wrap it

Kinder was 32 when he got to the bigs and he won 21 games in the Southern League 6 years earlier. And there was the war year in 1945 as an add-on.
If the MiL reserve clause issue is no longer at play, which I think by this time it's not, then I don't really see it any differently than Keller who was simply blocked, or alternately could be said to have been poorly scouted/evaluated by the parent team (since he was already better than the alternatives the Yanks had on hand). Same holds for Kinder, whether or not there was a parent team (I assume there was), MLB talent mavens did a poor job evaluating him. But that's baseball IF his rights were not reserved/chained-to a minor league club.

You can counter that this is unfair to them because they may have the ability to be more valuable if they were utilized differently.
Which is why we don't use lineup specific measures, we use ones that look at every batter the same so that we can consider them outside of the specific onctext in which they were deployed or uver/under-utilized.

Chris,

I disagree with you yet again. This is getting fun...

I woudl say that in your example Star B is more valuable/has more merit than Star A. I woudl say this because I dont' really care what replacement level player B does. It has no effect, in my opinion, on the HOM worthiness of Star A. If Star A plays at a higher rate than Star B but is injured to a point where his overall contribution to the team is less than that of Star B in a given season or seasons, then I would rather us elect Star B because he did more to push his team toward a title than Star A did. What player B who replaces Star A does is of no consequence to me because Star A has no effect on what player B does. I guess I am saying that I find it prudent to only look at the value given by that player and not by that player and what a resonable replacement would give should said player get hurt. It's jsut a diference of opinion I guess. This is why, even though I am a peak voter, I have not yet voted for Roush, McGraw, or Chance, though two of those (McGraw and Roush) may get a vote before 2007.

As for Kinder, as a guy who gives Keller MLE credit for one season I believe that Kinder may deserve some MLE credit. Of course there are two things to consider here. I first want to see what those MLE's would be. One woudl come in a half season (1941) and the other (1944 I believe) was in waht must have been a very weak southern league. I woudl also like to know what happened in 1942 and 1943, if Kinder wasn't very good then maybe something was up in 1940 when he won 20 games. Also, I am not sure if this would even put Kinder into my consideration set. I doubt taht he is close enough that a few years of MLB average MLE credit (which is what we may be looking at) is enough to put him even on the bubble.

I am also concerned that the electorate as a whole is overvaluing durability by accepting bulk WS as a straightforward measure of value.

I agree with Chris Cobb analytically but it is hard to agree with his diagnosis. Lip Pike, Cupid Childs, Heinie Groh and, who knows, maybe Charlie Keller, Ken Boyer, and Jimmy Wynn. The group has elected more players with short and medium careers, has inclined more to peak/prime, than I would have guessed in advance, based on knowing Joe Dimino and Dan Greenia (poorly). Rube Waddell, Dazzy Vance, Bob Lemon.

In a parallel universe, the project picks up Tommy Leach, Lave Cross, Deacon McGuire, maybe even that maligned HOFer Jake Beckley. How will Tony Perez and Dwight Evans fare?

Allen, Ashburn, Averill
Barnes, Bennett, Boudreau
Campanella, Caruthers, Childs

...which is the reason Barry Bonds, Babe Ruth and Ted Williams did not maximize their plate appaearances by batting leadoff, despite each having world-class on-base averages. They would have had 7% more plate apperances; and still would have been less valuable to their teams, as their positive contributions of walks, hits, and home runs would have accounted for far more than 7% fewer runs.

Ok, I'll agree that traditional metrics probably overvalue the plate appearances of a player who bats leadoff despite having skills that aren't specialized for the job (Soriano?). But what about the player with the classic leadoff skill set? If he draws a lot of walks and steals bases with a good percentage and isn't good at driving in baserunners, isn't it possible that his plate appearances will be more valuable when he's leading off the inning? I'm questioning whether there should be a single measure of "leverage" for batters, or different measures depending on the batter's skill set.

I don't know that this issue has been addressed by the research. Most studies using run expectancy matrices and the like have focused on the average lineup. To investigate the effectiveness of leadoff hitters with specialized leadoff skills (presumably followed by good batters who can drive them in), you need to move beyond an average run expectancy matrix; running simulations seems like the appropriate research strategy.

Mark S.,

I think it is misguided to assign merit to players in ways that don't correspond to the realities of what enables a team to win. If Star A (135 g) + Replacement Player (27 g) has more value to a team than Star B (162 g), then Star A's team is more likely to win a pennant than Star B's team, and the credit for that greater likelihood of winning belongs to Star A, because Replacement Player added only so much value as was freely available to any major league team. In fact, Star B's team _has_ a player like this (that's the definition of replacement level), but doesn't use him because Star B plays every game. Star B, even playing every day, however, couldn't produce as much value as Star A and Replacement in the same number of games. If Star A's team is better off than Star B's, then I don't see how it makes sense to argue that Star B is more meritorious than Star A.

Paul,

You wrote: I agree with Chris Cobb analytically but it is hard to agree with his diagnosis [that we are overvaluing durability by accepting bulk WS as a straightforward measure of value]. Lip Pike, Cupid Childs, Heinie Groh and, who knows, maybe Charlie Keller, Ken Boyer, and Jimmy Wynn.

I was speaking in the context of _seasonal_ value. I agree with you that the electorate has inclined much more than it might have towards peak/prime. Career bulk win shares is not in itself a ticket to election. However, all the short-career players that you have listed have put up big peak seasons partly through in-season durability. In-season durability is a very good thing, but low replacement levels (for bws in win shares, for frar in WARP) exaggerate its value. Now, players with short careers who aren't durable don't have much going for them. But as we sift through the borderline candidates, it seems as if the electorate is gravitating pretty strongly to the durability that produces bulk single-season win shares rather than to the talent that produces high rates of production, and I think that attraction to bulk seasonal win shares has been insufficiently examined. Perhaps the differences in rates of production among remaining candidates are so small that differences in durability make much more of a difference between the players, but until the matter is studied and debated, I won't be confident that the electorate's preferences are correct.

Chris,

I don't think we are going to resolve this one at all honestly. I would still pick Star B. Star B added more himself than Star A did. We dont' get the chance to vote for Star B and Star A/player b, we only get to vote for Star B or Star A. To me the question is, "Which player gave his team more value" and to me that is Star B who gave his team more value since Star A needs replacement player b to put him over the top. I don't want to be a GM here (where I could theorectically find a player far above replacement as Star A's necessary replacement), I want to assess the actual value given by that player to a baseball team.

But as we sift through the borderline candidates, it seems as if the electorate is gravitating pretty strongly to the durability that produces bulk single-season win shares rather than to the talent that produces high rates of production, and I think that attraction to bulk seasonal win shares has been insufficiently examined.

This is one of the reasons a number of voters have for not voting for Browning and Keller, is it not? You list a lack of in-season durability as one of the (several) reasons you have Browning rated so low on your own ballot, Chris.

Or am I misreading you?

But as we sift through the borderline candidates, it seems as if the electorate is gravitating pretty strongly to the durability that produces bulk single-season win shares rather than to the talent that produces high rates of production, and I think that attraction to bulk seasonal win shares has been insufficiently examined.

That's why I have stated numerous times that there has to be a combination of rate and bulk WS when analyzing a player's career. Relying on one or the other wont give you an accurate picture of that player's value, IMO.

Rod Nelson of SABR was kind enough to send me Kinder's TSN obit. It isn't explicit, but seems to be saying that Kinder's career got a late start because he had a better paying job in construction. Not clear on why he gave that up. Also, he skipped playing in 1943 because again he had a better job, working on the railroad. I may have said before he was in the American Association in 1944, but it was the Southern Association. He was in the KITTY League until the Yankees bought his contract in 1941, but they returned him after a brief trial. Then Memphis acquired him in late 1942.

Also, he skipped playing in 1943 because again he had a better job, working on the railroad.

In 1943, it's possible that a railroad job may have been considered a critical occupation that would have allowed him to avoid, or at least postpone, the draft.

This is one of the reasons a number of voters have for not voting for Browning and Keller, is it not? You list a lack of in-season durability as one of the (several) reasons you have Browning rated so low on your own ballot, Chris.

Or am I misreading you?

You are correct. The fact that I am arguing in principle that single-season win-share totals are not, by themselves, an accurate measure of a player's peak value, doesn't mean that I don't believe durability matters. It does matter, and in Browning's case the lack of durability (along with other factors) contributes to his low placement on my ballot.

It is the case, however, that I have only very recently _quantified_ the overvaluing of durability that appears in both WARP and win shares, so I have not attempted to adjust my system and re-evaluate every player on the basis of numbers revised to account better for replacement level. I have, up to this point, used a peak rate stat as the component in my system that counterbalances the bulk peak measure. I think that my system has worked decently, but as I examine the borderline candidates, I can see biases emerging in my rankings that I can't justify and must correct for. I'd like to have a system that is based more closely on measures of value that I trust are accurate, so that less subjective adjustment is necessary. Thus, I plan to reevaluate Pete Browning and all the other borderline candidates, and I may reach new conclusions about them as a result.

To me the question is, "Which player gave his team more value" and to me that is Star B who gave his team more value since Star A needs replacement player b to put him over the top. I don't want to be a GM here (where I could theorectically find a player far above replacement as Star A's necessary replacement), I want to assess the actual value given by that player to a baseball team.

The fact that Star A needs the replacement player to get him over the top is irrelevant, because every player by definition _has_ such a replacement available. That's what replacement level means. Since everybody can get a replacement of that calibre for free, value added below that level is not actually meaningful to a team. Replacement level is the _minimum replacement value_ that any minimally competent GM could plug in. It's possible for a competent GM to have better replacements available, in which case Star A's relative lack of playing time is even less of an issue for his teams. (It's possible for radically incompetent management to supply replacement players who are below replacement level, of course, but I prefer players who are more valuable to competent management than players who are more valuable to incompetent management. This is another reason to give hold-out credit, in my view.)

So when you are saying "actual value," you are not talking about actual value: you're talking about total wins added, and that's not the same thing. The value added by a player is the value that a team could not have gotten without him, and that value is value above replacement, because the team will always have 9 guys in the lineup. I know you mostly consider "value above average" and that is fine, and compensates significantly for the low zero point in win shares. However, it doesn't compensate competely, because the low replacement level means that players who miss time are penalized more than they should be.

Chris Cobb
I am also concerned that the electorate as a whole is overvaluing durability by accepting bulk WS as a straightforward measure of value.
Paul Wendt
I agree with Chris Cobb analytically but it is hard to agree with his diagnosis [that we are overvaluing durability by accepting bulk WS as a straightforward measure of value]. Lip Pike, Cupid Childs, Heinie Groh and, who knows, maybe Charlie Keller, Ken Boyer, and Jimmy Wynn.
Chris Cobb
I was speaking in the context of _seasonal_ value.

oops. As one who has distinguished longevity and durability, urging everyone else to keep them straight, I am ashamed of myself.

I prefer players who are more valuable to competent management than players who are more valuable to incompetent management.

John Kruk gets no extra credit because his strengths and weaknesses are so apparent that any buffoon would use him correctly?

Chris,

My biggest problem with what you are saying is that I dont' think that replacement level value that a team gets when Star A goes down with an injury should be attributed to Star A. I understand that it is better for a team to have Star A/player b than Star B in any given season, but I dont' agree that Star A should get all of the credit in the Star A/player b combination.

Mark,

I don't advocate attributing the value of the replacement player (which, incidentally = 0) to Star A. I advocate not attributing to the value either Star A or Star B the portion of their total wins that would have been earned by a replacement player in the amount of playing time logged by Star A and Star B, respectively.


Paul,

My phrasing probably wasn't sufficiently clear. In a multi-post discussion, the original context often becomes obscure by the later posts.

Ken Fischer 1991 Ballot

5-Carl Mays 256 WS
256 win shares in an offense dominated era is impressive.

Why? There are just as many win shares in a high scoring era as in a low scoring one. More specifically, here are some unelected pitchers from Mays' own time with more:

Wilbur Cooper 266
Burleigh Grimes 286
Jack Quinn 287

and from a slightly later offense dominated era:
Waite Hoyt 262

And that's just the pitchers. And ignores the ones with actually impressive WS totals, who were elected.

I'd be interested in a well-laid out case for Ellis Kinder for extra credit. I have him as the #2 swing man of all-time behind Firpo Marberry, with an outstanding DRA+ of 125.

I'd need to see at least 3-years of value equivalent to his 1949 or 1953 before he'd hit my consideration set, but anything is possible.

This reminds me, I still need to work out Sal Maglie's hidden credit . . .

>I'd be interested in a well-laid out case for Ellis Kinder for extra credit.

Joe, having asked the question myself, I have now become satisfied that there isn't such a case.

Yes, he was out of baseball (apparently working in defense plants) in 1943 and 1945, and he went 20-6 in 1944 but it turns out to have been in an AA equivalent league, not higher. And he was an approximately ML average pitcher in his first 3 years in the majors 1946-47-48.

And yes, he had a good looking year as early as 1940 but it turns out to have been in a Class D league.

If he had been pitching in '43 and '45, let's say his skills would have developed sooner AND he could/would reasonably have been expected to get noticed by a smart ML scout and signed by a ML GM. If the best possible case had occurred every step along the way, then I could argue 1945 for sure, and maybe 1944, but any more than that is a stretch. But if you want to stretch, I could maybe even imagine 3 years of MLE credit--but as an average ML pitcher at best. But more likely half of that.