I guess we're just never going to see a player elected unanimously, are we? Not even Jeter.

Nope, we won't. Even if the pinheads who refuse to vote for a first ballot candidate get swept up in Jeter's magic eyes, there will be a voter from the Detroit area who says, "Trammell was just as good as Jeter and he never got elected. So if Trammell wasn't good enough, then screw Jeter, he's not good enough, either."

Tim Raines
Alan Trammell

But doesn't he get points for not listing Smith and Mattingly?

Was there a blank ballot this year? Or any with just one name, or others with 2?

I guess we're just never going to see a player elected unanimously, are we? Not even Jeter.

Probably not, but I don't think it matters. Some here seem to be upset when players are elected with fewer votes than they "deserve", or even justifiably not elected with fewer votes than they arguably deserve, but it doesn't really matter. There are pluses & minuses from having a large voting pool, and IMHO the advantages far outweigh the disadvantages, including the likelihood that no one, no matter how deserving, ever receives 100%.

I don't think you have to search too hard to find out why Frank came in above Edgar. Frank had an insane peak. Edgar had a really, really good peak. Edgar remained good older but has a harder time compiling a career case (fairly or unfairly) because of his late start. I can see a case for him, but comparing him to Frank Thomas weakens his case significantly since it exposes what Edgar was not.

Yeah, Thomas vs. Edgar was one of the places where I really had to think seriously about the way that WAR deals with DHs. Because Thomas is SUCH a better hitter than Edgar that it's hard to square the closeness of their WAR numbers. Only a few wins apart, despite Thomas having 160 more Rbat. The fact that Edgar was actually a serviceable third basemen for a quarter of his career is part of the reason for the gap, but a bigger part is that Thomas was regularly scoring worse than if he had simply been the DH. If you zero out Thomas' years and simply treat him as if he were a full-time DH, it adds somewhere around 5 or 6 wins to his total, which seems like a reasonably fair assessment of his true quality.

That can't be right. Frank had an Rfield of -65. So he's only going to look 6 wins better if you assume he's an average first baseman.

If he were a fulltime DH from day one, he'd get a -15 position adjustment instead of -10 for a 1B. With 971 games at first, look like 6 full seasons worth. An extra 5 runs each year means -30 runs to the position penalty. So he'd look 3.5 wins better.

That can't be right. Frank had an Rfield of -65. So he's only going to look 6 wins better if you assume he's an average first baseman.

If he were a fulltime DH from day one, he'd get a -15 position adjustment instead of -10 for a 1B. With 971 games at first, look like 6 full seasons worth. An extra 5 runs each year means -30 runs to the position penalty. So he'd look 3.5 wins better.

Hmm, the back of your envelope is probably more precise than mine. I was thinking of the DH penalty being -13 or -14 (what he scored in 2006 and 2007 when he was a full-time DH). But obviously those weren't complete seasons. I think that's where I got the extra win and a half.

I guess we're just never going to see a player elected unanimously, are we? Not even Jeter.

there is really no plausible argument against him.

Allow me: Jeter's achievements rest primarily on his longevity (3000 hits!), his fame ("The Captain"), and the fact he played shortstop. Now, I don't care hugely about longevity or the big numbers, and I don't care about his fame. Playing shortstop, however, is valuable, and his bat as a shortstop is hall-of-fame quality (115 OPS+), but wouldn't be HOF quality at most positions. However, if you give Jeter credit for having played shortstop, you have to examine his defensive contribution, and that defensive contribution, as far as we can tell, was the worst of all time.

Some might stop there: good bat plus long career plus worst-of-all-time-defensive-shortstop might not equal hall of famer for many. But for myself, the question is: was his glove so bad that he falls short of the standards for induction to the Hall of Fame? And the answer is yes, his defense was that bad.

The best career-length estimates of defensive value are probably Michael Humphrey's DRA (available at seamheads) and Tango's WOWY. Other measures (TZ, UZR, DRS) have regressions built into their systems, and so overrate the worst fielders. DRA has Jeter at a hellish -350 odd runs for his career; WOWY is similar. Sub those fielding numbers into the WAR calculations for fWAR and rWAr, and Jeter's career value takes a severe hit: down to about 62 rWAR and 54 fWAR. That falls significantly short of the JAWS standards for SS. It's just not worthy of Cooperstown.

The fact that Edgar was actually a serviceable third basemen for a quarter of his career is part of the reason for the gap, but a bigger part is that Thomas was regularly scoring worse than if he had simply been the DH. If you zero out Thomas' years and simply treat him as if he were a full-time DH, it adds somewhere around 5 or 6 wins to his total, which seems like a reasonably fair assessment of his true quality.

Except (as I've pointed out here, ad nauseum, often to no avail) Hurt hit much better as a 1B than he did as a DH-even if you adjust for him DHing more during his decline phase.

#1508
Given your thoughtful analysis and summation, do you think for one minute that the collective body that is America's baseball writers will pass on this opportunity to spread their plumage and strut around the media grounds of America and not elect this guy overwhelmingly on the first ballot?

No way they will pass up this gigantic circle-jerk opportunity for their chance at fawning obsequiousness to America's hero.

I like Jeter, the player, a lot; I deplore idol worship by sycophants and that is what we will mostly get from MLB network commentators and the writer community.

Jeter will go in on the first ballot, unfortunately, despite being a borderline guy by my calculations. And yeah, a lot of people will take the opportunity to write obsequious things about him. But it'll be five years from now; the voting body will have changed significantly, and there will almost certainly be several loud voices pointing out things like what we see in 1508.

@jingoist

There's no doubt he'll be voted in (and he certainly won't be the worst choice in the world). I'm hoping he's limited to something like 93% (even if we hear endless claims about how unjust and disrespectful such a vote would be).

It's pretty inconceivable that 25% of the BBWAA would take such an unpopular stand in order to vindicate defensive value however nice that would be.

Except (as I've pointed out here, ad nauseum, often to no avail) Hurt hit much better as a 1B than he did as a DH-even if you adjust for him DHing more during his decline phase.

Is there good evidence that the 'DH penalty' effect really exists for hitting?

Generally, guys who split time between DH and 1B will take the DH spot when they are less healthy, right? Which would explain worse numbers in those games. And as with guys like Thomas, many players will be the DH far more regularly as they age (and presumably decline). Are there studies that have adjusted for this? My impression was that most attempts had found some tentative evidence but weren't able to conclusively state that DHing really does hurt offensive stats.

If those more conclusive studies aren't out there, I don't see any particularly good reason to punish Thomas for the random distribution of his stats at 1B and DH. It seems similar to folks wanting to downgrade Walker's stats even more than park effects indicate simply because his home/road splits are particularly large.

Is there good evidence that the 'DH penalty' effect really exists for hitting?

Even if there was, what difference would it make? War and other stats are about value evaluation, and regardless of why they are hitting poorly, the fact is that they aren't bringing anything else to the table than their bat, so the war penalty a player gets as a dh HAS to be greater than any other position.

Even if there was, what difference would it make? War and other stats are about value evaluation, and regardless of why they are hitting poorly, the fact is that they aren't bringing anything else to the table than their bat, so the war penalty a player gets as a dh HAS to be greater than any other position.

Surely you don't meant that the DH penalty by definition must be set at the level of the single worst defender ever to play baseball? If not, guys are occasionally going to 'lose' value by playing the field, and that is genuine lost value from the perspective of the team. But it doesn't seem particularly fair to punish the player for his team's poor decision to let him try to field.

Hence, zeroing out negative defensive contributions for the project of assessing an individual player's value.

For other arguments I'm fine with assessing different values or zeroing out etc, but as far as war/value stats is concerned, the positional penalty for a dh has to be bigger than the penalty for a first baseman. War and other stuff though is a value stat, it doesn't matter why the player is at the position, just that he is and the value they bring to the table. If a team decides to take a great defensive shortstop and put him at first base, he still gets the first base positional adjustment, just because it was the teams decision to weaken his defensive value, doesn't change the fact that he is no longer providing the same defensive value as he was before.

1505 why would you want to zero out Thomas' defensive numbers when he's playing the field? If he really was a horrifically bad fielder, having him out there cost his team runs (and hence wins). If what we're doing is measuring the value that he produced, surely the fact that he was a butcher in the field who cost his team runs should be a part of the calculation. It's not "fair", in that Thomas didn't write the line-up card, but whether or not he cost his team runs doesn't depend on whether or not it's fair to blame him for it. The fact is that he DID cost his team those runs, and that's what we're measuring.

1505 why would you want to zero out Thomas' defensive numbers when he's playing the field? If he really was a horrifically bad fielder, having him out there cost his team runs (and hence wins). If what we're doing is measuring the value that he produced, surely the fact that he was a butcher in the field who cost his team runs should be a part of the calculation. It's not "fair", in that Thomas didn't write the line-up card, but whether or not he cost his team runs doesn't depend on whether or not it's fair to blame him for it. The fact is that he DID cost his team those runs, and that's what we're measuring.

Because often times the argument is how good he actually was, not how valuable he actually was. If a guy is unnaturally hurt by his park, beyond even value stats can account for, simply because of his particular skill set, there is nothing wrong with looking at how good he actually was, above and beyond his actual in season value. Same with a guy who is playing out of position, if he should have been a DH, but his team didn't put him there for whatever need, he shouldn't be overly penalized by his inability to play. At the worse his defensive value should be adjusted to that of a dh(numbers out of my ass....say a dh penalty is -12 runs over 150 games, and an average 1b penalty is -9, yet because the player was such a bad defender his value is -14....on a war scale, that is perfectly fine, but on a hof argument, it's best to limit the negative to the point where he should have been at, which is -12)

Frank Thomas OPS splits:
1B: 1.087
DH: 0.899

Thomas was a fine DH but was nowhere near the monster he was when he played the field. Yes, this is a biased sample since he played more DH later in his career, but even doing the comparisons within seasons, you will find he hit much better when playing the field. And as a whole, players hit worse when they don't play the field, so there's reason to believe this is not a fluke. Giving the hypothetical "what if Thomas just was a DH?" does not only require adjusting defensive value, but offensive value as well (and in such a way that would not be easy to do).

Thomas was a fine DH but was nowhere near the monster he was when he played the field. Yes, this is a biased sample since he played more DH later in his career, but even doing the comparisons within seasons, you will find he hit much better when playing the field. And as a whole, players hit worse when they don't play the field, so there's reason to believe this is not a fluke. Giving the hypothetical "what if Thomas just was a DH?" does not only require adjusting defensive value, but offensive value as well (and in such a way that would not be easy to do).

Is this meant as a response to my 1513? Because it does not resolve any of my concerns, merely waves a hand at them.

Frankly, that late-career Thomas and somewhat-injured-Thomas and taking-a-day-of-rest-Thomas managed to hit that well as a DH seems like an argument for my position rather than the opposite.

Frank Thomas, in his 20s, was one of the greatest hitters in history. And he was primarily a first baseman in this period. In his 30s, he was merely a very good hitter, and was primarily a DH. If you want to ascribe a meaningful portion of that decline to his shift to DH, you are free to do so. But it sure seems like the more reasonable explanation is simply that he got older.

Beyond that, I would really like to see evidence that the 'DH penalty' is a real and quantifiable thing, rather than merely an anecdotal problem. Does such evidence exist?

Frankly, that late-career Thomas and somewhat-injured-Thomas and taking-a-day-of-rest-Thomas managed to hit that well as a DH seems like an argument for my position rather than the opposite.

During his prime hitting years, 1990-1997 and 2000, Frank had four years in which he had significant (over 100) Pas at both 1B and DH. In three, he hit far batter as a 1b than as a DH:

1991 947 OPS as a 1B in 240 PA, 1034 as a DH in 460
1995 1172 OPS as a 1B in 414 PA, 875 DH in 231
1997 1179 1B in 435, 943 DH in 214
2000 1193 1B in 132, 1023 DH in 573

Even in a down year in 1999 - 975 1B, 836 DH. Over 200 PA in each.

In 2003 he had a 146 OPS+ on the year. 1212 OPS as a 1b in 115 PA, 903 as a DH in 545.

The evidence is pretty compelling.

whatever the facts say the general belief in baseball is that the vast majority of players hit less well in the dh role primarily because the vast majority of players do not LIKE the dh role. the vast majority of players want to be full-time players even if they kind of stink fielding

that is why there are so few full-time dh players. managers are loathe to force a guy into a role he clearly detests and one in which he will perform at a lower level than if he was playing both ways

guys like ortiz and edgar martinez are the exceptions of guys who embraced the role and excelled

I know this isn't the exact correct way of doing this, but it's quicker than compiling all the base stats and gives a good approximation. Taking all the in season splits from 1521 and doing a weighted average shows an OPS of 1123 in 1544 PAs as a 1B, and 947 in 2392 as a DH. That's every season in which Frank had more than 100 PA in both roles. In 1994 he had 59 PAs as a DH and OPSd 400 points lower than as a 1B. In 1998 he had 64 PAs as a 1B and OPSd 100 points higher than as a DH. In 1990 as a rookie, he had 31 PAs as a DH and OPSd 180 points lower. I didn't include those numbers because it could just be noise.

Mussina gave one of my favorite-ever quotes by an athlete, when asked whether he would be able to accept a gay teammate in the locker room: “I'm going to make the assumption that I already have."

And who was that gay teammate? You guessed it... Frank Stallone!

I didn't include those numbers because it could just be noise.

Without knowing specifics, it's very possible that he often DH when he was less than healthy, and that might also figure into some of those numbers.

I don't think Thome, Baines or Molitor had problems with DH.

Without knowing specifics, it's very possible that he often DH when he was less than healthy, and that might also figure into some of those numbers.

Possibly. But when he was mostly a first baseman and partly a DH, the first base numbers were better. When he was mostly a DH and partly a first baseman, the first base numbers were better. When I calculated the average OPS, I used only seasons of 100+ PAs in both to try to wash out any of those effects. He had 6 of those years, (and in 4 he had over 200 in each), and in 5, his OPS was better, usually much better, as a first baseman. And those were in poor (for him) years, good years, and great years.

whatever the facts say the general belief in baseball is that the vast majority of players hit less well in the dh role primarily because the vast majority of players do not LIKE the dh role. the vast majority of players want to be full-time players even if they kind of stink fielding

that is why there are so few full-time dh players. managers are loathe to force a guy into a role he clearly detests and one in which he will perform at a lower level than if he was playing both ways

guys like ortiz and edgar martinez are the exceptions of guys who embraced the role and excelled

Then that is an argument for positional scarcity, right? That it is a completely different kind of scarcity than (say) finding an excellent hitting shortstop doesn't ultimately change that. You also have the fact that creating another regular position, by itself, makes players for that position scarcer.

I don't really buy the 'Thomas was incapable of hitting at his top level as a DH' idea, obviously. But I'm particularly skeptical of attempts to prove it that focus merely on his (very limited) splits over a few seasons.

If this is something that plays out more regularly, and can be measured for many players, then yes I'll start to believe it. But if it's just something with Thomas, the sample size is way too small to draw any meaningful conclusions. It might just be random. If it's not random, it might be because his time DHing usually corresponded with him being banged up. If it's not that, it's possible that playing DH sporadically does cause you to lose focus, in a way that simply being a full-time DH would not cause problems. Who knows? We can speculate, but without significant data behind the theory itself, it just seems like attaching a narrative to Thomas--one that doesn't necessarily beat the more boring theory that his differences are just part of the normal variation.

Anyways, I looked at a few other guys to check seasons where they did a fair amount of DHing and playing in the field. Baines in 89 hit significantly better as a DH. Same with Thome in 99, though he was much better when playing the field in 2000. Ortiz in 2003 hit WAY better as a DH. Edgar was a little bit better in 92 as a DH. Encarnacion was better as a DH in 2011, worse in 2012 and 2014 and the same in 2013.

I don't have the dataset or the wherewithal to really test this. But I would certainly be interested in seeing the results of a more wide-ranging look at the question.

"Surely you don't meant that the DH penalty by definition must be set at the level of the single worst defender ever to play baseball? If not, guys are occasionally going to 'lose' value by playing the field, and that is genuine lost value from the perspective of the team. But it doesn't seem particularly fair to punish the player for his team's poor decision to let him try to field.

Hence, zeroing out negative defensive contributions for the project of assessing an individual player's value."

We're talking about a couple different things here. CFB is talking about ability. 1515 and 1505 are not, they're talking about value. It doesn't make sense to zero out below-replacement fielding performances when talking about value, because the player who put up those performances really did cost his team runs with his glove. When talking about ability, that's something else entirely. But not something that 1505 was talking about (since it's about WAR) nor 1515 (quoted above).

And responding specifically to 1515, if you're measuring value, fairness has got nothing to do with it. It's just a measurement, and the quantity being measured (runs produced or saved) doesn't go up or down depending on whether or not the player was assigned a position he's ill-suited for.

Anyways, I looked at a few other guys to check seasons where they did a fair amount of DHing and playing in the field. Baines in 89 hit significantly better as a DH. Same with Thome in 99, though he was much better when playing the field in 2000. Ortiz in 2003 hit WAY better as a DH. Edgar was a little bit better in 92 as a DH. Encarnacion was better as a DH in 2011, worse in 2012 and 2014 and the same in 2013.

I don't have the dataset or the wherewithal to really test this. But I would certainly be interested in seeing the results of a more wide-ranging look at the question.

Try BB REF Play Index Batting Split Finder:

For single seasons, From 1973 to 2014, as DH (within Defensive Positions), (requiring PA>=501 for entire season(s)/career and PA as DH>=200), sorted by greatest tOPS+ for this split

                                           
Rk              Player Year tOPS+  PA PAtot
1           Don Baylor 1979   140 293   722
2           Carlos May 1973   135 312   617
3         Eddie Murray 1988   128 241   681
4          David Ortiz 2003   124 314   509
5           Don Baylor 1977   123 263   645
6          Ted Simmons 1983   122 282   650
7         Luis Polonia 1992   120 205   635
8       Gary Alexander 1978   119 280   564
9         Julio Franco 1997   118 302   505
10         Ryan Doumit 2012   118 200   528
11       Randall Simon 2002   118 265   506
12   Edwin Encarnacion 2011   117 294   530
13          Ron Fairly 1977   117 245   528
14     Rafael Palmeiro 2001   115 200   714
15       Darrell Evans 1988   115 246   522
16           Mo Vaughn 1999   114 289   592
17        Ruben Sierra 1996   113 351   587
18     Victor Martinez 2011   112 466   595
19          Luke Scott 2010   111 382   517
20         Mike Easler 1984   110 544   666
21        Chris Carter 2014   110 487   572
22         Jason Kubel 2008   110 327   517

[...]

272         Greg Vaughn 2000    82 220   545
273          Joe Carter 1997    82 276   668
274            Jim Rice 1976    82 217   624
275          Tim Salmon 2003    82 291   621
Rk               Player Year tOPS+  PA PAtot
276        Jose Canseco 1998    82 348   658
277     Rafael Palmeiro 2000    80 203   678
278         Larry Hisle 1978    80 221   600
279          Tom Grieve 1976    80 396   600
280        Frank Thomas 1997    79 214   649
281         Greg Vaughn 2001    78 325   562
282        Dave Nilsson 1997    78 240   629
283        Jason Giambi 2003    77 305   690
284         Ryan Doumit 2013    76 209   538
285       Larry Parrish 1984    76 263   664
286            Jim Rice 1978    75 225   746
287        Jason Giambi 2002    73 277   689
288        Johnny Damon 2007    73 228   605
289      Reggie Jackson 1985    72 206   541
290           Adam Dunn 2013    71 301   607
291       Jesus Montero 2012    69 321   553
292    Carl Yastrzemski 1979    68 227   590
293        Frank Thomas 1995    67 231   647
294           Jim Thome 2000    66 217   684
295           John Jaha 1996    62 273   636
296        Jason Giambi 2005    60 235   545

Provided by Baseball-Reference.com: View Play Index Tool Used
Generated 1/12/2015.

Anyways, I looked at a few other guys to check seasons where they did a fair amount of DHing and playing in the field. Baines in 89 hit significantly better as a DH. Same with Thome in 99, though he was much better when playing the field in 2000. Ortiz in 2003 hit WAY better as a DH. Edgar was a little bit better in 92 as a DH. Encarnacion was better as a DH in 2011, worse in 2012 and 2014 and the same in 2013.

What does any of that have to do with Frank Thomas? I understand not wanting to generalize from the specific example of Thomas to all DHs, and I'm not trying to do that. But in the specific example of Thomas, it's pretty compelling.

You claim the use of "very limited splits". Was Thomas banged up for 231 PAs in 1995 when he had an OPS 300 points higher as a 1B? Was he banged up for 214 PAs in 1997 when his 1b OPS was 236 points higher? Was he banged up for 573 PAs in 2000 when his OPS was 170 points higher as a 1B? Boy, for a first ballot HOFer, he was sure banged up a lot in his prime.

If Frank Thomas had been a DH every game of his career, getting -1.5 wins per 162, his dWAR would have been -21.5 instead of -23.4. In the history of baseball only 13 players rate worse defensively than they would have by taking the full DH position adjustment and never wearing a glove. Only 4 of them had more than a 5 win difference:

-5.4 Prince Fielder
-5.9 Danny Tartabull
-6.6 Frank Howard (played pre-DH)
-11.0 Adam Dunn

Manny Ramirez, for all his awful ratings, was only 1 win worse than a DH. Dr. Strangeglove comes in at 2.6 runs worse.

Out of curiosity I wondered how bad Jeter would have had to be in the field that he'd have the same value as a DH. Answer is 157 runs worse. That would mean a about -400 runs in the field. If he were really that bad he'd be a 55 WAR player.

So if someone like #1508 wants to argue that Jeter is not a deserving HOFer, what they are basically telling you is that Jeter wasn't just a bad shortstop, He was close to a DH in defensive value. Not as bad as Dunn or Thomas, but comparable defensive value to someone like Miguel Cabrera, Jose Canseco, or Bobby Bonilla.

What does any of that have to do with Frank Thomas? I understand not wanting to generalize from the specific example of Thomas to all DHs, and I'm not trying to do that. But in the specific example of Thomas, it's pretty compelling.

If one player has extreme home/road splits, does it tell us that there is something unique and special about that player that is different from every other player in baseball history? Or does it tell us that there are a lot of players and by simple random variation, some of them are going to have more extreme splits than others? If one player has extreme day/night splits, does it tell us something unique about that player? If one player has a big spike in performance on Tuesdays? If he has a big drop in performance in June?

If you encounter an outlier, the first impulse is always going to be to try and explain the outlier with narrative. And I'm not saying that's always wrong. I'm just trying to question the easy impulse to say 'look! splits! explain them!' Because there's every chance that the splits are just random and don't tell us anything about Thomas' unique abilities.

If there isn't any 'DH penalty' in general, that strengthens the case for random variation. If there is in fact a significant DH penalty, it makes Thomas' personal splits look far more like a normal problem. And if it's a normal problem, then of course you need to factor it into an assessment of the value that a guy provides as 1B vs. as DH. But I'm not particularly interested in looking exclusively at Thomas' numbers and presuming that this tells us all we need to know about his ability.

You claim the use of "very limited splits". Was Thomas banged up for 231 PAs in 1995 when he had an OPS 300 points higher as a 1B? Was he banged up for 214 PAs in 1997 when his 1b OPS was 236 points higher? Was he banged up for 573 PAs in 2000 when his OPS was 170 points higher as a 1B? Boy, for a first ballot HOFer, he was sure banged up a lot in his prime.

If you're going to insist on being a jerk, I don't suppose we need to continue this conversation. Cheers.