THE BOOK--Playing The Percentages In Baseball

Well, they’re both below Catfish Hunter (32.5) and he’s in the hall of fame.  Hunter’s career ERA?  105.  Same as Morris.  But you sure don’t hear the righteous indignation out of the blogosphere against Catfish now do you?

To say nothing of the far lower values for Sutter or Fingers (granted they’re relievers ... but if this is about career accomplishement, well, they “accomplished” a lot less now didn’t they?)

As for the JAWS conclusion; in what world would anyone say that Morris was a lesser pitcher than Radke??  Morris led the majors in Wins for a decade.  There’s not a single other pitcher who led the league in wins for their decade that isn’t in the hall of fame. 

Lastly; WAR has a tendency to be an accumulator stat.  If you’re mediocre but healthy you end up with a higher WAR than a better, but shorter-lived career.  Pedro Martinez had a career rWAR of 73.5 while Blyleven had a career rWAR of 87.6.  There is no one who with a straight face can tell me that Blyleven was a better pitcher than Pedro Martinez.  I consider Martinez one of the two or three best right handed pitchers of the last 50 years, hands down.  Blyleven was a run-of-the-mill starter who hung around for 20-some seasons and racked up a gazillion Ks and then suddenly, 15 years after he retired people started canonizing him as the great travesty of hall of fame voting.

Point is: no stat is all-encompassing, each has their flaws.  You can’t just look at one stat and declare that player A > player B.

Re: #1

If you aren’t trolling, wow.

Mr. Boss---look up pitchers with similar years pitched (length and time period), look at where Blyleven is in comparison with Ks, shutouts, 1-0 decisions, ERA, ERA+, playoff performance, etc., and come back with that run-of-the-mill.  Of just read Poz’ article on the HOF vote of a week or so ago, where he incinerates the Morris candidacy as it relates to Blyleven.

I don’t know what to say about your reliance on wins, in and of themselves.

That is all.

I think the point about WAR over-rewarding health and longevity is valid.  Goosing the replacement level would alter the Pedro-Bert comparison.

Catfish does actually take his lumps in cyber-space.  But it’s hard to see how he could have been much more successful.  Maybe by pitching planet earth to victory over a team of inter-galactic all-stars or something.  I’m OK with putting him in the Don Drysdale wing.

But you know, like with any counting stat you can check at what rate they accumulated it. 

For example, fWAR has Palmeiro at 74.3 WAR and McGwire at 70.6 WAR which might seem to suggest that Palmeiro was roughly equal to or even a little better than McGwire.  But it took Palmeiro over 12000 plate appearances to accrue that WAR while McGwire needed fewer than 8000 plate appearances to reach that number.  So WAR is not saying that Palmeiro is better than McGwire or that they are roughly equal.  I would say it strongly suggests that McGwire was superior to Palmeiro by a wide margin.

A similar thing happens with Pedro and Blyleven.  Blyleven had about 5000 IP in his career.  Pedro had under 3000 IP.  So once again the overall WARs are close enough to strongly suggest that Pedro was the superior pitcher by a gigantic margin. 

The stats aren’t lying or exaggerating anything.  They are telling you exactly what they are defined to tell you (and if you think they are telling you that Blyleven was just a run-of-the-mill pitcher then you have other issues to deal with before you worry about accurately using them to compare player to player).

You will ALWAYS have a problem when you turn a two-dimensional stat (Wins and losses) into ONE number.

ALWAYS a problem.

This is why I highly prefer to list a pitcher’s WAR-based W/L numbers.

My internet connection is superslow, so maybe someone can post Pedro/Bly by looking at my site at the top of this page that says “Indis W/L”.

Just eyeballing the numbers, so forgive me if I fail to carry a 1 or something. (I also only counted the pitching W-L numbers.)

Pedro: 114-8
Blyleven: 158-68

And just for the hell of it…

Morris: 95-76
Radke: 75-35

There’s not a single other pitcher who led the league in wins for their decade that isn’t in the hall of fame.

Frank Viola, Ron Guidry, Bucky Walters, Paul Derringer, Jim McCormick, Will White, and Tommy Bond have all led the Majors in wins for a 10-year period (Andy Pettitte as well, who might not be a lock for the Hall).

Radke’s a little bit of a weird comp, since he threw 1,400 fewer innings, and Morris’ entire case rests on a counting stat, which basically requires longevity. 

In the past 30 years, 16 guys have thrown 3,500+ innings, including Morris.  Picking out comps from that list is a lot easier, and still makes the same point about how Morris pretty clearly isn’t a HOFer. 

Morris ERA- of 95 puts him in the same category as Frank Tanana (4,200 IP, 240 wins, 94 ERA-, +60.1 WAR), Dennis Martinez (4,000 IP, 245 wins, 95 ERA-, +50.1 WAR), Charlie Hough (3,800 IP, 216 wins, 95 ERA-) and Jamie Moyer (4,020 IP, 267 wins, 97 ERA-, +49.0 WAR).

Jamie Moyer is a fantastic comp.  I had a blog post about it a few months ago.

I’ll take your assessments on performance and replacement as they serve a fine purpose, but I did want to point out that over the entirety of his career Radke pitched in a much more offense-conducive environment.  The American League averaged 4.5 runs per nine during Morris’s career, but that number goes up to 5.0 during Radke’s run.  Note: I used AL R/G and innings per season to properly weight these figures.  Plugging those numbers into your formula, I get 53.6 for Morris and 47.9 for Radke instead of the 42.9 I get for using 4.50 RA9. 

The difference is a handful of wins, but serves to bring them even closer as only 5.7 wins separate the two now.  It’s an interesting comp as Radke is probably a bit underrated in most people’s minds while Morris runs the gambit.

If you use a higher run environment, then you need to use a higher runs-per-win converter.  It’s not exactly a 1:1, but it’s close enough for what we’re doing here.

Hence, no real reason to do what you are doing.

BB-Ref.com has Radke just above Morris in WAR on its list of eligibles for this year. I’m as much a Morris detractor as anyone, but even I’d take Morris over Radke, and I always liked Radke. Heck, BB-Ref.com has Radke as the top-WAR pitcher on the ballot. The gap in career length is just too much for me, and even I have to give some nod to enormous gap in career W-L.

Tanana is maybe a better contrast than anyone, since Tanana had a run at the beginning of his career were he was legitimately a great pitcher: 5 years, WAR 30.4, 82-59, 2.86 ERA, 124 ERA+. Morris’s best 5-year run was 20.1 WAR, 94-54, 3.38 ERA, 120 ERA+. Radke’s best five years was 24.3 WAR, during which he was 71-65, 4.06 ERA, 119 ERA+. Everyone else: Martinez: 22.9 WAR, 68-42, 3.04 ERA, 135 ERA+. Moyer: 18.8 WAR, 75-41, 3.82 ERA, 119 ERA+. Hough: 20.8 WAR, 80-66, 3.57 ERA, 119 ERA+. Only Tanana really sticks out in that, as he had the best WAR AND got the wins AND was a huge K guy (I didn’t show it); his ERA looks better than it is since this was the low scoring late 1970s.

Brief question about the Indis W/L - maybe this is a secret, because I don’t see any explanation anywhere, but how did you calculate those things, Tango?

Suffice to say that if you did:
W - (W+L)/3
you would get WAR

That gets you most of the way there, but I do some little things to make sure everything works out perfectly.

As an example, someone said that:
Pedro: 114-8

So, 122 decisions would put the replacement level at 41 wins.  And 114-41 = +73, which sounds like it’s pretty close to Pedro’s WAR.

You’ll also note that it abbreviates to:
(2*W-L)/3

NO ONE is required to accept the replacement level as .333.  You can make it .300 or .250 or .400 or whatever YOU want.

Hence, the reason I prefer the Individualized W/L record