Last Year’s Hall of Fame Ballot and Voting, Compared with Wins Above Replacement (WAR)

My very first FanShot... wanted to see how this looked. Turns out (as anticipated), there's love for Raines and Trammell and a big fat "meh" for Jim Rice. Gotta love the holistic view of performance WAR affords.

about 3 years ago adarowski 19 comments 1 recs |

Story-email Email Printer Print

Comments

I enjoyed reading that, Adam.

One thing to think about is that career WAR isn’t the final word. Five 8 WAR seasons are, to me, more Hall worthy than ten 4 WAR seasons. Guys who hang on for a bunch of 2 WAR seasons don’t deserve nearly any credit in my mind.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Mar 21, 2009 3:57 PM EDT reply actions

Thinking out loud, how many points do we give for each season?

10 WAR: 15 points
8 WAR: 10 points
6 WAR: 6 points
4 WAR: 3 points
2 WAR: 1 point
0 WAR: no points

How’s that look? Just starting the conversation…

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Mar 21, 2009 3:58 PM EDT up reply actions

I wonder if there is any kind of formula to spit out some numbers

going by value to team (ie, playoff probability increase) those numbers are wrong. But I assume that you aren’t doing it like that.

vivaelbeñsheets

by vivaelpujols on Mar 23, 2009 2:40 AM EDT up reply actions

Right, I'm not following the same approach.

The pennants added article is here. WAR to PenAdd is pretty darn linear, about .06 per WAR. I’ve emailed David about getting the actual non-linear function.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Mar 23, 2009 5:20 PM EDT up reply actions

WAR/PA?

I actually had PA in the table because I was thinking of calculating the WAR per PA. Would that help? Does doing it in fewer seasons count more than doing it in fewer PA?

by adarowski on Mar 21, 2009 6:01 PM EDT reply actions

WAR per PA or WAR per season is useful.

Again, not the whole story. I mean, if a guy goes one season at 8 WAR, breaks a leg, and retires, well, I’m not putting him in the Hall.

This is sort of like the question that WAR, itself, solves on a seasonal level: how do you combine quality and quantity? Well, a player gets credit above and beyond some certain level of performance, not just zero. So maybe you only give a guy credit for any WAR above 4 (or whatever) in any given season (and don’t count seasons below that) That seems a little arbitrary to me, and prefer something non-linear, but I don’t have a good answer beyond that.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Mar 21, 2009 9:43 PM EDT up reply actions

I know some people at the HOM do something similar...

and only count seasons above average.

Granted, that’s a less demanding standard than 4 WAR, but it’s along the same lines.

Non-linear makes sense to me.

I need to re-read the Pennants Added articles, because the results I remember seem a little counter-intuitive.

by Dan Turkenkopf on Mar 21, 2009 9:56 PM EDT up reply actions

Where can I find the pennants added article?

Someone, maybe you, Dan, mentioned that in pennants added, two 3 WAR seasons rate better than a 6 and a 0 WAR season. That very well might be true, but I don’t think that’s what “we” should care about for determining the best players of all time. I’m open to hearing the argument for it, though.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Mar 21, 2009 10:03 PM EDT up reply actions

Here's the article...

Pennants Added

It wasn’t me who mentioned it, but I recall the conversation.

I’m sort of torn on the whole question. Do we measure players by how much they contribute to championships (which might suggest we look at the results from Gassko’s article) or by how “outstanding” they were (which puts a whole lot more weight on a 8 WAR season than a 4).

by Dan Turkenkopf on Mar 21, 2009 10:51 PM EDT up reply actions

Thanks, interesting stuff, and the method I was expecting (which is good).

I’m curious how Gassko’s adding Pujols to a 90-win team, though. Is he simply adding in his WAR? If so, would there be 600ish PAs of replacement level production to replace? It seems as though you’d need to remove something like an average player and then add in Pujols’ WAR. Maybe remove more production from a 90 win team than a 70 win team?

The conundrum about diminishing returns for great seasons would remain, though, even if we bumped the peak win total up from 4 WAR to 6 WAR. And I’m not sure how to resolve it.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Mar 22, 2009 9:30 AM EDT up reply actions

Right, the quantity vs quality balance

Again, not the whole story. I mean, if a guy goes one season at 8 WAR, breaks a leg, and retires, well, I’m not putting him in the Hall.

Of course, the same issue exists with OPS+, OPS, any other rate stat. I find it interesting that you value the same performance in the same number of plate appearances differently depending how it breaks down over multiple seasons. I’ll be watching the discussion… just always considered quantity and quality, not specifically how it breaks down over multiple seasons.

by adarowski on Mar 21, 2009 10:16 PM EDT up reply actions

Yeah, it's a bit different. We actually tend to value consistency within a season. (Except for Howard/Delgado situations.)

Tom Tango has pushed a philosophy that a player’s performance is a sample of his talent level. And even though his talent level actually changes over time, you can find a player’s highest peak talent level by finding his best consecutive streak of IP or PA. Now, the twist is that you need to regress. For 1 PA, it’s 99.999% regression to the mean. Over 1000 PA, it might be 70. At 10,000, it’s 40. (Just making up numbers.) So more PAs can help, but only if they’re really good.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Mar 21, 2009 10:29 PM EDT up reply actions

Thanks!

Everything I know about math I learned from baseball, so I’ll have to figure out what all that means. But interesting stuff!

by adarowski on Mar 22, 2009 10:59 AM EDT up reply actions

In English.

Think of a player’s performance as evidence of his greatness. The more evidence, the more sure you are of his actual talent level. So with just a little evidence, you’re unsure to believe in it. With tons of evidence, you take it almost at face value. Adding more seasons adds more evidence, but if it’s a 2 WAR season, that’s not evidence of greatness.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Mar 22, 2009 11:17 AM EDT up reply actions

Larkin

First ballot, nth ballot or never?

by Harry Pavlidis on Mar 21, 2009 11:22 PM EDT reply actions

His career is almost a clone of Ozzie's and Trammell's

Which means, well, next to nothing when it comes to HoF voters.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Mar 22, 2009 9:14 AM EDT up reply actions

Probably nth

I think he’ll get in, but it might take a while. I think he should be in right away, though. Alomar will probably be 1st ballot, but a lower percentage (80ish). Edgar’s only hope is the Veterans’ Committee (or sabermetrics going mainstream).

I’d love to see some of Sky’s WAR graphs starring Edgar…

by adarowski on Mar 22, 2009 10:51 AM EDT up reply actions

Win Shares of the same players

I think some of the voters recently have been considering them to help evaluate careers (reason Dawson is higher than Trammell). The year that Win Shares came out was the first time Bert Blyleven over took Tommy John in voting. Hopefully the list of lifetime WAR will help also. (Now if we can just get the list of pitching WAR.)

Rickey Henderson 535
Tim Raines 390
Alan Trammell 318
Mark McGwire 343
Andre Dawson 340
Mark Grace 294
Dale Murphy 294
Jim Rice 282
Matt Williams 241
Don Mattingly 263
Dave Parker 327
Harold Baines 307
Jay Bell 245
Ron Gant 206
Greg Vaughn 199
Mo Vaughn 200

by Jeff Zimmerman on Mar 21, 2009 11:59 PM EDT reply actions