Ripping Pos a New One

January 17, 2020
That’s what I’m not doing here, ripping Joe Posnanski a new orifice for re-starting his "100 Top Major Leaguers of All-Time" list after he was forced to abandon his previous attempt in mid-countdown because (I think) he learned belatedly that he had more top-quality players left to put on his list than he had places remaining in the list.  I think that’s what happened—it’s certainly what his readers were gently questioning him about, and then not-so-gently wondering about, and then complaining about when it became obvious that the only way he’d be able to squeeze 43 more top-rate stars onto a list that had 39 places left to fill would be to arrange for four spots to be tied, or something of the sort, which kinda negates the point of making a "top 100" list in the first place. I suppose you could just re-title it "My Top 104 Players" but Pos just bailed on the project, and now has brought it back to life, starting over from scratch. He’s currently halfway through his new list as I write.

But I’m not here to bust Joe’s chops. The point of such a list is never to read any writer’s opinion of which player ranks above or below another player, but rather what the writer has to say about each, or the writer’s methodology, or something other than the list itself. It’s probably a good idea to plan carefully who goes where before you get started publishing the list, but, hey, I’m not here to bust Joe’s chops.

I’m here to bust the chops of list-makers in general.

Who cares that you’ve ranked Grover Cleveland Alexander five places above (or below) Lefty Grove? (Early in the second half of the 20th Century, when I was first getting curious about baseball in the first half of the century, I used to get these two greats mixed up all the time—each was a top pitcher before I was born, and each name contained a "Grove." Even today, I sometimes have to think which giant is which.)  I certainly don’t care, though I’ll listen attentively if you’ve got something to tell me about either man. The ranking is the least part of my interest in what you’ve got to say. The order you list players in is, ultimately, going to rest on your opinions, judgment calls you make on a variety of unrelated subjects such as recency bias (and its lack), where in time you cut off your list (Do you include 19th century stars? Do you include current stars in mid-career? Do you penalize either? If so, how severely?), your definition of Major League-quality players (what do you do with Negro League stars, for example, or Federal Leaguers?), what you do with missing seasons for military service and other causes, how you’re balancing peak value with career value, and on and on. Depending on your answers to these and many other questions, your Top 100 listing is bound to differ from mine and everyone else’s in a way I find coma-inducing.

Unless you can write about the players, or explain your methods, in a fresh, insightful way. Fortunately, Joe Pos can, and so can Bill James and a tiny handful of other writers who just use their rankings as jumping-off points for mini-essays, without which I wouldn’t care a damn about their lists.

From a historical perspective, it might be interesting to see what experts in each section of these mega-lists make of players within their own spheres of expertise. If John Thorn would rank the 19th century’s stars, for example, and some expert on Japanese baseball or Mexican League baseball would rank his own players, and so on, and then perhaps some generalist could decide how to blend all the specialists’ lists together, so it’s not so much  the expression of some subjective opinions as much as it is a consensus of expert thinking.

Perhaps more interesting would be to examine the differing opinions of various experts within an area of specialization. One of the more complicated and oddly satisfying discussions I ever engaged in was on a Mets’ fansite that I founded in the early 2000s, where we attempted to rank each player on each yearly iteration of the Mets’ roster that worked off a model of specialists>general discussion>revision by specialists.

The discussions within each year were fascinating in themselves. One of us nutty Mets fans was given the responsibility for ranking the players from their chosen years, where presumably they knew the players better than anyone else, though everyone was welcome to chime in, offer dissenting opinions, re-rank the players-- the chief ranker could, and often did, then revise his original list depending on which arguments he found persuasive. As it happened, my task was ranking the 1962 Mets, and in the middle of the discussion, I realized that Bill had ranked that team in order to illustrate some of the principles of his Win Shares book. So I incorporated Bill’s rankings into my own, and those of various dissenters, and we looked closely at the contributions of that wretched hive of 1962 scum and villainy. As I recall, at one point I was seriously arguing whether Gus Bell (.149 in 101 ABs) or Don Zimmer (.077 in 45 ABs) contributed more to Mets’ wins in 1962 –Bill and I differed on their relative contributions.

Eventually, the entire yearly debate got morphed into a synoptic ranking of the value each Met player from 1962 to the present day has contributed to the team. Far as I can tell (I got kicked off my own website over a decade ago), it’s still going on to this day, and is updated down to the third decimal point on a constant basis. It’s a ranking project that makes some sense to me, in that its goals are clear, and its methods are transparent.

It is the blending of various experts’ lists that brought that Mets-project to mind. In drawing up my initial numbered ranking of 1962 Mets, for example, I would first make separate lists of pitchers and non-pitchers, just to keep straight my apples and my oranges. When I had sorted out the best (or least-worst) pitchers, which is a hard enough task, I would have to slot them into my list of least-worst hitters. Separating longer lists into shorter, more specific lists yields a more manageable result.

Bill spoke about this process a few years ago in a public interview Rob Neyer did with him on Cape Cod, where he described his own method as breaking down large, unanswerable questions into smaller and smaller specific questions that did have answers. In this clip, at around the 14th minute,, Bill credits the idea to his 19th century namesake, William James, who "has a section in one of his books in which he talks about the highest level of questions and the lowest level of questions." The idea is to understand the huge, broad, unanswerable questions by figuring out their more manageable components. James’ paraphrase of James: "Building understanding is a matter of breaking down the great questions, which have no real answers, into smaller questions which have slightly more objective answers, and smaller questions yet, which have slightly more objective answers, until you get down to the level at which you reach questions which have answers--and then you can build up."

If, for example, I divide my list of 1962 Mets pitchers even further into "relievers" and "starters," and then divide that partial list of an already partial list into, say, right-handed relievers and left-handed relievers, then I’ll find that I’ve got a few (but very clear) answers: Ken McKenzie led the Mets’ lefty relievers in IP, in ERA, in WHiP, in W-L percentage, etc. , while Bob (Lefty) Miller threw only 20 ineffective innings, so one of them goes to the top of this mini-mini-list and the other goes on the bottom.

In that way, we can work off a base of knowable (if somewhat trivial) information, or at least hash out a broad discussion in manageable terms.  If I were drawing my own list of Top 100 players of All-Time (which I never will, because who cares?), I would want to start as small as I could. To revisit briefly one of the thorniest questions I have grappled with in these electrons, rather than trying initially to rank all of MLB ever, I would begin by trying to settle the issue of which Willie is greater, Stargell or McCovey.

Deciding such a small matter enables me to decide what principles I will follow. How important is career WAR in my computations? How about Win Shares? Lifetime OPS?  Games played? Championships won? Five-year peak? All-Star appearances? H.R.?  If I follow one set of criteria to adjudicate McCovey v. Stargell, then how I can apply different criteria to et al.?  Applying Occam’s razor, I certainly don’t want to set up competing criteria unnecessarily. That smacks of custom-tailoring my rankings to accommodate my opinions to each case, which smacks of creating the whole list in the first place in order to foist my opinions on the world, and who wants that? Not you, not me, not Occam, and not Lefty Miller.

When you begin answering a large question by settling a smaller question, as Bill suggested on Cape Cod, you establish a small but solid base of knowledge. If you adjudicate McCovey v. Stargell to your satisfaction, or any other case, you’ve got principles to work with and you’ve got precedent. By doing head-to-head comparisons, you get a basis for ranking Gehrig above both Willies, and for Bill Buckner ranking below. It’s not just your opinion, and you can use the criteria you’ve established to settle closer cases at all points on your first-basemen list.

Breaking lists down by position is, of course, far from the only way to break them down, but it seems reasonable that all your breakdowns ought to be consistent or else made to be so. (It was only a foolish consistency that Emerson warned against—nothing wrong with consistency itself.)  It might be (I’m just hypothesizing now) that McCovey ranks lower on your list of "1960s NL sluggers" than Stargell does, and that Stargell also wins your list of "1970s NL sluggers," so how might McCovey rank higher than Stargell on your overall "First-basemen" ranking? If he does, then you’ve got to look at your lists and figure out where you went wrong, or else make sense of the seeming contradiction. (Some possible solutions: your "1970s NL" list doesn’t credit McC. for his time served in Oakland, and your "1960s NL" list doesn’t credit him for his year of 1959, so those might represent Willie M.’s advantage over Willie S.  Or perhaps Stargell’s edge in games played at first base (assuming he had one) would push him past McCovey as a first-baseman while McC. keeps the edge in career slugging. Or, of course, it might just be that all of these explanations are insufficient and you’ve screwed up on one list or the other. But settling the issue leads you to find flaws in your process.)  

The idea of making up disparate lists and then blending them is what I took from the task of ranking Mets players. If I ended up, as I did, with two lists of 1962 Mets batters and 1962 Mets pitchers, ranked in order, I had to figure out then how to make one list out of the two, which was challenging. The Mets, as I recall, who had the strongest seasons with the bat, Richie Ashburn and Frank Thomas,  exchanged places, #1 and #2, several times during the discussions we had. But there was a large drop-off between Ashburn and Thomas and the #3 hitter. Did their top pitcher belong in the #3 spot? Or were there four, or six, Mets batters who had stronger seasons than any Mets pitcher? Or was that an illusion based on the Polo Grounds' compact dimensions? (The final ranking, minus the protracted discussion, is here: .) I ended up deciding that Al Jackson needed to be slotted in as #3 Met behind Ashburn, and since Roger Craig had very similar numbers to Jackson’s, he would go right behind Jackson, and then perhaps the third-best hitter needed to be ranked as the #5 Met. By breaking down the whole roster into its component parts, I found the task far more manageable and sensible than tackling the whole roster at once.

I recommend Pos’s current list (available behind The Athletic’s paywall, I believe) of his 100 Greatest, because I’m pretty sure these will be 100 of the greatest mini-essays ever written about baseball players, though the concept of such a list doesn’t rise much above "clickbait." I clicked, in just this week, on a lot of worse websites than Joe Pos’s.  


COMMENTS (25 Comments, most recent shown first)

I remember that first ranking, he got about....60 percent of it done then bogged down, was going to try again--I thought this current try was that one. Missed the second run. Hrmmmm. Well---reading what he has to say about these guys can be fun, or, it can be a waste. I'm not one of his biggest fans to be honest. He should put out a book if he thinks it's all that.
5:22 AM Jan 24th
They are great articles. If we are whining about methodology, and that brings more attention to his writing, then we are accomplishing something.
1:54 PM Jan 23rd
Ah, I misunderstood. Thanks for the clarification.
6:23 AM Jan 20th
Yeah, Steven's approach, not Posnanski's. Bill's Twitter polls were based on a rigorous weighting system which he explained in an article here.
9:09 AM Jan 19th
Steven Goldleaf
I think what michaelplank is saying, steve161, is that MY method of starting with head-to-head comparisons, not Joe Pos's system, resembles Bill's Twitter-polling methodology. If that is what he's contending, I take it as a great compliment: another way to pose the issue is "How can you answer a broad question if the component parts of it are still in doubt?"

To use my example again, if we break the 30-odd 1962 Mets into seven groups-- righty batters, lefty batters, switch-hitting batters, righty starters, lefty starters, right relievers, and lefty relievers--we can arrange them into hierarchies of under ten players apiece, and then settle such questions as "Who was the best righthanded batter on the team?" more easily than we can the broader, and more difficult to measure, question of who was the best player on the team. Once we have each smaller group's hierarchy set, our task begins to fall into place.

You could do other ways, of course. Arrange them into groups like "batters with over 300 plate appearances," etc. and again they will be simpler to arrange than tackling the whole roster at once.
7:25 AM Jan 19th
Beg to differ, Michael: what Joe Posnanski is doing is the diametric opposite of a Twitter poll, which has no criteria at all, except for the random opinion of the tweeter's followers.

As someone whose baseball abilities were accurately described as "Poor field No hit", I don't believe I ever hit anything out of the park. So thanks, Radar, for the kind words, though most of what's in that post was written by Joe Pos, not by me.
3:26 AM Jan 19th
This apprach has a lot in common with Bill's Twitter polls: who's the best of these 4 Cardinals, these 4 firstbasemen, these 4 guys born in 1947, etc., then cross-reference the results.
5:36 PM Jan 18th
I apologize - as I re-read what I wrote, I can appreciate the inference drawn, but that certainly was not my intent. I think it's easy to differentiate this column (and the others that you write, and the others on this site) from the invective that some (but, honestly, not very many) authors choose to foist on us. Commenters (not on this site, thank goodness, which is yet another reason I'm a faithful subscriber) are far more frequently guilty of taking shots at authors and at each other.

Mainly what I was thinking is that, having read Joe's stuff for a bunch of years, and having read a couple of his incomplete attempts at the top 100, I'm disinclined to split hairs about his methodology. I know he's been thoughtful and not random about it, and I can respect that. Calling his compilation a list of the top 100 provokes discussion, which I think is great.

One of the many things I love about baseball is that it's easy and fun to talk about comparisons, whether it's Hall of Famers, Hall of Very Gooders, or the game's Duane Kuipers. HOF voting time is one of my favorite times of year because we get to read so many interesting takes on so many great and very good players. Sure, some stuff I read makes me roll my eyes. But I'm probably still at least a little better off for reading anyway. And if I don't like what a particular author keeps putting out, I can stop reading what s/he writes.

I think steve161 hit it out of the park with his comment.
12:48 PM Jan 18th
There's nothing to really disagree with! That's my point. You have hundreds or thousands of REASONABLE paths to create a list. It's easy to see a path that has Koufax at 70. Or at 50. Or at 150. Or at 300. It's easy to see Clemens at 5 or 5000, depending on your assumptions.

The only thing you can disagree on is results that can't be justified by any REASONABLE set of assumptions. You can't have say Barry Foote ahead of Gary Carter. There are no reasonable set of assumptions that will allow that ranking to exist.

If you can imagine the assumptions that Koufax is #70, great. If you can't, then you are trying too hard to reject reasonable assumptions.
12:19 PM Jan 18th
Steven Goldleaf
"You can literally come up with thousands of ways to rank players, all based on whatever assumptions you come up with."

Isn't that exactly what I'm saying here, Tom?

That's why I want to denounce the whole idea of list-making, other than for the purposes you describe: as a basis for writing essays about the players, or to explain your criteria, using the list for examples of how your criteria play out, or both.

If I understood Joe Pos's criteria better, or yours in aiding him, no doubt I'd disagree with some of them, but that's a more sophisticated and more interesting discussion than the Lefty Grove/Pete Alexander discussion., don't you think?
11:51 AM Jan 18th
You can literally come up with thousands of ways to rank players, all based on whatever assumptions you come up with.

The issue is that once you decide to create ONE list, you are locking in ONE set of assumptions. Those assumptions are treated as fact for the purposes of making the argument. That's how arguments work.

If you don't want to have an argument, then you don't make a single list, and don't lock in to one set of assumptions.

Personally, I find it ridiculous that anyone would argue with a list. As Bill James once said, paraphrasing: If you would have done it differently, that's because I did it, and you didn't.

So, don't complain, create your own list. Then have the same number of people complain to you. And you will see how ridiculous all the complaining is.
10:32 AM Jan 18th
Steven Goldleaf
Another part of Cranepool Forum rankings project, I just realized, was its crudity and efficiency. We arbitrarily alloted a fixed number of spots, 30 every year, awarding 30 points to the Met who in our collective judgment was the most valuable, and 29 points to second-Most Valuable Met, etc., so that in the total ranking a long-term productive player might accumulate a few hundred points as a Met in his career. It's obviously a somewhat crude system but it was also pretty efficient in assigning credit over the decades to contributors. I imagine Seaver or Strawberry or Wright might lead with close to 300 points, and who could dispute that this assessment is on the money. It was fun making these assessments every year, and it was easy, although not simple, to do.
10:09 AM Jan 18th
Steven Goldleaf
I hope you're not numbering me, Radar, among the critics who sit behind their keyboards typing stuff that would get them punched out in real-life, and I trust you're not. It did get a little testy on Joe's old page (I'm too cheap to subscribe to the Athletic, but that may change soon), as it does here at times. Joe is very civil, almost too much so, as he describes in the SABR-cast, but a writer does have a duty to entertain and to speak the truth, which on occasion gets in the way of that civility. I must admit, that was the quality that first attracted to Bill's writing--the caustic (bordering on nasty) way of offering criticism that I found fun to read, and which he has toned down considerably. But it takes a certain wit to do that, a kind of wit that few people can master.
9:03 AM Jan 18th
Steven Goldleaf
Yeah, he touches on some of that in the SABR-cast, too. The one thing I would fault him on, if we're faulting him, is not explaining the system he (and Tom Tango) came up with, so we can tell where we disagree with his criteria. The authoritarian part of lists is that part that gives us the results but not the methods, and I'm not big on authoritarianism. I'd probably be as pleased if Joe gave us a list of "A lot of players who I want to write essays on" rather than 100 players ranked in inexplicable order of greatness. If he doesn't give his methods, or if those methods don't stand up to scrutiny, or if he doesn't want to defend his choices, what's the point of a ranked list?
8:10 AM Jan 18th
Steven Goldleaf
Yeah, he touches on some of that in the SABR-cast, too. The one thing I would fault him on, if we're faulting him, is not explaining the system he (and Tom Tango) came up with, so we can tell where we disagree with his criteria. The authoritarian part of lists is that part that gives us the results but not the methods, and I'm not big on authoritarianism. I'd probably be as pleased if Joe gave us a list of "A lot of players who I want to write essays on" rather than 100 players ranked in inexplicable order of greatness. If he doesn't give his methods, or if those methods don't stand up to scrutiny, or if he doesn't want to defend his choices, what's the point of a ranked list?
8:10 AM Jan 18th
A number of us read Joe’s work because he’s an awesome storyteller. I have told many friends and acquaintances that Joe could write about paint drying or grass growing and make the story engaging. So reading his essays on lists like this is less about whether Clayton Kershaw should be above or below Miguel Cabrera, or whether PHil Niekro or Kid Nichols should be on the list at all, than it is about reading engaging stories about the players, and/or about baseball.

As one of Joe’s loyal fans, I have been frustrated with his multiple failed attempts to get all the way through the list. As he began this attempt, I wrote to him and suggested that he risks disenchanting even his most faithful readers if he doesn’t get all the way through the list this time.

As for engaging in list making, I don’t see a problem with the concept. Of course spririted debates will result. For a number of us, we learn more stuff that way. I don’t have to agree with Joe’s list, any more than I have to agree will Bill James’s approaches to ranking players, to enjoy reading the rankings.

I think that the problems with publishing lists lie not with list-makers, but with critics who sit behind their keyboards and type the kinds of nasty and insulting things that rightly would get them punched in the mouth if they said them face to face. That leads me to another enjoyable and refreshing aspect of Joe’s writing. People who comment on Joe’s posts are unfailingly civil. The commenter group is self-policing. The occasional troll is dealt with in such a way that the troll gets no pleasure out of diving in, and so goes away.

Now, I’m going to read Joe’s piece on Monte Irvin (No. 69), and I know I will be smiling when I am done.
8:04 AM Jan 18th
As a subscriber to The Athletic, I actually get to read Posnanski's essays (Number 69, Monte Irvin, appeared today, so he's still some distance from halfway through: the plan is to finish on opening day). On the one hand, he talks about working very hard to come up with his rankings, of which he says, "I spent many, many, many hours on them. I used the Tom Tango-inspired formula, added a bunch of wrinkles, did a bunch of research and made some hard judgments that I believe in."

On the other hand, "I don’t care much about the rankings." The main purpose of the list is to provide a framework for the individual articles. He predicts his readers' response: "You will probably get mad when you see which players I have left out, which players I have ranked way too low or way too high. You might want me to know just how dumb I am, just how little I know about baseball, just how insulting the ranking was. I totally get it. And I totally deserve whatever you are going to say because it takes some serious gall to believe that you can really rank the 100 greatest baseball players ever."

Many of the readers' comments meet this expectation. Many more sneer at the first group for missing the point, as described in the foregoing.

Posnanski does explain why his first two attempts failed: "At the time, I imagined writing just a few words on each player — a paragraph or two — and spreading it out over a baseball offseason.

"But it didn’t work out that way. The trouble is that I am utterly incapable of writing 'just a few words' on great baseball players. And so the stories began to get longer and more involved and longer and more involved until this project overtook my every waking thought...

"The series stretched out for so long that I began to find the rankings out of date. And finally, I simply ran out of time and space. The first attempt crashed around No. 30.

"I began the series again last year because I got very excited about this new ranking formula that that estimable Tom Tango helped me come up with. That series flamed out more quickly."

I also don't care about the rankings, whether here or anywhere else. To me they are the baseball equivalent of angels and pinheads. But there aren't very many out there who can write like Joe Posnanski. (Most of the few who can also write for The Athletic.) So far, every article has been up to his usual standard, and I don't doubt that the next 68 will be as well.
7:58 AM Jan 18th
Au contraire! Posnanski is rivaled perhaps only by the Current Occupant in his predilection for naming everything he does after himself. It's quite possible he's reading this while taking a Posdump....

I was an avid reader of Pos's initial list. It had no value as a definitive ranking, but as you say it provided a jumping-off point for essays in his inimitable style that were fun to read.
3:35 AM Jan 18th
Joe would probably appreciate not being called pos.

8:52 PM Jan 17th
Steven Goldleaf
OK--on the SABR-cast, Pos doesn't actually mention the specific glitch I pointed out (about having too many players left over to shoehorn into his remaining spots) but does describe his aborted attempt at a top-100 list as less than a masterpiece of planning. I didn't follow it that closely to begin with, for the reasons I cite above, but he dwells on the list stemming from some cockamamie ranking system he invented, which in mid-list he changed up, which is never a good idea, and which he blames for the abandoned list. As I say, not a candidate for the Guinness World Book of Brilliantly Conceived Ideas, but an entertaining read nonetheless.
6:10 PM Jan 17th
Steven Goldleaf
Did I come close to getting it right? I'm going to listen to the SABR-cast soon.
3:58 PM Jan 17th
He addressed why he stopped and why he re-started in the latest SABR podcast.
3:38 PM Jan 17th
Finished your article. All I know about the relative rankings of players on the 1961-62 New York Mets is:

"As a member of the starting rotation of the 1962 and 1963 Mets, who lost 120 and 111 games, respectively, pitcher Roger Craig posted a nightmarish 15–46 won–lost record during his two seasons with the expansion team... Despite his poor record, Craig was a stalwart of the legendarily bad team's pitching staff. Remarkably, he threw 27 complete games in 64 starts, demonstrating that he was one of the Mets' best pitchers.

His manager, Casey Stengel, told him, "You've gotta be good to lose that many."
1:39 PM Jan 17th
After reading your first paragraph, I thought, "This is the sort of thing I write that gets me once again 'disciplined for violating my double-secret probation bestowed for having unknowingly twice violated the site's posting guidelines." I'm impressed. I'm so impressed I am going to read the entire post, which I was planning to do anyway, but reckoned I should comment first before I did just because...
1:31 PM Jan 17th
I finally agree with you on something. This makes a lot of sense.
9:50 AM Jan 17th
©2020 Be Jolly, Inc. All Rights Reserved.|Web site design and development by|Terms & Conditions|Privacy Policy