I teased this on the last thread, and Mike Green made the mistake of biting, so, I can't resist laying out the case against Derek Jeter as a hall-of-famer.

Firstly, let me say that the case against him is purely based on his regular-season value. If you think that the round numbers are sufficient to get a player into Cooperstown (3,000 hits, career .300 average) or that his postseason exploits mean he can't be denied, that's fine - I don't have any quarrel with people who look at hall-of-fame cases through those lenses. My own approach, however, is to ask whether a player provides enough value to get him into the realm where the round numbers and post-season and intangibles can put him over the top. And it's on his overall value that Jeter falls short even of the .

By now, almost everyone knows that Derek Jeter was a poor fielder as a shortstop, that he didn't deserve his five (!!!) gold gloves. Even so, many don't know just how terrible he was, and the main statistical websites understate it. The problem with the way both Fangraphs and Baseball Reference calculate his WAR is that they use the least appropriate (and most favourable to Jeter) defensive metrics to calculate his defence. Statistics like UZR (used by Fangraphs), DRS (used by Baseball-Reference) and Total Zone (used by both for years up to 2002/2003) use stringers to describe the location and type of ball hit. These location-based metrics are subject to human error and bias. Even small biases will add up over time. Over larger samples, it is far better to use statistics that remove the subjective judgments of the stringers and rely on comparing the actual number of outs recorded to the average player (and adjusting for pitcher handedness, ground-ball tendencies, strikeouts and ballpark). Especially over larger sample sizes, these distribution-based metrics are significantly more reliable.

There are a number of different distribution-based metrics: Defensive Regression Analysis; Baseball Prospectus's FRAA; and (through 2008) Tango Tiger's fascinating WOWY measure (which is a little different - it uses pitchers as the fixed point, seeing whether Jeter got more or less outs than other SS with those pitchers). We also have an earlier version of FRAA (nFRAA) through 2010. Each of these distribution-based measures is much more negative towards Jeter's defence.

Over his career, these are Jeter's defensive values:
TZ/UZR (Fangraphs): -137 runs
TZ/DRS (B-R): -243
DRA: -352
FRAA: -305
WOWY: -436
nFRAA: -406

We have to add some 2009-2014 data to complete the WOWY data. DRA has Jeter at -95.4 runs during those years; FRAA at -76.3 runs. Being generous, let's use FRAA to round out WOWY to -512 runs for Jeter's career. If we substitute these superior defensive values for the faulty ones used by B-R & Fangraphs, Jeter's career WAR varies from a high of 66.4 WAR (B-R guts with FRAA) to a low of 36.4 (Fangraphs guts with WOWY/FRAA). It's only at the high end that Jeter's even a borderline hall-of-famer, where I'd let the round numbers and post-season make the difference.

I also think we have to question just how much credit Jeter should get for taking up the SS position while playing such egregious defence. All WAR systems give Jeter credit for playing a difficult defensive position (about 12 WAR for Fangraphs, 14 WAR for B-R). While it's sort of punishing Jeter doubly for his terrible defence, the fact he was so bad at defence also meant that the Yankees had poorly-constructed rosters for much of his career - most obviously, when they played Alex Rodriguez at 3B instead of SS, where he was still good defensively - and certainly better than Jeter.

TLDR: If you use defensive metrics appropriate to large sample sizes, Jeter is at best a marginal hall-of-fame candidate in terms of the contributions he made to his team, and at worst, clearly doesn't belong in the Hall.

Raul Ibanez got a text message from Posnanski congratulating him on making the HoF ballot.

Ibanez texted back: "I am? LOL"

OK, thanks. The argument is that he was the worst defender in the history of the game. Made Adam Dunn look like Ozzie Smith. I'll stick with UZR, DRS and what I saw.

As a note, the -436 number from WOWY is a number of plays; it was converted later in the same thread to -327 runs. I think it's likely that the advanced defensive metrics are right and Jeter was truly the worst defensive SS ever. Being compared to an average SS is a very harsh comparison group, because most SS's are very good defensive players, even the below average ones.

There's a good reason most players don't stick at SS for as long as he did, it's a tough and important position. His two signature plays as a SS, to my memory, are: 1. the jump-throw going to his right in the hole, which spoke to a lack of quick footwork and a strong arm that could occasionally make up for it - other SS's were just planting and throwing those plays, and 2. the "past a diving Jeter" play up the middle that he simply missed all the time, but seemed to be making a good effort on. He didn't make many errors, but it was more than offset by the lack of plays.

The end of this article about his defense (https://grantland.com/features/the-tragedy-derek-jeter-defense/) talks about how, at age 37, someone (Cashman?) actually told him and showed him how bad he was, and how he improved late in his career for several years. If that article is correct, Jeter was coached so poorly (i.e., not at all) regarding his defense, I'd almost give him some credit back for it. In the current shift-heavy environment, where players are positioned by the coaches (or given reference cards or something), he's probably "only" a bad SS instead of an amazingly bad one. Much like reliever-usage (entirely determined by coaches, some pitchers get extra leverage credit in WAR measures due to these choices), I think defensive metrics will increasingly have a problem disentangling coaching from player skill.

Jeter was a poor defender but the idea that he isn't a HoFer is a stretch at a minimum.
Bonds, Clemens, Jeter Schilling (unfortunately), Walker, Rolen, Ramirez, Jones, Helton, Sheffield

Pretty much in that order (though I didn't split hairs).

I would also vote for Pettitte and Abreu if I had more spots.

Notable 'No's - Sosa, Kent, Wagner, Vizquel

I think defensive metrics will increasingly have a problem disentangling coaching from player skill.

In fairness, though, isn't that true of all aspects of the game? It took Cito Gaston telling Bautista to open his hips earlier and pull the baseball to unlock Bautista's well-hidden underlying skill.

Metrics really just measure performance. So a player's performance could well be the by-product of substandard coaching that fails to exploit the player's skill.

Definitely true, and always has been I suppose. Jeter's defense is just one of the more prominent examples (Ortiz being coached away from power hitting by the Twins comes to mind too).

My 10 guys:
Clemens, Bonds, Walker, Schilling, Ramirez, Rolen, Sosa,
Sheffield (speaking of defensive problems), Pettitte, Helton

Yeah, I'll leave Jeter off. Since I argued for it, if you take off more defensive value to reflect what I think was likely the case, he's not a shoe-in at the least. Every so often there's debate about "what if player X makes a big round number" (right now I see it applied to Markakis and 3000 hits, previously it was Johnny Damon, or Adam Dunn and 500 HR), and whether they make the HoF anyway. I think Jeter is already that guy, he just happened to make it well past 3000 hits. Also, the years that Jays announcers spent slobbering all over him left a bad taste in my mouth.

Who's next to be scrutinized for their defense and Hall of Fame worthiness, Babe Ruth? Jeter not going to the hall, that's crazy talk.

So, what if Pujols wants to wear an Angels cap in the Hall.
Is that allowed?

I don't think that being overrated and a poor defender should be held against Jeter - neither of those things are his fault, unless you think he should've insisted that he be moved to 3B when Rodriguez was acquired. If the manager writes you into the lineup at SS, you're going to play SS whether you think you're the best option or not.

I never really liked Jeter, but I'm a round numbers guy. I'd vote for him in a second.

No offense Mike, but the analysis behind the WOWY and other metrics is a lot more advanced than a simple BABIP analysis.

FRAA describes Jeter here: https://www.baseballprospectus.com/news/article/12399/manufactured-runs-how-do-you-solve-a-problem-like-derek-jeter/

They account for pitchers, handedness of batter, parks, out-situation, etc. None of the systems were designed trying to make Jeter look bad, they just came out that way. Using bbref fielding totals, he's ~340 plays worse than average already. I think it's reasonable that the real number could be ~500. The "extra" bad plays only amount to one missed play every 16 games or so. For his career, 500 missed plays is only one every 5.5 games, barely more than one a week. My own simple analysis - he played 145 games or more 16 times, only rarely as a DH. He was top-ten in assists for a SS in 1997 (1st), 2005 (5th), and...that's it. He was pretty much always in the top-10 for games played at SS, just never made a lot of plays. Noted terrible fielder Michael Young played SS full-time for 5 years, and was top-ten in SS assists 5 times.

Jeter for the Hall? Sure, wouldn't argue against it. But obviously better than a fairly interchangeable half dozen guys on the ballot (Sosa, Abreu, Jeter, Pettitte, Helton, Sheffield, Jones)? Not to me at least. I'd put that entire list in of course, I love a big HoF to be consistent with the standards of the past. I think this is the first year a ballot with less than 10 players is defensible in a long time though, if you think more players should result in higher standards. If we use Harold Baines as a floor, there are maybe 20-25 guys you could put in?

AWeb, thanks for catching my error on WOWY. Filling out the missing years with FRAA or DRA, you'd be at -403 or -422 runs for his career.

Thanks also for linking to that excellent article about Jeter's defence. It's important to note that Jeter's poor defence isn't inexplicable - at least according to Bill James, Jeter positioned himself dramatically differently than Adam Everett, one of the best shortstops of the 2000s:

"Jeter played much, much more shallow than Everett, cheated to his left more, and shifted his position from left to right much, much more than Everett did ... Many or most of the good plays made by Jeter were plays made in the infield grass, slow rollers that could easily have died in the infield, but plays on which Jeter, playing shallow and charging the ball aggressively, was able to get the man at first."

It's an interesting question about how we should regard how poor coaching impacts player performance. On the one hand, it was hardly a secret that other shortstops positioned themselves differently, shifted position less, and so forth. I think Jeter probably needs to bear significant responsibility for his positioning and style of play, but you could share blame around to the coaching staff as well. Even the best defensive metrics can only tell us what happened on the field, not how to allocate responsibility for that performance.

More generally, I'm not telling anyone not to regard Jeter as a Hall-of-Famer. There are plenty of good reasons to put him in even if you conclude (as I do) he falls short of whatever regular-season-value benchmark one wants to use.

To be clear about Baseball Prospectus's numbers, the revised FRAA mentioned in the article about Jeter ended up being superceded - BP's current version of FRAA puts Jeter at -18 for 2009.

I mentioned this above, but the problem with the batted ball data used in UZR, DRS and TZ is that it's generated by stringers. On the margins, it's a judgement call whether to classify a batted ball as hard-hit or medium-hit, as a line drive or a grounder, as being in one zone or another. We should probably expect that Jeter's reputation and unusual positioning resulted in bias creeping in to a degree. Over smaller samples, batted-ball-based metrics will be more reliable; over larger samples (and Jeter's defensive record at SS is about as big a sample as any player ever generates) the distribution-based metrics should be expected to be better.

FWIW, Pettitte's ground-ball and pull rates are almost exactly the same facing LH and RH batters.  He's the other pitcher who was around for Jeter's whole tenure.  If you make broad assumptions from the L/R handedness split of batters facing Rivera and Pettitte about the number of ground balls that Jeter would see vs. the second baseman, you are going to be very wrong. 

Mike, everything you just mentioned is the type of thing that the advanced metrics take into account already (pitcher tendencies, L/R splits, etc., GB rates, etc.). They make Jeter look worse, not better, when they are taken into account. If the Yankees staff had weird SS-grounder prevention abilities, this would actually help Jeter, since he was bad and more chances means more bad results.

A quick check using the Jays as a comparison point in 2004 and 2005 shows almost identical numbers of grounders (in 2004: 1998 grounders against NYY to 2038 against the Jays, in 2005 the numbers were 2152 to 2132), at least on bbref. I know the grounder/liner/flyball numbers bounce around year-to-year sometimes as categorizing standards change, but there's not much reason to think this somehow make Jeter look worse. It's funny, and I'll probably drop the conversation here, because we're literally debating whether Jeter was the worst defensive SS in history, or the WORST defensive SS in history (obviously among those allowed to play SS for so long). Was he 25 wins below average (terrible) or 40 wins below average (super terrible), or maybe 50 wins below average (ultra-mega terrible)? In any case his bat (for a SS) makes his overall value strongly positive, the trade-off was worth it for the Yankees. I just happen to think he's a fine enough choice for the HoF, but I'd vote for other candidates I think are about equally deserving first instead.

Fascinating discussion, gentlemen. I've really enjoyed reading this. Thank you.

If the concern is that FRAA doesn't account for individual pitcher tendencies, then one should really like WOWY, where the primary mode of comparison is the pitcher with and without Jeter (it also adjusts for park, batter handedness and to a limited extent, batted ball classification). It's the exact answer to the concern raised.

Unfortunately, it doesn't help Jeter - WOWY is more negative on Jeter than either DRA or FRAA for the years we have. Unfortunately, we don't have WOWY for his entire career - I have no doubt its creator is selling the data to MLB clubs.

As an aside - TZ, which covers the first half of Jeter's career for both bWAR and fWAR, appears to have a flaw that results in systematically understating the cost of poor defenders (and the contribution of good defenders). You can see the way it works in comments #16, 20-26 in this discussion, where we got the WOWY numbers in the first place.

Small note - of course, there is no 4-5 hole, but 4-6. Unless we're calling Jeter the "hole" between the second baseman and third baseman.

Thanks, JamesW.  It was not a comment on Jeter's admittedly limited range...

On a Blue Jay note, I was and am inclined to the view that Wells was a somewhat below average defensive centerfielder (and was often criticized for that view here).  However the idea that he was the worst defensive centerfielder of the time and on average cost his club 15-20 runs a year with the glove as WOWY suggests would surely make some old-timers here shake their head. 

I think there's still a misunderstanding of what WOWY and TZ do, and I'll do my best to clear it up

WOWY has a world of other problems, in terms of scale, and Tango says so in the opening.

Tango initially has concerns about the greater range of ability implied by WOWY, true. But those concerns are addressed by the observations made in the comments that TZ structurally depresses the range - pointed out by Guy and agreed to by Rally, the architect of Total Zone: "I agree (confess) to pretty much every point Guy made" at comment 20.

And, by the way, WOWY compares Jeter with all other shortstops in the league in Yankee Stadium with a left-hander and a right-hander on the mound. It does not adjust for the particulars of his pitching staffs in terms of pull/spray by handedness.

This is not quite right - WOWY compares Jeter with all other shortstops playing behind each of the same pitchers. Adjustments are then made for park, batter handedness, and lastly, for GB/LD/FB splits. This is explained in comment #8. Because it looks at all the pitchers Jeter played in front of individually, it necessarily is adjusting for pull/spray tendencies of the specific staffs he's in front of (in fact, the specific pitchers).

As an aside, you can see from Rally's comment in #26 that he has things right.

This appears to misread Rally's comment. He already acknowledged in comments #20 and #25 the effect identified by Guy. In #26 he merely says he doesn't split responsibility for singles to left 50/50 between SS & 3B, but instead 40/60, and then says the technical difficulty of fixing the problem identified by Guy/Tango is beyond his ability/will to do it. It's a confusing because he's partly quoting from Tango in #24 without using quotation marks.

I don't agree but what about Statcast and centerfielders? Tango said in the article that hitfx will provide better evidence than WOWY. We now have that for outfielders and it suggests the scale is less extreme than WOWY would suggest.

A few observations about the Statcast range.

Firstly, Statcast is measuring in an era where teams have already integrated much better fielding data, and where defensive ability was already better understood and more emphasis was put on defence. The defensive revolution in front offices happened after Moneyball, basically hitting after most of Andruw's peak. This involved both better positioning, and defensive liabilities moving position or being dropped altogether. We should expect the defensive "floor" to have increased, and also the average.

Secondly, TZ thinks that Andruw had the 2 best seasons in history, and 3 of the top 9. He's second in career CF TZ to Willie Mays, who was an effective defensive CF for much longer than Andruw - and TZ is only available for about half of Andruw's CF career. It seems unlikely that the last 4 years would include a close rival to the player that TZ has already identified as the best CF in history.

Thirdly, Statcast doesn't include an arm component. UZR and DRS both rate his arm highly even in his post-peak years. You have to add that on to his range outs. I don't think WOWY takes into account the value from holding runners, but Andruw was well-above average in outfielder assists.

Lastly, I believe Statcast is based on the likelihood of making the out based on the initial positioning of the outfielder, so it doesn't capture any value based on superior positioning. I wouldn't be surprised if positioning was a very significant factor, especially in the 90s/early2000s before there was hit f/x data.

Putting that together, I'm not at all surprised by the variance of WOWY or the current variance of Statcast.

Well, if you believe the WOWY numbers, Andruw Jones was likely the real MVP in 1998, 1999 and 2000. He and Jeter would basically trade places on the career WAR list. I don't think it is true but I like Andruw Jones a lot more than Derek Jeter!

Just to add my annual post advocating for no bonds, Clemens or other proven PED users in the HOF.

It's the only enforcement left to punish cheaters. I truly believe that the lack of legitimate consequences for PED users is responsible for the Sox, Astros (and probably us) sign stealing scandals.

Until baseball takes a "kill one, scare the rest" mentality on cheating beyond Pete Rose this cycle will continue.

Andruw's a funny case for me, mainly because he went so completely off the cliff. Spoils one's recollection of him a little. Ultimately, I think he's quite a bit like Richie Ashburn - a tremendous defender in centre field, a good hitter - not one of the best in the league (or on his own team) - and someone who didn't age very well at all. (Obviously they were very different types of offensive players.) Ashburn did get in, eventually, some thirty years after he retired.

Hmmm. I think I'll go with Whitaker and Evans on the Veterans Ballot. Bonds, Clemens, Jeter, Schilling, Rolen, Walker, and Manny on the regular one.

Ashburn's always been an interesting player - hey, he was an original Met - mainly because his defensive numbers are absolutely insane. There's a simple enough explanation - Robin Roberts. It's as if they were made for each other. Practically every NL hitter who was active in the 1950s said Roberts had the league's best fastball (Ernie Banks said it was better than Bob Gibson's, but that might be the common case of a player being dazzled by the first great major league fastball he happened to see.) Roberts was an extreme fly ball pitcher, working up in the zone, in an era when hitters simply didn't strike out. And he threw so hard, it was hard to pull the ball against him. It was fly ball after fly ball to centre field. And Richie Ashburn ran underneath them and caught them.

I wonder who was faster, Ashburn or Devo? Ashburn could fly when he first came up. Saw him as a member of Eddie Sawyer’s 1950 Philadelphia Whiz Kids (several Toronto Maple Leaf grads on that team -- including Sawyer, Mike Goliat, Ed Sanicki, Stan Lopata, Jim Konstanty, and Willie (Puddin’ Head) Jones (1949 NL MVP), maybe others. My favourite name on that team was Putsy Caballero, also an ex-Leaf.) Old friend Russ Meyer was on that team, too, with several other ex-Cubs -- Waitkus, Borowy, Bill Nicholson.

Bob Carpenter, the Phillies owner, was apparently a friend of Jack Kent Cooke, the Leafs owner.

Ah, memory.

P.S. I’m hoping Dewey Evans makes the HOF, of course.

I may as well put up my hypothetical ballot:

Rolen, Schilling, Wagner, Walker

Obviously, after arguing Jeter's defensive shortcomings at such length above I can't vote for him here. I'm not going to be upset at his inevitable induction - he has the round numbers, the longevity, the fame, and the post-season success.

Apart from Jeter, Wagner maybe requires the most explanation. Wagner is the 2nd-most dominant reliever in history, after Mariano Rivera - and 3rd isn't particularly close. He didn't pitch as many innings as some other relievers, but once you're admitting relievers to the hall you're implicitly looking more at dominance than volume.

Obviously, I'm leaving the steroid-associated guys off. Bonds, Clemens, Sheffield, Pettitte, Manny and Sosa have the numbers, but aren't on my hypothetical ballot. In addition to Jeter, I'm open to discussion on Abreu, Helton, Kent and Andruw who all have decent arguments.

Well, if you really buy the WOWY over DRS and UZR, Andruw Jones is a must.  During the period 1998-2000 using DRS, Andruw Jones was 2nd in baseball in WAR just behind A-Rod (23.6 to 22.7 for A-Rod).  If you think DRS undersells him, he's the best player in baseball over a 3 year period and obviously the key player on a winning team.  Those guys ought to get in, leaving aside arguments about whether 67 or so career WAR is enough. 

Andruw definitely had the peak you'd associate with a hall-of-famer -it's the relatively short career that makes him a question mark. Yes, WOWY would imply he reached a career value that would make him a presumptive hall-of-famer; other defensive metrics such as DRA & FRAA (and TZ) are not quite as positive. I haven't taken a deep dive into his hall-of-fame case, but I'm definitely open to being convinced.

My point about WOWY was not that it's completely authoritative - but that it's a good methodology that directly addressed your concern about distribution-based metrics neglecting to account for which pitchers a defender played behind. I'd also note that I have a little more uncertainty about defensive metrics for outfielders as there's a lot more balls that could be caught by different fielders.

The Jays have signed Cole!!!!

AJ Cole, that is (to a minor-league deal).

Yeah, the headlines saying the Jays signed pitcher Cole messed me up too when I first saw them - thought 'you gotta be kidding me', then saw it was AJ Cole and it was 'makes sense'.

Here's a fine piece on Bobby Abreu from Jay Jaffe.  Jaffe is getting a ballot starting next year, and it is small changes like that which should in the long run lead to better Hall decisions.

I missed this fine piece on Andruw Jones from Jay Jaffe.  Jones is a better version of Vada Pinson, and was a key cog in the Braves' success for a number of years (indeed their best player over a prolonged period).