Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Hall of Merit > Discussion
Hall of Merit
— A Look at Baseball's All-Time Best

Wednesday, April 16, 2003

Token appearances

Let’s discuss token appearances here, I’ll lead off with DanG’s post . . .


While researching token appearances for upcoming elections, I ran across a few players that the “ballot committee” may have to pass judgement as to their eligibility.

The next election has one minor case, Joe Mulvey. After playing 55 games in 1893, he played none in 1894, then retired for good after playing just 13 games in 1895. Is he now eligible?

A more interesting case is Matt Kilroy. An outstanding pitcher for Baltimore in the late 1880’s, he pitched in 8 games in 1894 before disappearing. He resurfaced one last time in 1898, playing 26 games (13 as pitcher). When is he eligible?

Bill Hutchison is similar to Mulvey. He was done after pitching 38 games in 1895. After not pitching in 1896, he ducked back in for his final 6 games pitched in 1897. When is he eligible?

The first superstar the committee should assess is Sam Thompson. He played 119 games in 1896 before succumbing to injuries. He played only 3 in 1897 and 14 in 1898 (plus 8 G in 1906). When is he eligible?

Jack Clements is similar to Thompson. After 99 games in 1898, he played only 4 in 1899 and 16 in 1900. When is he eligible?

Finally, Hugh Duffy might deserve earlier eligibility. Retiring after 79 games in 1901, the Phillies made him manager in 1904. He proceeded to play 18 games that year and 15 in 1905. Also one more in 1906. When is he eligible?

Personally, I favor early eligibility for all these players.

BTW, the record books credit Duffy with 13 walks in 59 PA in 1904. That doesn’t seem right, it’s WAY out of line with anything else in his career. Is there some way to verify that?

I made a list of other upcoming candidates who ended their career with token appearances, but I don’t have it with me. Maybe tomorrow. One that I recall for 1894 retirement is 157-game winner Elton “Icebox” Chamberlain, who didn’t play in 1895, then finished with 2 games in 1896

Joey Numbaz (Scruff) Posted: April 16, 2003 at 09:44 PM | 93 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. Devin has a deep burning passion for fuzzy socks Posted: April 16, 2003 at 10:02 PM (#512192)
I don't really care that much, but my gut says 18 and 14 games played is more than just a token. I also tend to wait until the career is done if it was continuous. If David Cone keeps pitching decently (or gets back to it :)) but gets injured in mid-May, was this year just a token appearence? It doesn't feel that way to me. But I'm not going to get upset about it.
   2. jimd Posted: April 17, 2003 at 12:36 AM (#512193)
I agree with Devin. I really don't care much what the details of the eligibility rule is. However, we do need to decide on a rule, and then stick with it. No matter what rule we pick, someone will find two borderline cases that differ by one game and straddle the line drawn by the rule. So, here's a proposal.

For position players in a 154 game season: single digits in GP is token, double digits is not.

Generalize this to other season lengths and to pitchers as follows:
   3. Marc Posted: April 17, 2003 at 01:52 AM (#512194)
I thought we had agreed that 10 games by a position player and 5 by a pitcher was an active season, 9 and 4 or less was not. Thompson is an interesting case, even if 1906 counted as a "real" season, we wouldn't know about it in 1904.
   4. jimd Posted: April 17, 2003 at 02:15 AM (#512195)
If that's what we agreed to, then that's fine with me. I didn't pay much attention to that debate before.

To me, the most important thing is to pick a rule and go with it. I don't think it makes a big difference, though I suppose it could conceivably influence which borderline player gets elected when.
   5. DanG Posted: April 17, 2003 at 12:56 PM (#512198)
Our Constitution says this:
   6. RobC Posted: April 17, 2003 at 01:38 PM (#512199)
if we stick to a hard rule (like what we have in the Constitustion now) I will write a script this weekend to pull new eligibles by year from the Lehman (sp?) database. Then, we know we wont miss anybody. If our rule is flexible, I wont be able to do it.
   7. DanG Posted: April 17, 2003 at 01:55 PM (#512200)
Test.
   8. DanG Posted: April 17, 2003 at 02:14 PM (#512201)
"RobC
   9. RobC Posted: April 17, 2003 at 06:00 PM (#512203)
Joe -
   10. RobC Posted: April 17, 2003 at 10:08 PM (#512204)
I appear to have a working script. The logic evolved a little, as I will explain. There are 3 categories of years: token, nontoken and indeterminate. A year with <10 Games (hitter) or <5 games as a pitcher is a token year if it occurs after the last nontoken year. a year with 40+/20+ games is always a nontoken year. 10-39/5-19 are initially indeterminate. If the indeterminate year is immediately after a nontoken year then it is also a nontoken year. Otherwise, if it is after the last nontoken year, it is a token year.

I hope that isnt too confusing. If we want to change any of the numbers, that is easy to do.

Examples:
   11. RobC Posted: April 17, 2003 at 10:11 PM (#512205)
1899: Minimum 500 games for hitters, 250 for pitchers.
   12. RobC Posted: April 17, 2003 at 10:22 PM (#512206)
My first post didnt work right trying again.
   13. MattB Posted: April 18, 2003 at 01:47 PM (#512207)
The purpose of the "token appearance" rule, as it was first discussed, was to permit players like Jim O'Rourke (last "real" season, 1893, but played in 1904) to be considered with his peers who also retired around 1893.

I don't think we need a special token rule for players whose careers petered out in consecutive years. Many, many players have a last season with only 2 or 3 games played. Very few have Minnie Minoso careers.

If Dave Foutz played every year between 1884 and 1894, I say he retired in 1894. Similarly, Sam Thompson should be considered to have retired no earlier that 1898.

I believe any token rule should not kick in until a player has missed at least one complete season.
   14. RobC Posted: April 18, 2003 at 02:26 PM (#512209)
I agree with Matt and Andrew also. Its why when I was pulling these token appearances, my script considers the indeterminate case to be nontoken if it immediately follows any nontoken season. We do need to determine this now!!! because it affects Pete Browning for our next vote. Right now, he is eligible. He has to wait a year if we use the peter(out) principle.

O'Rourke would still be a token performance. But what about Mulvey (not that it really matters with him)?
   15. Howie Menckel Posted: April 18, 2003 at 02:30 PM (#512210)
hmm, I agree with this. Seems like the point was to ignore tokens after a guy basically retired. If he plays a few games after a semi-regular season, that keeps counting. So Browning should go to 1900, no?
   16. John (You Can Call Me Grandma) Murphy Posted: April 18, 2003 at 02:38 PM (#512211)
Put me down in agreement with the MattB rule.
   17. Carl Goetz Posted: April 18, 2003 at 03:13 PM (#512212)
I agree with the MattB rule. The token rule should only apply to players who left the game and came back later for a token appearance. The O'Rourke's and Minoso's of the baseball world. Otherwise, it should be based on their final season.
   18. Marc Posted: April 18, 2003 at 04:12 PM (#512213)
Any rule that can realistically be referred to as the "peter(out) principle" needs to be adopted, now!
   19. MattB Posted: April 19, 2003 at 02:28 AM (#512215)
Now, Pete Browning is an even harder case to make the argument that his last season shouldn't count because he played for two different teams in his last season (albeit for 3 games total). I don't know the story, but it seems pretty likely that he didn't think of himself as retired if he kept trying to play.

The issue is that many, many players have a taper off year. I guess I consider the HOF rules as a "baseline". Eventually, when we "catch up", we will be considering the same general group of people that the HOF is considering. Vince Coleman played 6 games in his last year. Rick Honeycutt played 2. Don Slaught played 20. Danny Tartabull played 3. Mickey Tettleton had 17. Mitch Williams had 7. Sid Fernandez had 1. Mark Gubicza had 2.

And that's just the National League in 1997. The 2003 ballot (1997 retirees) will consider all of these individuals for the first time. Depending on the "token" rule we use, any of these players could be considered by us earlier. If we don't consider these players the same time as the HOF, we're not really demonstrating how our election system is different/better.

The HOF rules can be changed if there is a good reason to, but those are the rules that have to be improved upon. Is allowing each of these players to be considered a year early worth the decision to not have them be considered the same time that the HOF does? Did all of these players stop playing in April? The "token" rule is different, but is it significantly better to warrant a change?

I see either rule as valid in a vacuum, but I don't see the token rule, as expressed, as significantly better in any respect to warrant a change, except as it applies to players who were out of baseball for years before their return -- an event that is rare, and is even more rare now among elite players.

My post this morning seems to have spawned a consensus among people who logged in today. I'm willing to listen to the other side, but I think it would take more than proving that the alternate rule is equally good. We are not writing on a blank slate, and I think there has to be more than just change for change's sake.
   20. MattB Posted: April 19, 2003 at 02:33 AM (#512216)
"That's a guy that makes it through spring training, plays for a week or two and bags it."

To respond directly to that, it could also be a guy who injures himself, works to rehab all year, and doesn't make it back to the big leagues until September, where he bombs.

In the case of Pete Browning, it's a guy who signed with a new team, and then played for another team. He was obviously trying to stick around. Was there an injury in the middle? Did he play for both teams in April? He could be a player-coach who played one game a month for the entire season whenever the starter was injured. There are lots of ways to only play a few games that don't look as much like retirement.

"First week of the season and out" is one way a player can play 3 games, but it's not necessarily the only way.
   21. DanG Posted: April 19, 2003 at 05:22 AM (#512217)
The consensus on this thread essentially agrees with the ideas I put forth in some of the first HoM threads back in January 2002. The idea was, as Carl Goetz wrote: "The token rule should only apply to players who left the game and came back later for a token appearance." MattB explained it like this:

"I don't think we need a special token rule for players whose careers petered out in consecutive years. Many, many players have a last season with only 2 or 3 games played. Very few have Minnie Minoso careers.

If Dave Foutz played every year between 1884 and 1894, I say he retired in 1894. Similarly, Sam Thompson should be considered to have retired no earlier that 1898.

I believe any token rule should not kick in until a player has missed at least one complete season."

Joe disagreed with this and decided to go with the rule as it is worded in our constitution, ie, no delay. Since it isn't a big-dot-deal, and because there is logic in either rule, I decided Joe's rule was acceptable. There really wasn't much discussion apart from mine and Joe's.

Now this issue has raised its head even after the HoM has started rolling. Personally, I don't feel strongly either way. We'll still need a "ballot committee" to decide borderline cases, no matter which rule we go with. I suppose I prefer Joe's rule, since it will mostly effect our early elections and I kind of like getting a crack at these guys a bit earlier.

But it does need to be decided now, since Pete Browning is sitting by the telegraph waiting for the word on his election.
   22. RobC Posted: April 19, 2003 at 01:35 PM (#512218)
If we are going to stick with the current rule (which I have no problem with) we need a decision on what the borderline cases are. Is <10 always token and >15 nontoken, with 10-15 being borderline? Or is it 10-20? Or 5-15? (Halve these numbers for pitchers) If someone will give me the borderline boundaries I will generate the first 25 years of borderline cases for us to start discussing.
   23. MattB Posted: April 19, 2003 at 06:29 PM (#512221)
I don't know.

Hornsby seems like an even easier case to me.

He played 20 games and had 63 plate appearances in his last year. That's over 10% of the season.

The question isn't whether he was still productive, or still at the top of his game. The question is whether he has retired. And if you are playing 20 games, you haven't retired.

An alternate rule would be to allow a voter to vote for a player as soon as he had garnered enough Merit to be considered. Say, Hornsby was a HoM player by the end of his 1925 season, so he gets a 15th place vote then, moving up to 12th after his 1926 season and 5th after 1928. Alternately, everyone could become eligible when they turn 45, irrespective of when they retired.

But the issue isn't when he stops being useful, it's when he retires. If a player plays 10 games a year for ten years from age 41 to 50, and goes 0 for 40 each year, I consider that hurting his team, and dock the player accordingly, something which I wouldn't do for "retired" players.
   24. MattB Posted: April 19, 2003 at 11:46 PM (#512223)
Yeah, but Hornsby was a player-manager. Any playing time issues would be of his own making (not that he was bad when he played).

Those 20 games in 1937 were likely scattered throughout the season whenever he saw fit to activate himself.
   25. John (You Can Call Me Grandma) Murphy Posted: April 19, 2003 at 11:59 PM (#512224)
I think I had an "Oscar Madisoy" moment there today at 4:29. :-)
   26. Marc Posted: April 20, 2003 at 03:17 AM (#512225)
The idea that we are "penalizing" a player by waiting vote for him til after he has retired is kinda like those budget cuts that are really increases, just less than somebody wanted. And we're not talking about penalizing a guy who hung around to play 2 games, I think the rule was always 10. Obviously we want to vote on Minnie Minoso in 1970. When we vote for Hornsby isn't going to matter, no "penalty." I don't think it will matter when we vote on Pete Browning but as I understood the proposed rule he would be eligible in '99. Thompson, however, as I had understood it, retired after 1898. Like most of you, I don't have a big problem with this one way or the other, but let's not trump up the issue by talking about penalizing Sam Thompson for playing 2 games. That's two misstatements. Having said that, I will do whatever Joe and Rob decide.

John (or Joh), I don't know about your "Oscar Madisoy" moment, but you are exactly correct re. Linear Weights. Until they start taking back World Championships from teams that can't play .500 ball the following year, I can't understand negative player ratings.
   27. John (You Can Call Me Grandma) Murphy Posted: April 20, 2003 at 04:58 AM (#512226)
John (or Joh), I don't know about your "Oscar Madisoy" moment,

I was referring to an "Odd Couple" episode where Bobby Riggs makes a bet with Oscar that he can't type his name correctly. When I saw that I posted my name as "Joh," I remembered it. :-)

As for your comments on Linear Weights, it seems we are of one mind.
   28. John (You Can Call Me Grandma) Murphy Posted: April 21, 2003 at 01:44 AM (#512228)
Secondly, if the Red Sox signed me tomorrow, I would bat .000, and I would make countless errors. That is a negative contribution, no matter what your baseline is.

Fine, but that has absolutely zip to do with what you accomplished prior to that. Take Pete Rose or Brooks Robinson: they had a few seasons where they were comfortably below average, so Linear Weights knocks them down a few pegs. If a GM was idiotic enough to allow them to play for ten more years, they would start to room with the Omar Morenos or Rey Ordonozes with Linear Weights. This is absurd.

In your mind, Player X going 1 for 87 as a 45 year old may be a non-negative contribution, in that it's better than being dead.

In my mind, he would get credit for it, except the number that I would give him would be so extremely small (decimal numbers) that he might as well have not played at all.

Fourthly, one of the major failings of Win Shares is that Bill James refuses negative numbers, despite having a non-zero baseline.

I think the failing that his Win Shares book had was not including WS per Games for each season and career. Fortunately, I can compute them on my own.
   29. Marc Posted: April 21, 2003 at 02:07 AM (#512229)
I'm with John. If Player A has accumulated 25 wins (TPR) through 2002 then retires, his final rating is 25. Meanwhile, Player B, has 25 wins (TPR) through 2002 but instead of retiring plays one more year and has a rating of -2, his final rating is 23.

Player A obviously ended up with a better career record (TPR) than Player B. I understand what the numbers mean and they accurately record what they mean to record. But...what bothers me is that if Player A and Player B are the same player, his legacy is harmed by continuing to play rather than retiring. Now, in this scenario, the numbers do not make sense. If the player in question had negative value (i.e. like you, he went 0 for 87 and made a bundle of errors) he would not play out the year. In order to accumulate a -2 rating, he would have to play out the year, which means his manager is an idiot OR he has some value.

Put the baseline at replacement value and say he had 50 points of value as of 2002 and he continues to play in 2003 at below average but above replacement, then maybe his career ends up 50.1. This makes sense and I think this is more approximately what WS does. His continuing to play does not negate what he has already done, which I think reflects the real world better than TPR.
   30. RobC Posted: April 21, 2003 at 04:09 AM (#512230)
TPR is the problem, not Linear Weights. Linear weighting can be used with the baseline set wherever you want. Heck TPR can be used with a factor to shift the baseline. Figure how far below average you want the baseline to be and multiply that by playing time. Add the result to the TPR number to get your TPR'. That extra year at slightly below average now has value.

I have a bigger problem with the assumption of a linear model than I do the average baseline. The latter can be easily adjusted for.
   31. John (You Can Call Me Grandma) Murphy Posted: April 21, 2003 at 05:27 AM (#512231)
RobC:

I agree there is some good aspects of Linear Weights. That's why I qualified my opinion of it before as such: That's one of the reasons I left Linear Weights, for the most part, a long time ago.

You are correct that TPR is the culprit.

BTW, I have used your TPR adjustments in the past.
   32. MattB Posted: April 21, 2003 at 12:46 PM (#512232)
It's Monday morning and balloting should resume today.

Do we have a rule?
   33. DanG Posted: April 21, 2003 at 01:41 PM (#512234)
Eleventh hour (as usual) commentary on the issue at hand, token appearances.

I may have finally puzzled out a compelling logic in support of the rule as stated in our constitution. There seem to be two concepts that need elucidating:

1) What does "retirement year" really mean?
   34. DanG Posted: April 21, 2003 at 02:35 PM (#512236)
Perhaps I should try and formalize a proposed rule.

Our Constitution has this regarding eligibility:

"We generally want players to appear on the HoM ballot with their contemporaries. Accordingly, we will ignore token appearances at the end of a player?s career in determining when a player?s HOM eligibility begins (i.e., the first HOM ballot he can appear on).

To discount token appearances, a player becomes eligible 5 years after the first time he plays fewer than 10 games in the field or pitches in fewer than 5 games, assuming he never plays in 10/pitches in 5 games again. If he does play in 10/pitch in 5 games later in his career, the HoM ballot committee will determine in which year the player?s HoM eligibility begins."

Proposal: Remove the last phrase, "the HoM ballot committee will determine in which year the player?s HoM eligibility begins."

Replace it with this: "his year of eligibility will remain unchanged if he plays in no more than 40/pitches in no more than 20 games for his remaining career, provided that in any single season he plays no more than 20/pitches no more than 10 games.

If he meets the 40/20 requirement but exceeds the 20/10 limit, his eligibility will be delayed one year from the original year."

Applying this specifically. So if this seems reasonable.

Hornsby finished his career like this:

Year Games
   35. DanG Posted: April 21, 2003 at 02:42 PM (#512237)
Correction for Hornsby. The reasoning should go like this:

Year Games
   36. DanG Posted: April 21, 2003 at 02:47 PM (#512238)
(1) Player plays 1920-1935, then plays 5 games in 1936 -- retirement year is 1936.

I would rather use 1935.

(2) Player plays 1920-1935, then plays 5 games in 1937 -- retirement year is 1935.

Agreed.

(3) Player plays 1920-1935, then plays 55 games in 1937 -- retirement year is 1937.

Agreed.

(4) Player plays 1920-1935, then plays 55 games in 1937 and 5 games in 1938 -- retirement year is 1938.

Would rather use 1937.

(5) Player plays 1920-1935, then plays 30 games in 1937 and 25 in 1938 -- retirement year is 1938.

Agreed.

(6) Player plays 1920-1935, then plays 55 games in 1937 and 5 games in 1939 -- retirement year is 1937.

Agreed.

(7) Player plays 1920-1935, then plays 30 games in 1937 and 25 games in 1939 -- retirement year is 1937.

That may be OK. My proposed rule would make it 1938. What's the logic here?
   37. RobC Posted: April 21, 2003 at 03:29 PM (#512241)
I think the rule needs to be simple so we can tell at a glance.

Here are 2 possibilities:

1. Inflexible: Eligible 5 years after last season playing at least 10/5.
   38. DanG Posted: April 21, 2003 at 03:51 PM (#512243)
Andrew--

Our main area of disagreement centers on the issue of consecutive seasons played, or seasons contiguous to the rest of a guy's career. You think this is meaningful and I don't.

I have decided that just because a season is contiguous to the main part of a career doesn't make it significant. Irrelevance is irrelevance.

I also disagree with you (and RobC) that our rule needs to be simple. I give higher priority to logic and fairness. I think we are sophisticated enough to deal with a multi-tired rule.
   39. dan b Posted: April 21, 2003 at 04:05 PM (#512246)
Is there really a need to change the rule we originally put in the constitution?
   40. DanG Posted: April 21, 2003 at 04:13 PM (#512248)
Mark--

There are many obvious problems with doing it the HOF's way. Refer back to some of the earliest HOM threads where these were discussed.

dan B--

No, the constitution rule is good. Some didn't like its lack of specifics, which I have tried to rectify.
   41. DanG Posted: April 21, 2003 at 04:45 PM (#512250)
Andrew Siegel wrote:
   42. MattB Posted: April 21, 2003 at 04:45 PM (#512251)
"Frankly, I don't care. We just need to pick."

I'm with Andrew. The positions are all out there. I like simple and quick: Five years after your last consecutive season, unless you later play in more than 10 games in a season.

I think it's ridiculous, though, to delay the second ballot because of it. I obviously didn't spend enough time reading the constitution, as I though the rule as I stated it, was the rule.

But it's not that important.

The positions have all been outlined. Joe has been fair before in concretizing the general sentiment. I'm fine with whatever rule he picks, but I think he should pick it today and start a ballot thread.
   43. RobC Posted: April 21, 2003 at 04:57 PM (#512252)
Actually, correcting my previous post, #2 isnt out constitution with numbers put in. But it was inspired by the constitution. Looking at the constitution the range for borderline should be 10-19/5-9. (Not that the constitution implies any level at the top of the range, but I dont think we want to include Walt Wilmot's 1898 as a token appearance.)
   44. Rob Wood Posted: April 21, 2003 at 05:23 PM (#512255)
I think the rule presented in the constitution is fine. Plus it allows for special cases to be decided by the ballot committee.
   45. John (You Can Call Me Grandma) Murphy Posted: April 21, 2003 at 05:24 PM (#512256)
The main thing for me is that Jim O'Rourke and Minnie Minoso are not newly eligible over twenty years after they retired. Other than that, who cares about when Sam Thompson, Pete Browning or whoever is eligible? They're still going to be included with their contemporaries.
   46. Marc Posted: April 21, 2003 at 11:58 PM (#512258)
I'm really flattered that Joe included my name among those to chime in on this but wouldn't you know I was out all day. In the spirit of Ed Wood, I think the rule in the Constitution is perfect.
   47. DanG Posted: May 20, 2003 at 04:22 PM (#512259)
I hope it's OK if we move this discussion here.

"Posted 11:46 a.m., May 20, 2003 - MattB
   48. RobC Posted: May 20, 2003 at 04:53 PM (#512260)
Thompson played 25 of 1407 games after 1896. That is nearly 2%. I dont know where I draw the line, but after 1898 he only played 8 of 1407 games; less that 1%. I would make 1898 his last year by that standard.

The fact that he played more games in 1898 than in 1897 (and it wasnt 1 then 3 games) seems to me that it wasnt just some token appearances.
   49. MattB Posted: May 20, 2003 at 05:24 PM (#512261)
I also think age should be a factor. 14 games at age 38 is a lot different then 10 games at age 48. Cap Anson had just retired at age 45. He is two years younger than Dan Brouthers, and the same age as Buck Ewing, who was primarily a catcher (so would tend to retire earlier).

The numbers 10 and 5 may be arbitrary, but they the numbers we have. I don't agree with the 2% standard that Dan has set, because in a standard 2000 game career, 2% works out to 40 games. I would accept a 2% standard if the arbitrary number were 40 games, but to use it with a 10 game base seems primarily intended to evade the number in the Constitution.

There is a difference, I believe, between "token" and "meaningful". My opinion of Sam Thompson will not be much swayed by 67 plate appearances in 1898, but honestly they wouldn't have been swayed by 200 or 300 plate appearances in 1898. That doesn't mean that 300 plate appearances should be considered "token".

Thompson had (1) over 100 plate appearances after 1896; (2) was attempting to come back from a back injury in 1897 and 1898, not from a previous retirement; (3) did not miss a single complete year until after 1898; played in more than 10 games after 1896; (4) received more than 1.5% of his games (and plate appearances) after 1896; and (5) was not yet in his 40s when the short seasons occurred, making a productive comeback a realistic possibility (and in fact he was successful when he was playing).

Those are my arguments. Discuss among yourselves.
   50. Carl Goetz Posted: May 20, 2003 at 05:45 PM (#512262)
What if we just say players are eligible the year they turn 45? Its a simple query for Rob and its hard and fast. We could add a 'Nolan Ryan/Hoyt Wilhelm' rule that says if a player is still in his consecutive career at 45, he is eligible the 1st year that he does not play. We really wouldn't need that rule since Nolan and Hoyt were probably HoMers by 45 anyway, and most others did nothing of note after that age. I know this would be a radical change at this point, but it just seems easier to me.
   51. DanG Posted: May 20, 2003 at 05:52 PM (#512263)
For purposes of this discussion, Thompson's games in 1906 are not applicable, as he will be eligible before he plays them.

After 1896, and through 1901, Thompson played 1.22% of his games (17/1399) and had 1.25% of his PA (81/6470). IMO that's token, i.e., not meaningful.

The fact that he was attempting to come back from injury simply makes him like most players: he was out of the game due to diminished physical capacity.

As for his age, he turns 42 in early 1902; a comeback hardly appears imminent. This ain't like Bob Caruthers, who came on the ballot at age 35.

As for evading the Constitution, don't let what is written there cloud your thinking on this issue. Remember where those numbers came from: I tossed them out there in January 2002, they were taken up by Joe, and they have somehow achieved a sacrosanct status in the minds of some people.
   52. RobC Posted: May 20, 2003 at 06:09 PM (#512264)
I have trouble counting any appearances in a year immediately following another year in which the player played as a token appearance. 1896 was a full season, 1897 was at attempt to continue playing, as was 1898. 1906 was token. Im willing to consider something like 1897 as token because it really doesnt matter, lots of players have that small number of appearances before they finally hang it up. Im not hung up on the numbers in the constitution, if Thompson had played 14 games in 1906, I would still think they are token. But 14 games in 1898 isnt token, its a legitimate comeback attempt, that worked for 63 At Bats.
   53. RobC Posted: May 22, 2003 at 05:55 PM (#512266)
It doesnt really matter to me which way we go as a general case, but Thompson's case is different. Thompson's 1898 wasnt meaningless. He added on 1.5 wins of career warp-3 and 1.8 wins of warp-1. His 1898 was a productive season which added on to his career value, and not in a meaningless way. His 1897 and 1906 are meaningless (-.1 and .0 of warp-3; -.1 and .1 of warp-1). If a player is truly done, I should be able to throw away his stats and it not affect where I rank him. I cant say for sure that Thompson's ranking wont be affected by throwing away his 1898 season.

63 at basts with a .959 OPS isnt meaningless. I recommend 1904.
   54. John (You Can Call Me Grandma) Murphy Posted: May 22, 2003 at 06:35 PM (#512267)
It doesnt really matter to me which way we go as a general case, but Thompson's case is different. Thompson's 1898 wasnt meaningless. He added on 1.5 wins of career warp-3 and 1.8 wins of warp-1. His 1898 was a productive season which added on to his career value, and not in a meaningless way. His 1897 and 1906 are meaningless (-.1 and .0 of warp-3; -.1 and .1 of warp-1). If a player is truly done, I should be able to throw away his stats and it not affect where I rank him. I cant say for sure that Thompson's ranking wont be affected by throwing away his 1898 season.

63 at basts with a .959 OPS isnt meaningless. I recommend 1904.


I agree with you, Rob, but in the end, Thompson is going to be on the sidelines cheering Brouthers and Ewing at their induction ceremony.

Same goes for 1903.
   55. Marc Posted: May 22, 2003 at 06:49 PM (#512269)
I don't like the subjectivity that is being injected here. It's not a question where 10 and 5 came from (thin air), it's a question of where they ended up (in our constitution).

I don't like the possibility (and I am not accusing anybody, but the possibility exists nevertheless) that somebody who really really likes Sam Thompson might look at the upcoming eligibility list and decide he has a lot better chance of getting elected in 1902 than 1903, or 1903 rather than 1904, or whatever, and THEN slot him into that more favorable spot. We DO need a hard and fast rule that does not respond to subjective judgements of whether he "altered" his "record" or not, or what his age was, or what his injury status was. Will we always know injury status?

I think it should be by the numbers, it should always be the same, and it should never be subject to question if and when a borderline case gets elected.
   56. Marc Posted: May 22, 2003 at 06:50 PM (#512270)
I don't like the subjectivity that is being injected here. It's not a question where 10 and 5 came from (thin air), it's a question of where they ended up (in our constitution).

I don't like the possibility (and I am not accusing anybody, but the possibility exists nevertheless) that somebody who really really likes Sam Thompson might look at the upcoming eligibility list and decide he has a lot better chance of getting elected in 1902 than 1903, or 1903 rather than 1904, or whatever, and THEN slot him into that more favorable spot. We DO need a hard and fast rule that does not respond to subjective judgements of whether he "altered" his "record" or not, or what his age was, or what his injury status was. Will we always know injury status?

I think it should be by the numbers, it should always be the same, and it should never be subject to question if and when a borderline case gets elected.
   57. Marc Posted: May 22, 2003 at 06:53 PM (#512271)
And that goes double ;-)
   58. RobC Posted: May 22, 2003 at 07:29 PM (#512273)
Joe,

I disagree. He wasnt done after 1896. He had a not insignificant portion of his value left. His 1898 was twice as valuable (in warp3) as
   59. DanG Posted: May 22, 2003 at 07:41 PM (#512275)
I agree with Joe, it's a playing time issue. That Thompson played well in his brief stint in 1898 is interesting but not statistically significant. From my posting here on April 17:

The spirit of the rule, I think, is to allow players' eligibility as soon as (five years after) they are finished contributing significantly to their careers. Joe and I seem to agree that this is the idea, as we both advocate earlier eligiblity for the players in question. For Thompson, 3 (0 WS) and 14 games (3 WS!) adds nothing to the weight of his accomplishments or career totals.

I don't believe in a hard rule. I thnk we're smarter than that, that the ballot committe can make more reasonable decisions for eligibility. For Thompson and others, I see no reason for delaying eligibility until five years after the last drop is squeezed out of their careers.
   60. RobC Posted: May 22, 2003 at 07:46 PM (#512276)
It shows that 14 games is long enough to not necessarily be meaningless by default. The way the constitution reads, if <10 games, we can assume its token and if its 10 or more "the committee" should use judgement. Well, in this case, its clear that 14 games were more than a token appearance, based on how well he played. If his 14 games had been 18 at bats, I would have no problem with it being token. The 63 at bats is the biggest thing for me, he was trying to be a full time player, and in fact, actually did it well.

If he had sucked it up, even in 63 at bats, I would be willing to call it token. But, I dont want to call a player done, until he's done. Let him prove he's done by either not playing or sucking.

I guess my question is: In what case what 14 games not be token, by your standard? We(you) put 10 in the constitution, but it looks like you are trying to move the default to 15. I cant think of 14 games a player could play that would look more like a real season than this.
   61. DanG Posted: May 22, 2003 at 08:05 PM (#512277)
Another possible solution. We could use a scale like this:

Season after playing 10+ games (say, 1897): <10 games that year
   62. MattB Posted: May 22, 2003 at 08:07 PM (#512279)
Joe wrote:

"It's got nothing to do with the value of the time, it's the amount Of time that matters."

Dan wrote:

"For Thompson, 3 (0 WS) and 14 games (3 WS!) adds nothing to the weight of his accomplishments or career totals."

I agree with Joe, and that's why I think 113 plate appearances (after 1896) has to be significant. In theory, he could have gone 0 for 113, or he could have hit 113 home runs in that time. If he had, in fact, gone 113 for 113, with a 4.000 SLG in that time, no one would question his greatness, and he'd certainly bump up a few notches on my ballot. If the game COULD make a difference, it should not matter whether they did or not.

1.8 Wins above replacement (after 1896) may not seem like a lot to a guy who has around 100 either way (1 or 2 percent of his total), but when comparing top candidates, all of whom generally fall within 20 wins either way, 2 wins becomes 10% of the difference.

It is entirely possible that Thompson could fall 10th on your ballot excluding post-1896 accomplishments, but 9th including them.
   63. RobC Posted: May 22, 2003 at 08:11 PM (#512280)
I think you misread what I wrote. I know this is a judgement call. I think you are trying to make it a non-judgement call. I want to know in what case you think a 14 game season would count as a season? If there isnt one, then all 14 game seasons are token and we move the line from 10 to 15 (which is okay, but we need to change the constitution). I think this is the about the most non-token a 14 game stretch can be. He was playing at a 16+ warp-3 pace!
   64. RobC Posted: May 22, 2003 at 08:20 PM (#512282)
MattB - you are right. In fact, in terms of career warp-3, 1898 moves Thompson from below Hardy Richardson to above him. Admittedly, a .2 season could do the same thing, but a 1.5 season is likely to do it, as close as these players are bunched together. And that was the first player I looked at to see if he passed. He moved above 100% of the players I checked him against. :) Now, that may or may not affect my ballot, but it could, because 1898 counts too.
   65. John (You Can Call Me Grandma) Murphy Posted: May 22, 2003 at 08:35 PM (#512287)
I don't know if I have a point here, but I'd like to point out that Thompson's 14 games of insignificance would equal a majority of a season for some of the players that he's competing for ballot spots against.

Except Thompson's season is comparable to a couple of games played before the NL. It's not really a good analogy, IMO.
   66. MattB Posted: May 22, 2003 at 08:37 PM (#512288)
1.5 WARP3 may not be statistically significant, but if could move a player from a statistical tie (say, 1.5 WARP3 difference) to a significant amount (3.0 WARP3 difference).

I don't know if Mark has a point either, but Lip Pike played 28 games in 1871,and I'm giving him a lot of credit for that.
   67. RobC Posted: May 22, 2003 at 08:44 PM (#512289)
Joe,

I get what you are saying. Really. I think we have a disagreement on what a token appearance is. Those 14 games in 1898 fit no meaning of token or meaningless that I can think of. Minnie Minoso in 1976/1980 is token/meaningless. I think the best way of looking at it is, can I drop the year from the career records and it not affect anything? In Minoso, the answer is yes, with Thompson's 1898 the answer is no. For 1897 and 1906, the answer would be yes.

Mark - I dont think you have a point, but it made me laugh.
   68. DanG Posted: May 22, 2003 at 09:28 PM (#512293)
Geez, the wheels of this discussion slipped the tracks somewhere along the line.

The quality of Thompson's performance in 1897-98 is totally irrelevant to the discussion of when he is eligible, because this play is being considered whether he becomes eligible in 1902 or 1903 or 1904.

The issue is more basic. Some players careers end abruptly, others dribble along to an end. I fail to see a compelling argument as to why these dribblings should delay the eligibility clock for a player.

Sometimes a player will fall below ten games played and come back to play fairly regularly.
   69. John (You Can Call Me Grandma) Murphy Posted: May 22, 2003 at 09:29 PM (#512294)
I don't know if Mark has a point either, but Lip Pike played 28 games in 1871,and I'm giving him a lot of credit for that.

Why would that matter, Mark? One would have to draw back his numbers towards the mean because of the short schedule, but the amount of scheduled games for that season shouldn't be discounted as somewhat less significant. Lip Pike had no say in that matter.

If you're arguing competition, that's a different story.
   70. Marc Posted: May 22, 2003 at 09:43 PM (#512295)
Mark wrote:

>I am concerned because I believe TomH's assumption about the year of eligibility not having
   71. RobC Posted: May 22, 2003 at 09:46 PM (#512296)
Sam Thompson:
   72. MattB Posted: May 22, 2003 at 09:49 PM (#512298)
Joe,

Hey, I was just looking at the intro article you wrote well over a year ago, and noticing how many things changed between now and then.

So, I was just wondering, what happened to your co-author Robert?
   73. RobC Posted: May 23, 2003 at 02:22 AM (#512301)
DanG -

14 games is absolutely worth delaying eligibility for if its not a player playing out his career, but a player adding value onto his career.

Intesting question: does this break down along career vs. peak lines? To someone voting primarily on peak, these 14 games clearly dont matter. But, as someone who mainly votes on career values, these 14 games can, and possibly do, make a difference to where Thompson is placed on my ballot. If it is going to affect the ranking, its worth waiting on. Otherwise, we might as well vote on the players as soon as their career value is high enough.
   74. RobC Posted: May 23, 2003 at 02:42 AM (#512302)
Im going to try to clarify the issue going on here, if Joe and Dan can answer some questions for me, I might be able to go along with the 1902 eligibility, but I want to see if I understand.

Here are some situations for some unknown (yeah right) players last 3 years:

a. 119 games - 3 games - 14 games
   75. John (You Can Call Me Grandma) Murphy Posted: May 23, 2003 at 05:22 AM (#512303)
I can't see any justification for having O'Neill ahead of Wright on a ballot.

O'Neill is trailing my boy Pearce, so that tells you what his chances for the Hall are. :-)
   76. RobC Posted: May 23, 2003 at 01:00 PM (#512305)
Joe,

At least we are in agreement on my a,c, and d scenarios. For yours, I would go with, b and d are 1905 and 1908 (assuming those 20 and 22 game seasons are 80 PAs and not 20 PAs). A and C would depend on the number and quality of PAs. They are clearly the cases that need to be looked at. I think ~70 PAs is enough to reset the clock, if player C got 70 PAs in 11 games, then reset it. If Minoso had 70 PAs in 1980, I would keep him off the ballot until 1986 (or possible remove him from the ballot in 1980 and then reput him on the ballot in 1986).

I guess my "rule" would be, if this player where a mid-ballot kind of guy, would his season possibly affect his ranking? In Thompson's case, on my ballot, he will be fighting Richardson and Galvin, not Dunlap and Browning, so 1904 value doesnt matter, but if he were a little bit less of a player, the value he got could make a significant difference in his ranking. That seems like enough to reset the clock to me.

I like the idea of using judgement in these kind of situations, but if Thompson's 14 games arent enough to reset the clock, then no ones 14 games will be, and we arent using judgement anymore. We just have an unwritten rule.

Anyway, I vote for 1904.
   77. MattB Posted: May 23, 2003 at 01:02 PM (#512306)
Two points (well, actually one point from two sides):

Either:

(A) The year of eligibility won't make a difference because Player X either will be DEFINITELY IN or DEFINITELY OUT no matter what year he is eligible; or

(B) Causing Player X to wait another year or two will prevent him from ever being admitted.

If (A), then the argument is all about nothing, and it doesn't matter what we decide. If (B), then isn't it best to decide to be conservative and not admit a borderline player who's case is so tenuous that one or two years of eligibility will make the difference between enshrinement and eternal damnation in the pits of "really good player" Hell?
   78. MattB Posted: May 23, 2003 at 01:31 PM (#512308)
BTW, there is a real life example coming up after Sam Thompson. HOFer Hugh Duffy's career ended like this (games played, 1900-1906):

55 - 79 - 0 - 0 - 18 - 15 - 1

Even though he had fewer PA's than Sam Thompson after the date of his "retirement", I think he's a stronger case for delaying eligibility.
   79. DanG Posted: May 23, 2003 at 03:24 PM (#512310)
Joe, we're in pretty close agreement. You pretty much nailed it when you wrote: "The key is getting perspective...."

One small disagreement with this notion on the hypothetical "D" Thompson:

"The logic on D is normally eligible in 1902, before he played the 20 games. Then since he played 20 games, I'd say delay a year and then put him back on the ballot."

I think players should remain eligible once they're on the ballot, not be pulled off and put back on. In the rare case of a Babe Herman or Jose Rijo (someone who returns more than five years later and plays more than a token amount) we simply need to remind our voters to disregard the as-yet-unplayed part of their career when considering their ballot. Anyway, I don't think there are any likely HoMers who returned like that.

As for Duffy, I could see a case for delaying him a year, until 1908. If he comes on in 1907, it's immediately after he just played a game. At that moment, we can't be really positive that he won't come back and alter our perspective of his career. Again, not too big of a concern as he's not a likely HoMer.

Finally, Thompson. In 1902, we can be pretty positive that he will not come back and alter our view of his career. The main reason we wait as many as five years is in case a player makes a comeback. Usually, such comeback attempts fail and the guy never plays again; in Sam's case, he did manage to eke out a few more hits. From an early-1902 perspective, those 17 games were a comeback attempt that ultimately failed: a 17 game/3 win shares contribution is hardly a return to being a regular contributor to the pennant push of the team.
   80. DanG Posted: May 23, 2003 at 04:25 PM (#512312)
While we're in the mood for debating eligibility, I want to mention Matt Kilroy again. While not a HoMer, he may garner a few votes from peak lovers. Here are his games played and games pitched in his last few years:
   81. Marc Posted: May 23, 2003 at 04:29 PM (#512313)
By all means, let's make this even more complicated than we have already made it. In fact, let's make sure that the average HoMer voter has absolutely no way of knowing who is eligible and who is not until "the committee" makes a declaration.

This is ALL going in the wrong direction.
   82. DanG Posted: May 23, 2003 at 04:30 PM (#512314)
Correction. The Koufax comment should have been something more like, "he was the youngest VIABLE CANDIDATE to appear on a BBWAA ballot." He was certainly not the youngest player.
   83. John (You Can Call Me Grandma) Murphy Posted: May 23, 2003 at 04:46 PM (#512315)
How about, for the questionable eligible players, we vote up or down for that upcoming election?

At any rate, I think we're making a mountain out of a molehill, gang. As long as Minnie Minoso and Satchel Paige do not make their first ballot when they're in their sixties, chill out.
   84. DanG Posted: May 23, 2003 at 04:55 PM (#512316)
Hey, Marc. To quote Murphy, "chill out".

Despite the sarcasm, your objection is noted. But c'mon, we're just talking here. Nothing has changed. If you can offer a bit more than "This is ALL going in the wrong direction" it owould be helpful.

BTW, when do people think Kilroy should be eligible? I assume 1904 is the preference.
   85. John (You Can Call Me Grandma) Murphy Posted: May 23, 2003 at 05:06 PM (#512317)
BTW, when do people think Kilroy should be eligible? I assume 1904 is the preference.

1904 would be my choice (not that he will make my ballot anyway).
   86. John (You Can Call Me Grandma) Murphy Posted: May 23, 2003 at 05:13 PM (#512318)
While not a HoMer, he may garner a few votes from peak lovers.

Since his peak wasn't anything special, I can't see him on a single ballot.
   87. DanG Posted: May 23, 2003 at 06:36 PM (#512320)
OK, here's a few more upcoming debates. I much prefer actual example to theoretical ones. I show them not to stir up s**t, but to demonstrate the real-life gray areas that arise, and to resolve them before they're up for election. IMO, some subjectivity is OK, which our Constitution allows for.

I assume that whatever is decided for Thompson will also apply to Jack Clements and Frank Chance, who are very similar cases.

How about these three similar cases: Cy Seymour, Kid Elberfeld and Orval Overall. They each were retired for two years, than came back to play kinda-sorta-but-not-too regular.

Seymour
   88. John (You Can Call Me Grandma) Murphy Posted: May 23, 2003 at 06:54 PM (#512321)
Seymour: 1915

Elberfeld: 1916

Overall: 1915

Donlin: 1917? (You got me on this one)
   89. DanG Posted: May 23, 2003 at 07:15 PM (#512323)
I concur also, making their proper years:

Seymour: 1916

Elberfeld: 1917

Overall: 1916

Donlin: 1918
   90. John (You Can Call Me Grandma) Murphy Posted: May 23, 2003 at 07:17 PM (#512324)
I agree with John on all of them, except that you are off by a year. The 5 year wait means we actually make players eligible in retirement year +6.

Oops! Yeah, I screwed up. So sue me! :-D
   91. Marc Posted: May 23, 2003 at 07:45 PM (#512325)
Dan, I have provided additional detail in the past and there has never been any indication the objections were noted....

But anyway, my criteria would be 1) the average slug can apply the rule without expert help and 2) there would be no possibility of an appearance that a player got preferential treatment--ie. slotted in a particular year--so that he would have a better chance of getting elected.

Those are the two specific problems with these complicated (possible) rules.

Am chilling.
   92. DanG Posted: May 23, 2003 at 08:09 PM (#512326)
Marc: chill noted.

As for your 1), that is a worthy goal, but IMO an unattainable one. The rule would tend to be too simplistic, so as to occasionally make obvious errors.

As for 2), there is no way to completely eliminate someone choosing to view some aspect of our procedures this way. We can always be accused of starting in the year we did just because we wanted to elect Soandso. Or we made the rule 10-and-5 just so we could get Whosisnuts on the ballot at a particular time.

So, yes, we want to avoid the appearance of impropriety. Have we yet suggested anything that would clearly validate such a charge?
   93. RobC Posted: May 23, 2003 at 08:27 PM (#512327)
1919

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Harveys Wallbangers
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 0.8448 seconds
49 querie(s) executed