Hot topics.

Is this HOM voters only, or open to the entire BBTF universe?

And are we voting based on HOF rules (including character, sportsmanship, fame, etc), or strictly on "merit"?

Hall of Merit voters only Howie. New voters can join, but with the same restrictions as for normal ballots.

Ballots will still have to be justified, etc..

Our vote is strictly on Merit.

Wow, that's a tough ballot.

I think this is "HoM" voters only (which only means if you haven't voted previously we need a prelim). I think we're voting by HoM rules.

1) Bert Blyleven - pitcher of the same caliber as Niekro and Perry which is really good.
2) Tim Raines - As good as Tony Gwynn, best leadoff hitter in NL history?
3) Alan Trammell - A median HoM SS.
4) Lou Whitaker - A median HoM 2B.
5) Dwight Evans - Strong defense, well rounded player
6) Mark McGwire - Below average defender and baserunner, often injured, massive power
(Rick Reuschel)
7) Bret Saberhagen - I like the peak he has better than Stieb.
8) Will Clark - Better hitter than Hernandez, not quite the glove
(Lee Smith)
9) Keith Hernandez - Best fielding modern 1B, great OBP
10) Dave Stieb - Solid pitcher from the 1980s, why do people like Jack Morris better?
(Tommy John)
11) Willie Randolph - Worst 2B in my PHoM, defines the in-out line.
(several players)
12) Andre Dawson - OBP is too low, how could he possibly rate ahead of Raines?

Dawson is the only one not in my PHoM.

I've got one obvious placement: I placed 11 of these 12 players on my HoM ballots, so the one I didn't would have to be 12th. That would be Dawson. In his case, I never could get past the low OBP. (Or, as I phrased it, he made too many outs.)

DL from MN's list looks like a good place to start. Do I have any disagreements with that? Well I'd put Stieb ahead of Saberhagen, that's one. And I'll have to at least take a close look at McGwire versus Evans and Whitaker.

For measures of consensus and/or inter-voter agreement, the fact that every voter must vote for all 12 enables to me to use mathematically simpler models - I'm thinking of something closely related to correlation coefficients.

Just rough first guesstimate:

1. Tim Raines (2008)
2. Dwight Evans (1997)
3. Alan Trammell (2002)
4. Lou Whitaker (2001)
5. Bert Blyleven (1998)
6. Mark McGwire (2007)
7. Bret Saberhagen (2008)
8. Dave Stieb (2002)
9. Will Clark (2006)
10. Willie Randolph (2001)
11. Andre Dawson (2005)
12. Keith Hernandez (1996)

I'm not 100% sure yet, but that's my gut.

I would add that I do think Raines is a pretty easy #1 choice.

HOF rules (including character, sportsmanship, fame, etc),

HOF rules don't include fame.

A quick prelim:

1. Raines
2. Blyleven
3. Clark
4. Trammell
5. McGwire
6. Dawson
7. Evans
8. Whitaker
9. Hernandez
10. Stieb
11. Saberhagen
12. Randolph

Randolph's the only one not in my PHOM.

Is this a prelim or a one-shot final ballot? And when does balloting close--1 week or 2???

I have them in 3 categories:

Guys I support:

Raines
Blyleven
Dawson
McGwire

Guys I think are close (highlights my bias against middle infielders):

Trammell
Whitaker
Evans
Randolph

Guys I don't think are close:

Stieb
Saberhagen
Clark
Hernandez

Picking the order within those three groups will be tough.

Addressing #12 . . . This is just the discussion thread. Balloting will be next week.

"I do think Raines is a pretty easy #1 choice"

Among position players, I agree. I think Blyleven is in a higher percentile among pitchers than Raines is among outfielders.

Here's one of my offense-only charts, in the same style as I've always done.

Since the base of this is a specific book whose last recorded year was 1999 all years from 2000 on are somewhat cruder estimates. This affects only McGwire and Raines, and only some of their less valuable years, so it doesn't matter much.



Some of the usual caveats: this looks at offense only, using RC and outs. Since the baseline is the average hitter (still a fairly low baseline for 1B and corner outfielders), it undervalues durability. I suggest making extra allowances for games played and durability on top of this. The sorting has removed the information about which years were consecutive; this matters the most in the case of McGwire.

The following is an oversimplification, but I'll say it anyway:

Evans is a career candidate.
Raines is a prime/career candidate.
Hernandez is a prime candidate.
McGwire is a peak/prime candidate.
Clark is peak candidate.

I don't know how many of you are reading the McGwire thread now raging in the newsblog. For my own sake, I really don't see how you can say "his numbers weren't good enough." Of course, Clark's numbers are pretty darn good if only you read them the right way.

And here are the pitchers with their RA+ equivalent records arranged by year, with the years sorted by year-by-year equivalent FWP. In each case below, the three numbers under each pitcher are W-L FWP.

And one more chart, this time for the infielders.

From #17: Blyleven doesn't quite have the record of Stieb and Saberhagen added together. The arithmetic is more like Blyleven = Stieb + Saberhagen - two super Cy Young seasons. ("Super Cy Young" being 21-6, better than any of the three ever did.)

I don't see any way not to have Blyleven ranked first of our three pitchers. Blyleven vs. Raines is a whole different argument, and what I've presented doesn't help you much with that. And Stieb versus Saberhagen seems to be a fairly close call.

Looking at all of those 18-21 equivalent FWP seasons lined up, I guess that means we can think of Blyleven as a prime candidate -but what is that, a 10-year prime? Or if you go down to the 17-13 and 15-12 equivalent seasons, a 15-year prime?

My prelim:

Raines
Blyleven
Clark
Trammell
McGwire
Saberhagen
Stieb
Hernandez
---[my pHOM line at present]
Dw. Evans
Whitaker
Dawson
Randolph

"HOF rules don't include fame."

Right, by that I meant that 'fame' plays into how Hall of Fame voters vote. But it's not a rule.

This is a rough ballot for Randolph, who as we know staggered in with a small pct of the vote in a bad year. He's a guy we took only because we're taking a full set of 235 or so.

Ignoring steroids effects, my first pass would be:
Blyleven
McGwire
Trammell
Raines
WClark
KHernandez
Stieb
Saberhagen
DwEvans
Whitaker
Dawson
Randolph

But boy, this is a challenge.

"Raines is a prime/career candidate."

I'd call him a peak/prime candidate. 1983-87, and a couple of other nice ones.

Prelim.

1. Blyleven
2. Trammell
3. Raines
4. McGwire
5. Whitaker
6. Evans
7. Saberhagen
8. Clark
9. Dawson
10. Randolph
11. Hernandez
12 Stieb

I propose to use the rankings I had for these guys when they were elected (this won't work properly for Groups 2-4, but works fine for this one, which is temporally concentrated. Thus my ballot for this election would be:

1. Blyleven 1998 2
2. McGwire 2007 4
3. Raines 2008 5
4. Whitaker 2001 6
5. Trammell 2002 9
6. Saberhagen 2008 11
7. W. Clark 2006 14
8. Evans 1997 16
9. Dawson 2005 20
10. Stieb 2002 23
11. Hernandez 1996 61
12. Randolph 2001 75

I will insert the justifications I had on their respective ballots for each one when we come to a final vote, but I think I'm pretty happy with this ranking overall.

For now:

(1) Raines
(2) Blyleven
(3) McGwire
(4) Trammell
(5) W. Clark
(6) Hernandez
(7) Stieb
(8) Whitaker
(9) Evans
(10) Dawson
(11) Saberhagen
(12) Randolph

Here's my preliminary ballot, brought over from the earlier thread, with the fortunate Rich Gossage removed and some numbers added:

1) Bert Blyleven -- 370
2) Tim Raines -- 342
3) Mark McGwire -- 303
4) Dwight Evans -- 290
5) Lou Whitaker -- 279
6) Alan Trammell-- 267
7) Will Clark -- 274
8) Keith Hernandez -- 259
9) Bret Saberhagen -- 229
10) Dave Stieb -- 228
11) Willie Randolph -- 264
12) Andre Dawson -- 249

The numbers are my system's composites. The in-out line for position players is about 240, for pitchers about 225. You can see that I have not followed my system's results precisely. I definitely need to think more on Blyleven vs. Raines for the top spot. Blyleven in the aggregate looks as good as Steve Carlton and Gaylord Perry in my system, but I'm not sure he was quite _that_ good. It's hard to judge his continued excellence in the lower-usage 1980s vs. their monster peaks in the innings-happy 1970s. If Blyleven is that good, then he merits the top spot. I'd be interested in seeing how OCF's numbers show him in comparison to his 300-game-winning contemporaries. I'm curious why Joe has him below guys like Trammell and Whitaker. I really don't see that: Blyleven was as good a pitcher at his best as they were as position players, and he retained his quality (and durability) a lot longer.

McGwire is a solid #3 and Evans a solid #4. The relative order of the three first-basemen is clear. But I'm not sure how to rank Clark/Hernandez vs. Whitaker/Trammell.

I may be too hard on Willie Randolph: perhaps I am swayed by the consensus about him too much. By the numbers, Saberhagen and Stieb should be last, but I don't think we yet have a really solid way to assess the difficulty of being an outstanding pitcher in the 1980s, so I won't put them at the bottom of my rankings. Dawson clearly brings up the rear among position players, so I'm comfortable having him last overall. I see him as a deserving HoMer, but it's nevertheless ironic to me that the player who is likely to finish dead last in our assessment of the HoM-not-HoF, BBWAA jurisdiction players got more support than any of the others in the BBWAA vote, with many of these players having dropped off their ballot. How pathetic is that?

I may be too hard on Willie Randolph:

Not hard enough would be more like it. And, what Howie said.

I think there's a bit of revisionist history going on in the Randolph bashing:

Howie wrote and sunnyday2 concurred:

This is a rough ballot for Randolph, who as we know staggered in with a small pct of the vote in a bad year. He's a guy we took only because we're taking a full set of 235 or so.

This may be true, but if it's true for Randolph it must be even truer of Dave Stieb, whom Howie ranks 5 places ahead of Randolph but who finished well behind Randolph when they were on the ballot together in 1998, 1999, 2000, and 2001. Stieb was in turn elected in 2002, well ahead of Dawson, who was eligible that year for the first time. Now, the electorate is often cautious with new arrivals, so Dawson probably placed lower than his true rank that year, but he never passed the intensely debated Pete Browning in the balloting, and Stieb had finished well ahead of Browning.

So the voting history suggests that in the HoM electorate preferred Randolph to Stieb to Dawson, and the phrase "He's a guy we took only because we're taking a full set of 235 or so" applies with nearly equal justice to all three.

The results of this ranking exercise may turn out to be different, but let us not pretend that Randolph received weaker support in the actual election process than Stieb or Dawson did.

Nobody's running down Lou Whitaker, and, back in the day, there were some pretty tightly constructed arguments that Randolph was as good or better than Whitaker, which it might be worth re-visiting now.

If I get a ballot together, which will be valid only if I provide substantive comments this week, it will match Chris Cobb at the granularity shown here by quoting him and inserting three blank lines. There will be no change between groups, only within groups. So McGwire will be number three.

1) Bert Blyleven -- 370
2) Tim Raines -- 342

3) Mark McGwire -- with three capital letters in his name, where else but third?

4) Dwight Evans -- 290
5) Lou Whitaker -- 279
6) Alan Trammell-- 267
7) Will Clark -- 274
8) Keith Hernandez -- 259

9) Bret Saberhagen -- 229
10) Dave Stieb -- 228
11) Willie Randolph -- 264
12) Andre Dawson -- 249

I have a semester ending this weekend or Friday, then expected travel, so it's iffy.

I'll be about the 100th person to second a Chris Cobb comment over the HOM years - even responding to my own off-hand comment!

One of the things I will be doing is re-checking my own HOM ballots and rationales on these guys.

But even if I did like Randolph a little less than some others (and I had all of these guys on my ballots when they were elected, except maybe Dawson), I shouldn't stick a shiv in Randolph's tally vs other candidates here too casually.

On the 1998 ballot, Howie was tied with jimd for being Stieb's best friend. Howie had Steib 3rd on his ballot and Randolph 12th. I was one of Stieb's better friends - I had Stieb 6th and Randolph 12th. This does suggest that the ranking Howie implied with his comments is consistent with his own earlier ranking.

From that 1998 ballot:

Voters with both Randolph and Stieb, with Randolph ahead of Stieb:

ronw, rawagman, Mike Webber, Chris Cobb.

Voters with Randolph on ballot, Stieb off ballot:

DL from MN, TomH, Rusty Priske, Juan V. Thane of Bagarth, rico vanian, Rob Wood, Al Peterson, DanG, Joe Dimino, SWW, DonF, KJOK, EricC.

Voters with both Randolph and Stieb, with Stieb ahead of Randolph:

favre, OCF, Jim Sp, Howie Menckel, Got Melky.

Voters with Stieb on ballot, Randolph off ballot:

Andrew Siegel, Sean Gilman, Eric Chalek, jimd, Mark Shirk, Tiboreau, Devin McCullen.

Voters with neither on ballot:

John Murphy, karlmagnus, Daryn, sunnyday2, Mark Donelson, Andrew Schafer, yest, AJM, Rick A, dan b, Michael Bass, Andrew M, Esteban Rivera, Chris Fluit, Ken Fischer, David Foss, mulder & scully, Brent, Max Parkinson.

---

Of the first four groups, a tiny number may have changed their minds about the order and some may fail to vote in this upcoming election. But I doubt that that many minds have changed. More interesting is the fifth group, who hadn't previously ranked Stieb versus Randolph (although some may have done so by listing a vote order beyond 15 - I didn't record that.)

---

In my opinion, the only two rank orders that are beyond question are that Blyleven > the other two pitchers and Whitaker > Randolph. Many other things can be argued.

I've not participated in the HoM voting before, so I'm going to have to come up with a methodology on the fly. It'll be interesting because I don't have the math skills many of you have. I saw the word "coefficient" earler, and while I've heard of the term, I have no idea what it means. Nor do I want to know :) I'll turn in a preliminary ballot sometime this week.

I tend to look at peak more than career, particularly if the career was long and not particularly distinguished. (yes, I'm one of the few that doesn't think Blyleven is a HoFer). He is, however, head-and-shoulders above the other two pitchers in this group. I think part of my method will be, somehow, to compare each player's season with the top players in their league to determine some kind of rating.

Gives me something to think about when I should be working (or, like now, sleeping).

The results of this ranking exercise may turn out to be different, but let us not pretend that Randolph received weaker support in the actual election process than Stieb or Dawson did.

As OCF points out re. Howie, I was also talking about my own eval of Randolph which is non-PHoM by a fairly wide margin. Stieb is also non-PHoM but not so far off. I wasn't commenting on HoM voting generally. Once upon a time I would have known what that was, BTW, now it's getting to be pretty much ancient history.

I think I would be consistent with my HoM ballots if I had Trammell > Whitaker > Randolph, which I will. The three 1B are interesting and I think McGwire > Clark > Hernandez. And Raines >>> Dawson (if not >>>>), and Blyleven >>> Stieb and Sabes. Now to integrate the different lists.... I thnk the four #1s on each list will be 1-2-3-4. McGwire > Blyleven > Raines > Trammell???? That's not an easy one.

"Coefficient" means working well with others, I think. (Peak voters always welcome, math skills not required.)

Apparently everything is > DwEvans.

:)

"I saw the word "coefficient" earler, and while I've heard of the term, I have no idea what it means."

You should sue your school district. In (8 x Randolph) the coefficient of Randolph is 8.

I have changed my mind on Randolph (dropped due to Dan R's data) and Stieb (raised slightly) since those ballots.

This is a pretty rough prelim. I'm not sure I will do the work to get it more substantial or not (If not, I won't actually vote.)

1. Tim Raines
2. Bert Blyleven
3. Alan Trammell
4. Mark McGwire
5. Andre Dawson
6. Lou Whitaker
7. Bret Saberhagen
8. Keith Hernandez
9. Dwight Evans
10. Will Clark
11. Willie Randolph
12. Dave Stieb

Stieb is the only one not in my PHoM (which is odd, because I was a big fan.)

My preliminary ballot is fairly similar to Mark Donelson's; however, I'm finally in the process of incorporating Dan Rosenheck's work so things may change by next week, particularly on the bottom half of my ballot.

Tim Raines
Bert Blyleven
Alan Trammell
Mark McGwire
Will Clark

Dave Stieb
Bret Saberhagen
Lou Whitaker
Keith Hernandez
Willie Randolph
Dwight Evans
Andre Dawson

Years they were on HOM ballots (* if elected that year)
in order of finish by year

1996 - KHernandez*
1997 - DwEvans*
1998 - Blyleven*, Randolph, Stieb
1999 - Randolph, Stieb
2000 - Randolph, Stieb
2001 - Whitaker*, Randolph*, Stieb
2002 - Trammell*, Stieb*, Dawson
2003 - Dawson
2004 - Dawson
2005 - Dawson*, Saberhagen
2006 - WClark*, Saberhagen
2007 - McGwire*, Saberhagen
2008 - Raines*, Saberhagen*

reviewing my actual ballots, I now move Whitaker from 10 to 6, Stieb stays 7, KHernandez from 6 to 8, Saberhagen from 8 to 9, DwEvans from 9 to 10, Randolph from 12 to 11, and Dawson from 11 to 12:

Blyleven
McGwire
Trammell
Raines
WClark
Whitaker
Stieb
KHernandez
Saberhagen
DwEvans
Randolph
Dawson

Initial ranking:
I might revise the rankings within each group,
but doubt that players would move between groups.

1) Blyleven
2) Raines

3) Hernandez
4) Whitaker
5) Stieb
6) Trammell

7) Saberhagen
8) Clark
9) McGwire

10) Randolph
11) Evans

12) Dawson

My prelim ballot: I don't know how much info is needed for a prelim ballot, but here's how they rank.


1. Raines: Ahead of McGwire and Blyleven by quite a bit.

2. McGwire: Less than 1% difference between McGwire and Blyleven. I said I liked peak over career.
3. Blyleven: Ok, I was wrong, Blyleven is a HoFer. His career reminds me of one season when I bowled. I averaged 210 with a high game of 258 and a low game of 165.

4. Trammell: With positional advantage, he edges ahead of Dawson.
5. Dawson: Dawson gets credit for power/speed that overshadows his poor OBP.
6. Stieb: A better pitcher than I'd have given him credit for. About 3% better than Saberhagen

7. Saberhagen: Just edges out the next two. Will continue to review my system as regarding ranking pitchers and hitters together.
8. Whitaker: I was surprised to see Sweet Lou and Dewey in a virtual tie.
9. D. Evans: I'd have thought he'd rank higher, but the difference between Raines and Blyleven is about the same as the difference between Blyleven and Dewey.

10. W. Clark: One of my favorite players of all times. I was distressed to see him rank this low until I really took a hard look at how short his career was compared to the others.
11. Randolph: His positional modifier dropped him almost exactly between the two 1B on either side of him. No peak to speak of.
12. Hernandez: He's Will Clark +(career value-peak value).

Edit on #41.

A poor choice of words on Randolph. He'd have been quite a bit below the others without the positional modifier. The modifier lifted him into 11th on the list. I'm quite surprised he's in the HoM.

pocket8pin:

You're using certain language there - "less than 1% difference," "about 3% better," "positional modifier" - that hints that you are trying to create a single numerical scale on which to place these candidates. Can you tell us what how you put together that number? What goes into it?

This is what I said above:

Guys I support:

Raines
Blyleven
Dawson
McGwire

Guys I think are close (highlights my bias against middle infielders):

Trammell
Whitaker
Evans
Randolph

Guys I don't think are close:

Stieb
Saberhagen
Clark
Hernandez

This is what I decided on:

1. Raines
2. McGwire
3. Blyleven
4. Evans
5. Dawson
6. Clark
7. Whitaker
8. Trammell
9. Randolph
10. Hernandez
11. Stieb
12. Saberhagen

Strangely, I found the top 3 to be harder to order than the rest.

OCF,

I'm using OPS+ as a base for the hitters, looking at 3 yr and 6 yr stretches for peak and career OPS+ for length. I'm adding in modifiers for defensive position and SB. Since I think one should have excelled against their peers as well, I'm using the black and grey ink numbers as well.

For pitchers, I'm using ERA+ as a base, also considering 3 and 6 yr peaks. I'm factoring in career totals and winning pct and the black and grey ink tests.

Since others have included their composites, I'll put in mine as well. I'm please to say that in comparisons to other ballots, mine doesn't stick out like a sore thumb.

1. Raines 526
2. McGwire 439
3. Blyleven 436
4. Trammell 384
5. Dawson 374
6. Stieb 367
7. Saberhagen 352
8. Whitaker 348
9. D. Evans 347
10. Clark 335
11. Randolph 333
12. Hernandez 330

Quite the challenging ballot here. There seems to be some consensus who are the top 3 spots then more divergence. Here's where I come out when the numbers are plugged in.

1. Blyleven
2. Raines
3. McGwire
4. Clark
5. Whitaker
6. Trammell
7. Hernandez
8. Stieb
9. Randolph
10. Dw. Evans
11. Saberhagen
12. Dawson

It's assumed you have to put Whitaker and Trammell next to each other, wasn't it? My surprise was Dewey sliding down the ballot. Then I looked again at his career and it is not so impressive when you're comparing to the guys in this list.

For reference, here's the WARP3 numbers

Blyleven 146.1
Trammell 129.4
Whitaker 129.3
Raines 123.9
Dw Evans 120.2
Randolph 112.8
Hernandez 112.5
McGwire 109.3
Dawson 105.3
W Clark 105.2
Saberhagen 93.6
Stieb 92.4

And here's the position players in Dan R's WAR (out of my spreadsheet so I've removed bad seasons).

Trammell 82.1
Raines 78.5
Whitaker 71.1
Dw Evans 68
McGwire 65.3
Randolph 59.7
W Clark 58.8
Dawson 55.5
Hernandez 55.3

McGwire doesn't come anywhere near Trammell or Whitaker in either measure. Trammell is also consistently ahead of Whitaker (which is the HoF consensus also). I'll leave the peak/prime analysis to someone else.

If DanR is lurking, what are the batting/baserunning/defence breakdowns for Trammell and Whitaker in your WAR?

Player BWAA BRWAA FWAA
Trammell 16.7 3.4 4.7
Whitaker 28.3 2.2 3.9

Prelim:

1. Bert Blyleven
2. Tim Raines
3. Will Clark
4. Keith Hernandez
5. Dave Stieb
6. Bret Saberhagen
7. Alan Trammell
8. Mark McGwire
9. Lou Whitaker
10. Andre Dawson
11. Dwight Evans
12. Willie Randolph

McGwire looks a lot worse than I remember. He's the one who stands out from my HOM ballots.

His career reminds me of one season when I bowled. I averaged 210 with a high game of 258 and a low game of 165.

Interesting. I think McGwire matches me from last season. I had a 206 average, a high game of 288, a low game of 123, and I missed time with injuries.

Yikes. You guys are good bowlers!

For contrast here's big Mac

Player BWAA BRWAA FWAA
McGwire 60.6 -2.1 -0.6

Yikes. You guys are good bowlers!

Comes from having parents who owned a bar across the street from a bowling alley. Might also explain why I'm a good bowler and a lousy drinker. Not enough practice.

You should sue your school district. In (8 x Randolph) the coefficient of Randolph is 8.

Nah, I'm 47 and have a job where little math is required. I've efficiently forgotten things like co-efficients, which I probably knew in high school. Then again, since I got a D in high school algebra, I probably never did understand coefficients.

Yum! The Hall of Merit is back!

Going straight down my dollar values....

1. Trammell, 250
2. Blyleven, 242
3. Rock, 230
4. McGwire, 188
5. Whitaker, 183
6. Evans, 175
7. Clark, 162
8. Saberhagen, 158
9. Dawson, 156
10. Hernandez, 156
11. Stieb, 151
12. Randolph, 151

If I miss the actual vote someone please post this.

And here's the position players in Dan R's WAR (out of my spreadsheet so I've removed bad seasons).

Trammell 82.1
Raines 78.5
Whitaker 71.1

Wow, that's nothing like DanR's dollar value, where Whitaker is at 70% of Trammell. I suppose that is DanR's personal, not WARP, preference for peak seasons. Whitaker bowled a lot more like pocket8pin than like AJM.

Except Trammell bowled all 12 frames more often than Whitaker, who sometimes had an opposite-hand bowler toss for him...

Nah, I'm 47 and have a job where little math is required. I've efficiently forgotten things like co-efficients, which I probably knew in high school. Then again, since I got a D in high school algebra, I probably never did understand coefficients.

You're using them, you just don't like the jargon. :-) In post #45, you spelled out your method and how you give weights to things like OPS+, SB, position, grey ink, black ink, etc. Another name for those weights could be "coefficients". A*(OPS+) + B*SB + C*PositionFactor + D*GreyInk + E*BlackInk. Your method might be more complicated than that, but A,B,C,D,E are "coefficients".

Just call them "weights" or "factors" if you want to keep claiming you don't like math. :-)

Except Trammell bowled all 12 frames more often than Whitaker

Someone should've informed Whitaker they added 2 frames.

Someone should have told Whitaker to bowl a strike in the 10th once in a while.

Call it "10th extended" if you like...

Just call them "weights" or "factors" if you want to keep claiming you don't like math. :-)

I like math, as long as it doesn't extend past the "creative accounting" I use when balancing my checkbook.

I have Trammell with 86 and Whitaker with 74. Are you using old numbers, DL from MN? The most recent ones (posted in September, I think) are in the HoM Yahoo group.

And yes, Trammell is quite peaky while Whitaker is freakishly steady. The standard deviation of WARP2 of Trammell's seasons is 2.84, for Whitaker it's 1.18.

I threw out the cruddy seasons. I don't fully trust your pitcher WAR yet so I'm not going to concur with that Blyleven placement.

Also people should recognize the difference between the $$ values and the WAR numbers.

Preliminarily,

Bert Blyleven
Lou Whitaker
Mark McGwire
Tim Raines
Alan Trammell
Dwight Evans
Dave Stieb
Will Clark
Andre Dawson
Keith Hernandez
Bret Saberhagen
Willie Randolph

#2-#5 are close. I like Whiteaker an awful lot, and already knew that I wasn't quite so high on Raines as most others here. Maybe I undervalue his baserunning. I've been working on a PHOM, and only Hernandez, Saberhagen, and Randolph might not make it, but they'd be among the next one or two players at their positions, even if they don't. In other words, this is a good group of selections, in my opinion.

DanR #56,

I have the same separation between the top 3 and the next group. I have them in a slightly different order, but that might be my Canadian bias creeping in..

Do you dock Blyleven even slightly for pitching much better when the game was out of reach?

DL from MN, you shouldn't fully trust my pitcher WAR yet--there's a reason I call it preliminary. It's totally reliant on BP's DERA and makes no adjustment for career length. The pitcher *hitting* numbers are definitely right, though! :)

Sure, the $$ values reward peak seasons and ignore sub-replacement ones, the WARP are straight linear.

Kenn, take a look at my piece on Raines' baserunning: http://www.nytimes.com/2008/01/06/sports/baseball/06score.html. He was over 100 runs with his legs for his career, a remarkable feat. And I don't know if anyone would agree with you on Whitaker > Trammell.

Mike Green, I definitely do not do any WPA-based analysis. My understanding is that the guy who really did underperform his ERA+ in terms of win probability was Stieb, not Blyleven.

And I don't know if anyone would agree with you on Whitaker > Trammell.

Looking at the thread it looks like more than a handful of people agree with him.

68. Daryn Posted: January 17, 2008 at 11:09 PM (#2670943)
And I don't know if anyone would agree with you on Whitaker > Trammell.

Looking at the thread it looks like more than a handful of people agree with him.

I'm with DanR on this one, it's a surprise. But I haven't put in the evaluation time that others have.

I was going to say the same thing as Daryn. Looking at OCF's chart in 18, I would initially suspect a peak versus career split, but I can't recall whether that's consistent among those of us with Whitaker above Trammell.

I did see DanR's article on baserunning, which is precisely why I think my different evaluation of Raines comes from that subject. I'm only giving him about half as much credit, but based on much cruder methods. Giving him 100 runs would move him up to 2nd in my rankings.

Prelim ranking largely based upon a career value perspective:

McGwire
Blyleven
Whitaker
Raines
Trammell
DwEvans
WClark
Randolph
Dawson
Hernandez
Saberhagen
Stieb

Kenn: remember that the chart in #18 is offense only. What you see in it is not enough for me to move the second baseman ahead of the shortstop. I'll have Trammell rated higher.

There's a second factor: that's based on RCAA and it doesn't much penalize a player for missing a few games. There were portions of his career during which Whitaker was largely platooned. I think the measure I was using ranks him higher than if he had faced all pitchers - at least that may be true of the 14-14-13-13-12 part of the chart.

Thoughts about the top spot:

As a pitcher, I have Blyleven in the same general neighborhood as Perry, Carlton, and Niekro. How large is the gap between Perry/Carlton/Niekro and Seaver/Johnson/Alexander?

As a corner outfielder, I have Raines in the same general neighborhood as Waner, Gwynn, Clarke, and B. Williams. How large is the gap between Waner/Gwynn/Clarke/B.Williams and Ruth/T.Williams/Musial/Aaron?

As a shortstop, I have Trammell a little behind Ripken/Yount/Appling/Cronin and a little ahead of Reese/Boudreau. The gap to Wagner is enormous, but who else is there?

As a first baseman, I have McGwire in the same neighborhood as Allen and a war-credit version of Greenberg; Killebrew and Murray are lurking but the peak/career stuff gets pretty wild. How big is the gap between Allen/Greenberg/Killebrew/Murray and Gehrig/Foxx/Mize?

As a second baseman, I have Whitaker in the same general neighborhood as Sandberg, Frisch, and Grich. How big is the gap between Sandberg/Frisch/Grich and Hornsby/Collins/Morgan? But then I think second basemen should also be compared to shortstops.

I still don't know the answer, but I think I'm talking myself into putting Blyleven in the top spot.

OCF--or John Murphy or whomever may know--what HTML tags do you use to create the charts above? Does the code button above the comment box actually work?

I use (left square bracket)pre(right square bracket) with /pre at the end. I write everything up first in a text editor set to plain text and a monospaced font. Don't use tabs. Every line must start with a character, not a blank, and (unless the rules have changed again), you can't have more than two blanks in a row. This last requirement accounts for some of the choices I made, including the periods after some shorter names and all those vertical lines in the pitchers chart above. It looks wrong in preview. You just have to post it anyway and hope you got it right.

John may be doing things a little differently.

OCF, not all positions have to be equally distributed...would anyone put a catcher in their top 20 players overall?

I definitely have some disagreements with your groupings. I see Seaver as well behind Johnson/Alexander, Yount/Appling/Cronin as well behind Ripken, and Greenberg with war credit much closer to Mize than the other 1B you list (Killebrew is barely in my PHoM).

Who else is there?

DanR,

Both Stieb and Blyleven pitched much, much better when the margin was more than 4 runs than otherwise. There are relatively few pitchers who have had long careers and who have this pattern. It is reasonable to dock each of them something for it.

The standard pattern is on-base percentage goes down and slugging percentage goes up as the game gets out of hand.

I am not advocating a WPA analysis generally, but when a starting pitcher over a long career displays the Stieb/Blyleven pattern (which leads to "inefficiency") or the contrary Koufax pattern (which leads to efficiency), there is merit to an adjustment.

> would anyone put a catcher in their top 20 players overall?

Josh Gibson

Yes, there's Vaughan. He's not really in a group with anyone. And I wasn't really claiming that Seaver was the equal of Johnson or Alexander, but he's quite easily the class of the (retired) modern pitchers.

Who else is there (part 2)?

Sorry, I meant among 20th century MLB players. I should have clarified. Gibson is easily top 10 with a reasonable catcher bonus.

Mike Green, I think part of being fair to all candidates is treating them all the same. If I'm going to open the WPA can of worms for Blyleven, I need to view every P through the same lens. I am sure there are more than just 2-3 outliers of that magnitude in all of baseball history. Needless to say, I don't have the data to do this (nor do I believe it is publicly available). Moreover, I'd need to see a study showing that this type of "clutch pitching" really exists as an ability--in other words, that the standard deviation of "pitching to the score" over all of baseball history is greater than what would be expected by random chance (as the stdev of "hitting to the score" famously is not).

OCF, not all positions have to be equally distributed...

Understood, and it's in part a reason not to give Whitaker top billing or to place him above Trammell. The part of this thought that I'm least sure what to do with is the oversupply of RF/LF.

McGwire doesn't come anywhere near Trammell or Whitaker in either measure. Trammell is also consistently ahead of Whitaker (which is the HoF consensus also). I'll leave the peak/prime analysis to someone else.

Well, considering how I've advertised myself as a peak voter all these years, I'll give that a try. . . .


The first column for each uber-stat is the player's best 5 non-consecutive years, while the second column is what I call the jschmeagol method--recording a player's seasonal total after reaching a certain level, say, 21 WS or 7 WARP3 or 4.5 WARP2. (I've always liked jschmeagol's method--it takes a player's prime, his good years outside his Top 5, into better account.)

McGwire and Whitaker flip-flip; McGwire is among the best in career value, while Whitaker hovers among the bottom of the group. However, Trammell more than holds his own among the peak candidates--except in Win Shares.

DanR:And I don't know if anyone would agree with you on Whitaker > Trammell.

Daryn:Looking at the thread it looks like more than a handful of people agree with him.

Paul Wendt:I'm with DanR on this one, it's a surprise. But I haven't put in the evaluation time that others have.

Kenn:I was going to say the same thing as Daryn. Looking at OCF's chart in 18, I would initially suspect a peak versus career split, but I can't recall whether that's consistent among those of us with Whitaker above Trammell.

Like Dan and Paul, I'm just as surprised that people are place Whitaker over Trammell. But is it just a career vs. peak thing?

Talk about 3 different perspectives on their career value!

Whitaker ends up with more raw career value than Trammell, packing more PAs into his 19 years than Trammell did into his 20. But because of his bigger seasons--and the fact that he packed more time into them--Trammell easily beats Whitaker in peak value.

Why then does BP's numbers see their career value as equal, and why do DanR's numbers see Trammell ahead of Whitaker career-wise? At first I thought it was defense, but looking at BP's numbers Trammell sees a bigger jump in his BR & FR than Whitaker after adjusting for all-time.Why? I dunno. However, Win Shares is notably tough on the defensive side of the spectrum--look at the WS peak chart: the middle infielders easily rank among the bottom of the pack. I still do think defense is a part of the reason.

Regardless, while I understand a career voter's train of thought, I do think that some account for peak should be made, and that the gap between Whitaker's career value and Trammell's isn't big and is offset by the latter's peak advantage, if not in Trammell's overall favor.

Thanks OCF!

Tiboreau, I think you must be using version 1.0 of my WARP. I posted updated numbers to the Yahoo group around September.

Here are charts for the 80's Tigers' keystone combo. I'm using the stdev-adjusted numbers, but it barely matters since a) Trammell and Whitaker were exact contemporaries and b) the adjustment for the 80's AL is always right about 1. TXBR is career total excluding sub-replacement seasons.

Trammell



Whitaker



So, as you can see, on raw value above average, Whitaker comes out sizably ahead, 36.4 to 28.7. Whitaker also played nearly a season's worth of games more than Trammell. The rub, of course, is that Trammell was a shortstop and Whitaker was a 2B, and the second half of the 1980's represented the biggest gap in replacement level between those two positions in MLB history. Here are my replacement levels (measured in standard deviation-adjusted wins below average per full season played) for 2B vs. SS. Over time, they have usually been about 1.0 wins apart. Not so in Trammell and Whitaker's day:



As you can see, the gap between 2B and SS went from 0.5 wins in 1960 (its all-time low) all the way up to 2.0 wins in 1986 (its all-time high). This is not a statistical blip or random noise--these are averages taken of 88 player-seasons (the worst 3/8 of starters on 26 MLB teams in rolling nine-year intervals). You will see it if you look at positional averages as well (although it will be more clear if you just look at the NL, since the Holy Trinity of Ripken, Yount, and Trammell will skew the AL average). By the late 1980's, the positional OPS+ for 2B was actually above 100, meaning it was technically an offensive position, while SS was just as bad as ever (and with the highest within-position standard deviation ever, since you had your Jackie Gutiérrezes, Alfredo Griffins, and Onix Concepcións playing in the same league as the Big Three, which is why I have all three of them so high in my all-time rankings). And this is not a case of a few stars dragging up the 2B average--it's that decent-hitting 2B were freely available on the scrap heap (as they remain today). (If anyone wants evidence/examples, I've got plenty).

So WS prefers Whitaker and BP has them as comparable because both of them just sort of glop 2B and SS together as "middle infielders," so to speak (the same reason, in my view, that we unconscionably elected Nellie Fox). In fact, 2B was a mid-spectrum position, similar to 3B and CF, while SS was still the toughest position to fill by far. That's just an insurmountable advantage in Trammell's favor, IMO.

I think 1985, 1987, 1989, and 1994 might have something to do with it, kevin.

That's because Saberhagen's peak years were far, far, far greater than Rice's. Just using my own WARP2, I have Saberhagen's top four seasons at 8.6, 7.3, 7.0, and 6.9. He should have been the AL MVP in 1989, deserved his Cy Young in 1985, was the second-best pitcher in the NL in 1994, and was the third-best pitcher in the AL in 1987. If that's not a Hall of Merit peak, I don't know what is. Rice's top 4, by contrast, are 7.3, 5.9, 5.5, and 4.8, and that 7.3 is in part due to his compiling a ridiculous 746 plate appearances in 1978 due to playing on a high-OBP team in a huge hitter's park. He did deserve his '78 MVP, but in overall historical context it wasn't nearly as impressive as Saberhagen's 1989, and after that his next-best showings are an 8th place in 1979 and a 9th place in 1986. Lots of voters are amenable to voting in a guy with just a few great years--look at Hughie Jennings or Ralph Kiner. But a guy with one MVP season and then just three years as a mid-level All-Star? There's a reason no one suported hi for the HoM.

kevin wrote:

So, pulling out a few great years and focusing on them works for Saberhagen but not for Rice.

I see.

If Jim Rice had four seasons that were as good as the ones Dan R has mentioned for Saberhagen, the HoM electorate would undoubtedly have given him some support. But he doesn't have those years, not when one adjusts for park, GIDP, and defense (Rice was a below average defender at a low-value defensive position).

To take a convenient measure, WARP3, here are Rice's and Saberhagen's top 4 seasons:

Sabes -- 12.4, 10.3, 10.0, 8.7
Rice -- 9.8, 9.8, 8.5, 7.4

To a voter who favors peak, Saberhagen's record is simply substantially better than Rice's: it's not that they are being held to different standards. In career value, Saberhagen is also well ahead of Rice: 93.6 WARP3 to 83.2. 10 wins is a big difference!

Your intimation that the HoM electorate is in some way prejudiced against Jim Rice or shallowly inconsistent in its determinations shows that you don't have much sense of how we have worked. Most of the electorate employ purely statistical methods of evaluation: we develop system to determine value, and we evaluate each player's record using that system. If Saberhagen's record shows that he had outstanding value, despite the fact that he missed a lot of time with injuries, he gets support. (You might note that the four seasons Dan R. mentioned are the four seasons in which Saberhagen was _not_ hurt, and he was brilliant in all of them.) If Rice's record shows that he did not have outstanding value, regardless of his gonzo counting stats, he doesn't get support.

Because his top 4 years are the answer to why Saberhagen was elected, see #89. You're the person who brought up comparison of their top few years.

I also don't think it matters what cutoff you have, saberhagen still beats Rice comfortably.

Would you care to share what you believe to be the flaws in said formulas, or what "parameters" you think we are ignoring?

I picked four because you asked why Saberhagen got in, and the answer is obviously those four years. But no, adding or subtracting years does not change the calculus. Here are my WARP2 numbers for the pair, ordered from best to worst:



I'm afraid there isn't a single set of endpoints that come close to favoring Rice. If you're interested in the ethodology behind these numbers, check out the thread on my WARP data.

Let's just remember what this discussion thread is about. We have precisely 12 players to find a relative rank among, and we must rank all 12. Jim Rice is not one of the 12 names. So let's not let kevin sidetrack the conversation.

I'm not much of a fan of Win Shares or WARP3, but I still don't vote for Rice and I did vote for Saberhagen.

It's harder to have 4-5 great SP seasons than to have 4-5 great OF seasons. There are plenty of non-HOM OFs with better, equal, or slightly worse cases than Rice. We've elected most/all of the SPs with better or equal cases than Saberhagen, and there aren't as many slightly worse players in his case.


I have no idea why someone would assume that because a certain player doesn't get HOM votes, there must be some sort of agenda. It's just silly.

Neither Rice nor Saberhagen held up as long as we'd prefer in an ideal candidate. Both were great at times, but Saberhagen was greater at his peak than Rice was in his, in a far more important role at a position where it is far more difficult to repeat success on a longterm basis.

Point taken, OCF. I wish I had seen it sooner. I'm done on that subject.

I haven't voted for either Saberhagen or Rice, and I actually have them rated fairly close. But taken in the context of '80s-'90s pitching, Saberhagen is for real. I had him the the 20s on (off) my ballot, very similarly to Stieb and Hershiser, none of whom I would have thought to rate that highly before digging into it a bit.

As to the "problem" with Jim Rice being that the HoM participants use "statistical models".... Well, like I say, I had Rice in the general vicinity on my ballot of Saberhagen, meaning substantially higher than most voters. I also have Dave Parker in the same vicity, substantially higher than most voters. I remember both of them as hitters who were "feared." Ah, there's a parameter that works for Rice. So, I was shocked to find that he (Rice) ranks about 30th in IBB for the decade he played, even with selective endpoints that should work to his advantage.

There have been lengthy (understating it by a lot) discussions of Rice in the newsblog. It's all been said. Still, I'd be curious to know what "parameters" of Rice's career we're missing by using "statistics" to evaluate him.

Here's a revised prelim. The most significant change from my first ballot is the placement of Trammell and Whitaker. I discovered that I had never integrated the conclusions of Dan R.'s WAR into my evaluation of them, which kept them lower than they should have been.

The new list, with some comments:

1) Bert Blyleven -- 370. I think this number overrates him a bit: he wasn't as good as Carlton and Perry. However, using my win-based system alone (not integrating it with WARP, which is derived from detail stats, not results), he still ranks slightly above Raines, even though he wasn't as good as Carlton. His peak is not as good as Raines' or Trammell's, but he was well above average for over 15 years and tremendously durable. There's a lot of value there.
2) Tim Raines -- 342.
3) Alan Trammell -- 330.5. Dan R's WAR loves Trammell. If I relied only on his system, Trammell would be handily in the top spot. However, I do not entirely accept contextual value above replacement as defining merit: in my view, even with a broadly based replacement baseline, there are still variations caused not by real talent demands on defense but by widely held errors of judgment about ability. The 1970-1990 shortstop trough I treat as one of those errors. Still, Dan R's system captures the value of infielders much better than WARP or WS, so I weight its conclusions more heavily, and that raises Trammell to the third spot.
4) Mark McGwire -- 303. Not sorry to see him drop a little.
5) Lou Whitaker – 301.5. Rises with the inclusion of Dan R's WAR in his assessment, though not as much as Trammell.
6) Dwight Evans -- 290.
7) Will Clark -- 274
8) Willie Randolph -- 264. Bringing him back up to where my system puts him.
9) Keith Hernandez -- 259
10) Bret Saberhagen -- 229
11) Dave Stieb -- 228
12) Andre Dawson -- 249

12) Andre Dawson -- 249

Is the 249 a typo or did you do a subjective re-ordering after you calculated the number?

Is the 249 a typo or did you do a subjective re-ordering after you calculated the number?

249 is not a typo, but the reordering is only partly subjective. Since I include Dan R's numbers for position players but not for pitchers, their totals in the system, while similar, are not exactly commensurate. I have found, over time, that the in-out line for pitchers falls around 220, while for position players it falls around 240, so when I rank pitchers and batters together, I usually add 15-20 to the score of the pitcher, especially near the borderline. That makes Saberhagen, Stieb, and Dawson pretty close to equal, and I order them subjectively from there. Since I think replacement level is too low in WARP and WS, I tend, in subjective adjustments, to favor peak over career, as I have done in this case.

I should have explained that above: I think I noted it on my first preliminary ballot, but since I didn't change the rankings here (except for moving Randolph up), I forgot that I needed to comment on my apparent rejection of my system's conclusions.

Tiboreau, I think you must be using version 1.0 of my WARP. I posted updated numbers to the Yahoo group around September.

Yeah, those numbers are probably from before then. I must have missed the memo on the availability of v2.0. By the by, does that include preliminary pitching info?

Of course, I should've realized that the reason your WARP2 had Trammell over Whitaker, career-wise, was due to standard deviation. I wonder if that is the reason for Trammell receives a bigger boost than Whitaker in BP's transition from WARP2 to WARP3. . . .