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 03, 2002

More to Come Soon

Okay—we are going to get this rolling. One of the reasons we’ve been kind of slow on this is that I’ve been waiting for the Win Shares book to come out. I think that’s going to have a big impact on the Hall of Merit.

The other reason is that fantasy baseball has taken over the better part of the last two weeks, as I’m the commissioner and owner for baseball, hockey and basketball leagues and this is the busy season. After the draft this weekend my life will start to come back into order.

Anyway, I’m not saying the Win Shares method doesn’t have it’s flaws, especially the low replacement level. But once we know what the flaws are, we can adjust for them (and season length, which is important pre-1904)and we’ll really have a nice evaluative tool on our hands. I’m especially encouraged by the pitching/defense splits to see where the credit falls for historical players.

I’ve also been working on updating my career offensive W-L numbers w/some tweaks. Once that’s updated (hopefully by the weekend of the 19th) I’ll present numbers for every viable candidate on the ballot.

If any of you have questions, thoughts on the process, etc.; let’s get that discussion rolling as well. This is going to start picking up over the next few weeks, it’d be great if we could get the first ballot underway by Memorial Day weekend.

I’ll set up a few threads now to organize the discussion. One for ballot structure (relating to how points will be awarded), one for candidate discussion (let’s limit ourselves to players retired before 1900). If anyone has any other ideas for topics that need to be discussed, let me know and I’ll post them as well.

Joey Numbaz (Scruff) Posted: April 03, 2002 at 03:43 PM | 19 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. tangotiger Posted: April 08, 2002 at 08:33 PM (#509684)
James does not use "replacement levels". His totals add up to the absolute number of wins. He confuses everything, but would have cleared everything up if he simply also reported the equally valuable loss shares. I don't think that James has made any advancement with his win shares technique.
   2. scruff Posted: April 09, 2002 at 04:06 PM (#509685)
Tom, I checked out the Win Shares assessment on your home page, interesting stuff. Was that written before the book came out? I just received my book in the mail last night, and I noticed that he uses 1.52 and .52 instead of 1.5 and .5.

Overall I think the formula is very good, but with some flaws, most notably the lack of a replacement level. I think that's easy enough to overcome though. Just figure the replacement level per out and subtract this from the WS total based on the outs the player made. For pitchers, figure the replacement level per IP and reduce by that amount. I'm guessing it's about 8 WS per 162 games (454 outs) for batters.

I especially like that the book has numbers for all players ever, although he neglected the National Association. This a HUGE step forward, because of the different baseline (from Linear Weights), many questions can be answered that couldn't be before w/Linear Weights and it's average baseline. I buy into that premise, let me know if there is a reason I shouldn't.

I have a deeper question. My question regards allocating the pitching credit among the individual pitchers. James separates the pitching credit from the defense by accounting the K's, BB's, HR, etc.. But for individual pitchers he doesn't do this. So pitchers that play w/Randy or Pedro look better (because the team's pitchers get more credit) and pitchers that play w/junkballers lose credit.

Am I missing something here? Is there a way that K-rates, etc. are used in the individual apportionment indirectly?
   3. tangotiger Posted: April 10, 2002 at 01:59 PM (#509686)
Yes, I wrote that when the hist abstract came out.

Again, there's no need to discuss replacement level, since not everyone has use for those. By providingn win shares AND loss shares, the reader has enough information to decide if he wants to:
   4. tangotiger Posted: April 10, 2002 at 05:09 PM (#509687)
By the way, I don't understand how your point system is going to work. Say you give points 1 to 10. You get Brett with 10 points, Yount with 9, Fisk with 8, Guidry with 7, and Carter with 6. Some other people might have Carter at 10, and Yount at 3 points. In a ballot where you might have alot of qualified candidates, you might end up with Fisk getting written on 80% of the ballots, but only getting 60% of the maximum points. Are you going to account for this? Even worse, if you have alot of unqualified candidates, you might have some people put in a borderline guy at 9pts, and all of a sudden, a borderline candidate comes in at 75% of the points.
   5. tangotiger Posted: April 10, 2002 at 07:29 PM (#509688)
I see you guys have discussed this in the other thread. I like the idea of letting the voter split up his votes, whether to give 20% to the #1 guy, or 10% or whatever, to a max, say of 20%. But I still think that's alot of trouble. In/out would be easier.
   6. scruff Posted: April 10, 2002 at 08:28 PM (#509689)
Tango -- Let me clarify, we aren't using a 75% (or any number) standard.

We are taking the top X players for each "annual" ballot. The number will range from 2-5 depending on the year. In the early going, we'll start with 5 electees in 1906 (a catch-up election for the entire 1871-1900 period) and then we'll drop to 4 and 3 before finally settling at 2 per year around 1915 (I don't have the spreadsheet handy). Sometime around the 70's we'll start electing 3 every other year and eventually 3 every year then alternating between 3 and 4 per season, as more teams/players come into the league, etc..

So the top X candidates each year will be elected, regardless of the % of votes or points.

In/out would be easier on some level, but I definitely don't think it's the way to go. You have no room for degrees of voting. You have to ask the question, "what is a Hall of Famer" and that answer is different for everyone. Asking the question of who are the best players not enshrined is a much easier question for everyone to get on the same page with.

Also, players never lose eligibility. This is important, because as new data becomes available, some previously overlooked candidates may all of the sudden become viable. There is no reason to remove anyone from the ballot when you are asking who are the best available. By definition, the people who the group deems the best will rise to the top.
   7. tangotiger Posted: April 10, 2002 at 09:05 PM (#509690)
Ahh! Interesting... I think you mentioned that somewhere earlier as well. Might be good to get a recap of everything.

Any thought of doing a "pyramid" type of HOF, as some article suggested. This would answer your question about what is and is not a HOF, as there are by some people's perceptions degrees of HOF. This way, instead of having 12,000 on the ballot in 1992, you'll have filters in place. Just a thought...
   8. DanG Posted: April 11, 2002 at 02:45 PM (#509692)
I want to revisit the topic of how many we'll induct each year, which was discussed in an earlier thread. I don't recall that we really settled on a scheme, but scruff seems to have it clear in his mind: "We are taking the top X players for each "annual" ballot. The number will range from 2-5 depending on the year. In the early going, we'll start with 5 electees in 1906 (a catch-up election for the entire 1871-1900 period) and then we'll drop to 4 and 3 before finally settling at 2 per year around 1915 (I don't have the spreadsheet handy). Sometime around the 70's we'll start electing 3 every other year and eventually 3 every year then alternating between 3 and 4 per season, as more teams/players come into the league, etc.."

Catching up in the early years is fine, but if we're not careful we'll find ourselves electing players near the bottom of the Hall of Fame line in the early 30's. This could happen because there is an 8-year gap of good new candidates in our elections from 1925 to 1932. Check it out.

By 1900 there are 18 players retired who are in the HoF. Chronologically these are Cummings, Spalding, Wright, Radbourn, Welch, Galvin, Keefe, Kelly, O'Rourke, Clarkson, Ward, McCarthy, Brouthers, Ewing, Anson, Connor, Thompson, McPhee. Add in White, Caruthers, Browning, etc. and we have a bit of a backlog. Fine.

By 1910, another 20 HoFers have retired, so we can keep electing lots of guys to the HoM.

By 1918, another 15 HoFers have retired, but then it stops.

From 1919 to 1926 only 4 HoFers retired. (Baker, Hooper, Marquard, Youngs.) Three of these are weak choices.

Quality retirees pick up quickly after then. The question is, who will we have left as top candidates by our elections of 1930-32? By 1932, after 27 elections, the Hall of Fame has 57 players. If we plan to have many more than that, we'll be scraping the bottom of the barrel in the early 30's.

It's something to be aware of. We should be careful not to load up the HoM too much in the early years. The elections of 1925-32 will be the time when we earnestly debate the qualifications of the early marginal candidates.

Dan
   9. scruff Posted: April 14, 2002 at 02:36 PM (#509693)
DanG -- I'll give the current plan (I can send the spreadsheet we used to figure this if you'd like) in a second.

Basically, I overstated the number of candidates above for the catchup period of 1906-14. I was basing my numbers off of the original 1915 start date plan. Through earlier discussion on these threads we decided to drop back to 1905.

Because of this we'd have to drop back the early elections.

We based the number to be elected as a function of the teams in the league, with a 13 year lag (middle of a 16 year career plus the 5-year wait). We didn't strictly use the number of teams for the 19th century, we used:

1876-1884: 8
   10. DanG Posted: April 15, 2002 at 03:04 PM (#509696)
Scruff,

I think your scheme is logically based and it should work fine through 2001. It gives us 52 HoMers through 27 elections, in no danger of scraping the bottom for good candidates.

As for 2002 and beyond, I don't agree with the scheme. I think we should always maintain our membership at the same level as the HOF.

ChapelHeel:

The reason we choose the same number as the HOF (~213 now) is because that defines the quality of players we elect. The HOF itself has never defined how exclusive it wants to be, so we can never say whether anyone they elect doesn't deserve it.

One of the things we want to know in the end is, who exactly are the HOF "mistakes" as well as omissions. By making our Hall exactly as inclusive as Cooperstown, we will have an answer to this question.

As for electing a player at each position the first year, I originally proposed this, as well. While there is a certain appeal to creating this "All-19th Century Team", there are problems also. The main one is you're electing players out of order of their true merit. You would have to elevate a secondbaseman and a thirdbaseman above Brouthers or Anson, since only one of them is the top firstbaseman.

Dan
   11. scruff Posted: April 15, 2002 at 04:58 PM (#509697)
ChapelHeel and DanG -- I think you guys are forgetting one thing -- The expanding population. There are more people now (both U.S. population and the foreign players) than ever before. I think it logically follows that we will have more HOMers.

In the 30's what was the US population? 125-150 million? Now we are pushing 300 million. I think we need to account for that.
   12. DanG Posted: April 15, 2002 at 06:06 PM (#509698)
When you start talking about the population pool available to MLB baseball I think you have to consider a vast number of issues besides what the U.S. population happens to be. I'm doubtful that accurate comparisons can be made across time.

What else is there to consider? Things like how many people are seriously trying to attain professional status in the sport? What percentage is likely to be unscouted by MLB? How many are unwilling to leave their home for an uncertain existence in pro ball? How much talent drain is occurring from other professional sports? What foreign lands do we need to include in this assessment and what assumptions must we make about their populations? How have medical advances contributed to the available talent pool? Professional athletes have always come from the lower socio-economic classes. Shouldn't we account for this in some way?

The ground is constantly shifting, the answers to these questions continually evolving and very different now than in previous generations. Indeed, many are practically unanswerable.

For the HoM, I think it makes sense to tie the number of members to the number of major leaguers, not US population. Indeed, your system pretty much does this.

But I think it's better if we keep our overall membership in line with that of the HoF. If our number elected becomes much higher than the HoF, it's harder for us to argue for players as deserving to be in the HoF.

I think there is a semi-serious side to the HoM, that there is a mission to try and right some of the errors of omission perpetuated by the HoF. Perhaps in some small way, through our thorough justification of worth, players that we elect to the HoM, that the HoF have missed, will finally be given their due from Cooperstown.

Dan
   13. MattB Posted: April 15, 2002 at 07:29 PM (#509699)
I agree with DanG.

If, after a few years, there HOM is electing 5 people per year, and the HOF only one or two, the original goal of comparing the top X players as chosen by the different groups becomes meaningless. THe HOM will eventually have twice as many inductees.

Also, the system has a built in assumption that voters will select many more recent players. Assume, that for the mid-90s elections, the top 4 or 5 players turn out to be players from the 1910s. Stars of the '80's aren't elected because the voters examine the data and decide that the previously unelected Jack Fournier, Bobby Veach, Fred Luderus, Larry Doyle, and Gavvy Cravath are more worthy marginal candidates on the strength of 1915 seasons that were overshadowed at the time by Ty Cobb.

But that's not the assumption. The assumption going in is that there are more good players now than then, and in exactly the proportions that will be elected. But that's putting the cart before the horse. The rules are assuming the final complexion of the outcome. They are assuming that people will put more weight on the tougher competition that a world market presents today, and not more weight on the ability to lead a team singlehandedly to victory. The rules assume that players will vote the fifth best player of today over the third best player of then.

I know, the response is that the older players will never lose eligibility, and therefore can be elected at any year. But if you really thought that a player from 1901 would still be viable in 2001, the rules would be set up to permit an earlier entry. Expanding the number of entrants later on is based on the assumption that there will be more good players later, which is, I thought, what this exercise was attempting to establish.
   14. scruff Posted: April 15, 2002 at 08:55 PM (#509700)
"The assumption going in is that there are more good players now than then, and in exactly the proportions that will be elected. But that's putting the cart before the horse. The rules are assuming the final complexion of the outcome. They are assuming that people will put more weight on the tougher competition that a world market presents today, and not more weight on the ability to lead a team singlehandedly to victory. The rules assume that players will vote the fifth best player of today over the third best player of then."

Actually Matt, that not really what we were assuming. The idea was:

1) come up with an objective criteria for the number elected.

2) try to give the voters the most flexibility

So we aren't assuming anything. It's more conservative to have more players in the later elections because more players are available, which gives more choices.

If you allow too many in early, you are forcing electors to take candidates from players up through that time. This is one of the reasons we moved the first election back to 1906 -- so we force electors to choose players from the 19th Century.

By allowing more people to be elected in later elections, you are giving voters more of a choice on who they elect, not less.

As far as a player from 1901 not being elected in 2001, yeah that probably won't happen. But it's not unreasonable that a player from 1955 could be elected. That's what the "slow" years (without Ruth/Gehrig/DiMaggio's, etc.) are for, to catch up on the borderline candidates.

I know my personal decisions will not be too biased in terms of current players. I don't believe in the time machine concept (what would this player do if you transported him back or forward in time), I think you have to look at how the player dominated the competition he played against as the main factor, with some room for leeway on adaptability of skills. Sure the Jim O'Rourke of 1890 would have trouble w/Randy Johnson, but not as much if he had grown up in the (19)80's and been exposed to the advancements in the game from an early age.

Recently the Hall of Fame has been electing about players a year:

2001 - Winfield, Puckett, Mazeroski, Hilton Smith
   15. scruff Posted: April 15, 2002 at 08:57 PM (#509701)
I should say "Recently the Hall of Fame has been electing about THREE players a year:"

Somehow I missed that word . . .
   16. MattB Posted: April 15, 2002 at 09:11 PM (#509702)
So, when will the first ballot be available? And will tuxedoed men from PriceWaterhouseCoopers by tabulating them?
   17. scruff Posted: April 16, 2002 at 03:12 PM (#509703)
I'm shooting for Memorial Day Weekend. I'm trying to get WS numbers adjusted to a 162-game season for every player retired before 1901. I'm almost done with the "B's" going alphabetically. By far the 3 best candidates so far are Cap Anson, Charlie Bennett and Dan Brouthers. No one else is really close, maybe Tommy Bond.

So the answer is soon. Hopefully I can finish this over the next two weeks. It's a good starting point, I don't mean for it to be the be all and end all if that's how I'm making it sound. But it will be nice to present a decent resume for the top candidates.
   18. MattB Posted: May 24, 2002 at 08:33 PM (#509704)
"I'm shooting for Memorial Day Weekend."

I guess you overshot. :-)
   19. scruff Posted: June 07, 2002 at 03:34 PM (#509705)
Yes MattB, we overshot.

Robert is transitioning his move back from Europe, and I've been busier than expected.

I'll be starting a new thread today, asking for much needed help. We appreciate your (and everyone's patience).

I realize I've possibly cried wolf here, but more will come soon. I'll be posting the new thread later today, and we can finally start moving this forward.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Adam M
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 0.3925 seconds
49 querie(s) executed