitched 2,749.1 innings. Only two pitchers above the JAWS average pitched fewer than 3,000 innings: Ed Walsh (2,964.1 IP) who pitched during the dead-ball era (pre-1920), and Pedro Martínez (2,827.1 IP) who had one of the strongest peaks of the live-ball era. The pitchers in the Hall of Fame with fewer IP than Halladay were dubious choices. When Halladay gets in, his IP total will rank only 61st all time among Hall of Fame pitchers, according to the Play Index. Kershaw is currently at 2,096.1, which would rank him above Dizzy Dean, a poor inclusion for the Hall, and right below Hoyt Wilhelm. He was primarily a reliever!

It is not exactly cutting edge analysis to state that starting pitchers are seeing fewer and fewer innings, but it makes evaluating Hall cases for starting pitchers more and more difficult. Looking back at the list of Hall of Famers by IP, a lot of the high innings totals belong to pitchers from the dead-ball era. Cy Young (7,356), Pud Galvin (6,003.1), and Walter Johnson (5,914.1) pitched entirely or mostly during that time.

So that gave me an idea: What if we adjust the JAWS average by filtering out pitchers from the dead-ball era? Not only will this help to get rid of those high innings totals, but it also helps with something else. The dead-ball era was a different game, much more so than today’s game differs from previous eras. WAR can adjust for a lot, but not for how different dead-ball era baseball was.

So I took the JAWS list for starting pitchers and filtered out anybody who pitched in the dead-ball era. This was really hard for me because Baseball Reference does not have an option to export to Google Sheets, and I don’t have Excel, so I had to copy and paste all the data individually. Yep, I am just that dedicated. Some pitchers obviously spanned both eras, so I had to make a judgement call. If the pitcher had his best years before 1920, such as Walter Johnson, Pete Alexander, and Stan Coveleski, I filtered him out. (I promise that Stan Coveleski was a real player and that I did not just make him up.) So here are the grand results:

Yeah, that was useless. Only Charlie Buffinton and Al Spalding would move above the JAWS average, and ironically they pitched during the dead-ball era. Part of the problem is that there are quite a few members of the Hall of Fame from that era who really drag down the averages. Just for fun, I repeated the same exercise with the additions of Curt Schilling, Mike Mussina, and Roger Clemens. The numbers were even closer to the original averages, mostly because of Clemens.

I doubt declining innings totals are going to have major impacts on future considerations for Hall of Fame pitchers. Some kind of adjustment would be helpful, but what I tried to do here clearly isn’t it. However, as the gap between starters and relievers shrinks, as well as starters’ innings totals, something is going to need to be figured out. At least we have plenty of time until it becomes a real problem.