rolling sabermetrics and statistics thread

win probability added divided by leverage index

it's generally about the same as the batting component of WAR though, so not very useful. plain WPA is more interesting and kind of like the sabermetric equivalent of RBIs.

Thanks. I'm dying to find out what Ray Oyler's Leverage Index was for 1969--not very good, I'm guessing.

clemenza otm. James revolutionized my thinking during the mid/late 80s, but my enthusiasm for the game waned in the 90s when it seemed like the New Math was press-ganged into service to fantasy baseball. In the late 90s I sort of got my mojo back by actually watching a lot of baseball instead of reading about a lot of baseball. Sabermetrics are good corrective lenses, but I had to remember to use them to watch baseball games, not read box scores. I get a lot more fun out of the game that way.

leverage index is just a value which quantifies how "important" any single plate appearance is to winning that game

so bases loaded, 2 outs in a tie game would have a really high value whereas bases empty in a blowout would be really low

Has a leverage index been developed for relief appearances? Inherited runners stranded/scored drives me bananas.

Admission I'd rather not make: I'm still stuck in a place where I follow baseball primarily through the lens of statistics (more traditional statistics, but statistics nonetheless). Getting back to actually watching more baseball is my next therapeutic goal. (Part of this does have to do with the overload of baseball on TV. Somewhere along the way, it just became too much.)

well yeah it works both ways

if a game situation is a 1.5 LI for the hitter (1.0 is average) then it's a 1.5 for the pitcher too by definition

tbh i'm a math nerd and i love all the baseball stats stuff but i think sabermetrics folks tend to have too much confidence in their own metrics, there's not nearly enough self-evaluation in the "field".

i think the offense stats are pretty close to complete but there's still so much we don't understand about pitching, let alone defense or player development

I think there is plenty of self-evaluation in the field; the BP guys debate stuff all the time, and most if not all recognize that these metrics are imperfect tools.

(not that I have the time to read all the articles or watch a game every day, understand)

i think the offense stats are pretty close to complete but there's still so much we don't understand about pitching, let alone defense or player development

Agreed. And just to really horrify VORP disciples, I think you can even learn something from Joe Morgan when it comes to "in the field" stuff. I realize Morgan is considered a human punchline by most everyone who's been influenced by James, but if you can look past the many blind spots that someone of his generation probably carries around (having to do with character, clutch play, the value of a .300 average in and of itself, etc.), there are going to be some things that he's learned about the game that I just don't believe you can arrive at through abstract statistical analysis. So treat him skeptically, for sure, but don't try to ridicule him out of extistence. (When the influence of James on me was at its peak in the late '80s/early '90s, Kubek used to drive me up the wall for the same reasons.)

So treat him skeptically, for sure, but don't try to ridicule him out of extistence.

Not sure I can follow you this far.

Morgan does not deserve to have the most prestigious color commentary job for baseball in the world. That doesn't mean he's never insightful. But the fact that he has insights into baseball is meaningless. He lacks the ability to express those insights or the work ethic to learn about the teams he's watching. Orel Hershiser is a far better analyst. Keith Hernandez is another guy I've enjoyed. Neither of those guys is a stats guy, but both actually do the legwork to bring some on-the-field insight to the presentation.

Also, while I have my issues with Morgan, there are plenty of guys who are worse. Rob Dibble springs to mind.

Orel Hershiser is a far better analyst.

Which reminds me, John Smoltz has been an absolute treat this year since he sorta retired.

Keith Hernandez is listenable mostly for the crazy shit he comes out with. He calls Jeff Francoeur "a streaky hitter."

I probably haven't listened closely enough to Morgan to be defending him. The back-and-forth between Miller and him is easy enough on my ears that I've never quite understood the intensely negative feelings about him that I keep encountering, but maybe that's all credit to Miller. And I have the additional bias that the mid-'70s Reds were my favourite team. This goes back a ways, but I used to think Palmer, Seaver, and Reggie were great in the booth. As analysts, I can't remember. I just liked them.

to clemenza, i don't know dude, you seemed pretty "blinders-on" when discussing Bonds' achievements and were in the process of pooh-poohing him on the Pujols thread so........... idk, was tbh hard to take you seriously in your short shrift dismissal of him as a legit triple crown candidate given the 232 BBs that got in the way of him chasing such a "retro-cool" counting achievement (all the while destroying almost every offensive record in the process).

but kudos to all y'all who were reading bill james in the summer of love~~~

i don't mind Morgan and Miller, because they're both local guys. Morgan seems way worse on the page then in the booth ime.

francoeur is a streaky hitter in that he has a lot of cold streaks

Steroids notwithstanding, I really wasn't discounting the magnitude of Bonds's statistical achievements--just that I thought the walks ruled out him ever having a realistic shot at the Triple Crown. Not just in terms of RBI, but, I thought, also in BA. But Ciderpress's math made me realize that he in fact likely would have won one, maybe even two. Which was your point to begin with--you were right, I was wrong. What I didn't appreciate, though, was pulling out VORP as kind of a gotcha moment, like I'd just been teleported out a 1974 issue of Baseball Digest. (Not to knock BD, which I used to love.) Again, I've been reading James forever.

That's cool, just like I said, you appeared to be full blinders in your take on Bonds achievements.

Also, VORP was introduced 9 years ago by Keith Woolner. Bill James has always preferred win-shares and runs-created in my 10 years of being familiar with SABR. IIRC.

who went to the pitchF/X summit in SF?

i liked this little bit from the pitchFX summary on bpro:

5:39: Brad Hawpe play: starts with a >80% chance of catching the ball, but freezes in place and fails to make the play. Difficult to represent visually, because the out probability plummets while Hawpe stands in place and time elapses. In a different Hawpe play, his first step gives him a lower probability of catching the ball, since he broke in the wrong direction. Rumor has it no Rockies reps are in attendance, but they’re not missing out, since they’ve enjoyed a front-row seat for this sort of action for the last several years.

haha

http://www.baycityball.com/wordpress/wp-content/uploads/2010/08/chernoffteams.png

wau

source?

http://www.baycityball.com/2010/09/01/wednesday-graph-major-league-chernoff-faces/

players w/an OPS of 1.000 or more in 2000

Todd Helton
Manny Ramirez
Carlos Delgado
Barry Bonds
Jason Giambi
Gary Sheffield
Vladimir Guerrero
Frank Thomas
Sammy Sosa
Moises Alou
Jeff Bagwell
Nomar Garciaparra
Richard Hidalgo
Alex Rodriguez
Brian Giles
Jeff Kent
Mike Piazza
Troy Glaus
Edgar Martinez

players w/an OPS of 1.000 or more in 2010

Josh Hamilton
Miguel Cabrera
Joey Votto
Albert Pujols

jim thome and justin morneau deserve a mention on that too for partial seasons of 1.000+

man some of the names on that 2000 list

Richard Hidalgo is the one that jumps out at me. I sort of remember him...I think.

I just looked him up--wow, his 2000 was huge. Should have mentioned him on the fluke thread a while back (although he did a few other good seasons).

troy glaus didn't even get an mvp vote that season (though his teammate darin erstad did with one of the weirdest fluke seasons of them all)

for good measure, the players this season who finished between .900 and .999 --

José Bautista
Paul Konerko
Carlos González
Troy Tulowitzki
Matt Holliday
Jayson Werth
Adrián Béltre
Robinson Canó
Adrián González
Luke Scott

and in 2000:

Jim Edmonds
Bobby Abreu
Chipper Jones
Edgardo Alfonzo
Will Clark
David Justice
Carl Everett
Bernie Williams
Rafael Palmeiro
Jermaine Dye
Darin Erstad
Geoff Jenkins
Tim Salmon
Jorge Posada
Ken Griffey Jr.
Luis Gonzalez
Mike Sweeney
Jim Thome
Jeffrey Hammonds
Scott Rolen
Jose Vidro
Magglio Ordóñez
Phil Nevin
Bobby Higginson
Ryan Klesko
Travis Fryman
Andruw Jones
Steve Finley
Sean Casey

Didn't realize Luke Scott had such a good year.

I know nothing about sabermetrics. (Well, I read Moneyball once.) Where do I start? Go buy some of Bill James's old 1980s Baseball Abstracts on eBay?

That's a good question. Is there a beginner's section at BP? Sabr 101?

http://playahardnine.wordpress.com/saber-101-saber-201/

http://www.baseball1.com/bb-data/bbd-bj1.html

What about OPS+ for 2000 vs 2010? The league difference is probably 40-50 points of slugging. I'm betting that those lists nearly even out if you use OPS+ > 135 or 140 as the cutoff.

context makes the length of the list -- see 1930 vs 1912

Lisa Simpson as sabermetric Little League coach right now...

Bill James: "I made baseball as much fun as doing your taxes."

"It's the triumph of number crunching over the human spirit"

I don't know how many of you (if any) subscribe to James's site. He's been great lately. Right now, I get the feeling he is to sabermetrics what Andre Bazin ended up being to auteurism: original inventor (not quite true of Bazin, who was more the inspiration), now a somewhat skeptical onlooker. Bazin had that famous quote: "Auteur, yes--but of what?"

Here's James a couple of days ago on jargon:

Along the lines of your "PDO" story. . .at a spring training game in 2004 I was sitting in front of Mark Bellhorn's wife and son, and this boy, who I think must have been four years old at the time, is kind of chattering to his mom about the game. A player comes up that he doesn't know, and he says. .. I swear I am not making up one word of this. . . "Is he a good player? What's his on base percentage against left-handers?" That will always stick with me as the moment at which I realized that sabermetrics was mainstream, hearing this kid who I am sure hadn't started school yet ask about a player's on base percentage against left-handers.

But the question you pose is more central than you realize, I think, because what you are really asking is "How do you reach the public with your information?" I think the distinction I would make is between careless and careful progress. After RBI were introduced to the public and explained to the public, about 1912, there as eventually a column added to the Sunday batting summaries in the paper, "RBI". Somebody who saw the new column and didn't understand it could ask "What is this, RBI?", and there was probably a code at the bottom of the column that explained it.

That is CAREFUL progress. On the other hand, people will write articles in which they introduce LIPV (Leverage Index Performance Variation) and PAD/1000 (Pythagorean Advantage per 1000 games) and EBOR (Enhance Base/Out Ratio) and sixteen other measures, and then toward the end of the article they'll write that Michael Bourn had a 163 LIPV despite his -43 QXTR and his pathetic .721 M2D2, and you're thinking "What in the hell is he talking about?" That's careless progress.

My attitude has always been "Be sure that you take the public with you,"--or, at least, do what you can to take the public with you. Don't start speaking your own language that only you and two other people understand; take the time and make the effort to give anybody who wants to understand what you're saying a fair shot at it. I'm sure that sometimes I have failed to do that, but that's what I believe in. Take the time to type out "Batting Average on Balls in Play", rather than BABIP, and "Wins Above Replacement", rather than WAR. It just takes a few seconds.

Three things James often says that I love:

1) I need to look at that again.
2) I was wrong.
3) I don't know.

Especially the last one. I don't see those words too often around here. From at least a couple of you, I'm not sure if you're even familiar with those concepts.

Let me preemptively provide you with your comeback:

"I sometimes think Clemenza makes sense. I need to look at that again."

Every major saberoriented writer I'm aware of says there are tons of things we don't know.

But this was Neyer the other day disagreeing w/ James about the lingo thing:

http://espn.go.com/blog/sweetspot/post/_/id/5960/lets-not-call-the-whole-babip-thing-off

There's this really cool thing called Google. There was a time, not so long ago, when if you were reading a book or a magazine and you came across some obscure technical term and couldn't figure out what it meant, you were basically stuck.

You're not stuck anymore.

...Bill knows what BABiP means. The great majority of Bill's readers -- all of whom are interested enough to spend actual money to read his missives on the Internet -- know what BABiP means. BABiP's been around for 10 years, and is well-established among the people who pay to read Bill James. In that particular space, spelling out Batting Average on Balls in Play would be almost as pointless as spelling out Earned Run Average.

I think Bill just doesn't like BABiP because he didn't grow up with it. When I worked for him, he didn't like it when I wrote that a player slugged .472 (or whatever)....

You wanna put me on TV, before the great unwashed masses? Then I'll spell out anything you like. Until then, I'm going to reserve my right to use acronyms and abbreviations that I believe you can handle.

"There's this really cool thing called Google."

No kidding, plus sabermetric writers are writing for a sabermetric audience. That audience (the one that presumably pays their bills, not the "great unwashed masses", not that I have any idea how these dudes make money) isn't going to want to read every acronym spelled out.

Every major saberoriented writer I'm aware of says there are tons of things we don't know.

I'm sure that's true. Like I said, I just don't hear it very often around here. I guess I could do a search, but if Morbius ever came on and said "You know, you're right about that--what was I thinking?", I think I'd have several heart attacks on the spot.

As far as the acronyms go, I realize James is being disingenuous; he used to use things like RC/27 habitually. But I agree with his central point that jargon is odious. I deal with it every day in my job. No one can dream up ridiculous acronyms like educational resource teachers. Our big focus right now is "TLCP": Teaching Learning Critical Pathways. Which'll be in place till they dream up a new one.

clemenza, I try not to post anything here I'm not certain of. I save that for the politcs thread.

also, jargon SAVES TIME.

Okay--I'm headed over to the politics thread to sample the humble, truth-seeking, "You know, I was wrong about that" version of Morbius!

no, that's not what I meant...

I'm with Morbs on this one!

I'm excited about the Negro Leagues stats announcement, this will hopefully lead to a lot of new research and debate.

However, incorporating their stats into MLB doesn't really address the real problem. We don't know whether Josh Gibson would have had a higher BA than Ty Cobb had they played in an integrated MLB. A lot of the weaker pitchers that Cobb faced wouldn't have been in the majors if the league had been integrated. I don't know what Cobb's BA would have been, but surely a lot lower than .366. And you could make the same argument about Gibson. Putting Gibson and Cobb on the same leaderboard gives the illusion that they were contemporaries whose stats can be compared (which is possible post-1947, or post-1961 expansion), but it's not so straightforward.

(actually, if we're going to fantasy book an integrated MLB pre-1947, then a racist like Cobb might have refused to join it and instead found a white-only minor league to play in, in which case we might not know his name today)

One of the most extreme disparities between the two WARs that I've seen in a while: Tyler Anderson leads all MLB pitchers in bWAR with 4.0, while in fWAR he's 57th with 0.9.

I almost always use bWAR as a matter of habit--I just find it easy to look up and navigate--and the two are usually more or less in sync. But fWAR is clearly the better one here. I don't know how they arrive at that 4.0. Good ERA (2.63) for a lousy team, and under 7 hits per 9 innings, but that's about it. Anderson leads the AL in walks given up, he strikes out under six batters per 9 innings (I didn't think anybody was that low anymore), and his FIP is 4.77.

bWAR is supposed to be less accurate for pitchers iirc? But yeah I’ve seen the difference between his ERA and his FIP (the latter basically in line with his career numbers) as partial explanation; he’s been getting bailed out by position players.

Must be the case, although that generally happens with good teams. Is fielding the one thing the Angels can do?

No. I think it’s just insane luck.

I don't think he's less than a one win pitcher. His game logs look very good, only once below five innings. He accrues about the same amount of WPA over his starts as Crochet, Tarik Skubal, Tanner Houck, and Cole Ragan, more than Erik Fedde, Logan Gilbert, and Jack Flaherty.

Seth Lugo is the AL starter that really outdistances everybody there.

His change-up is really, really good.

Siri, post an overlay of his 4SB/CU deliveries.

(^^^referring to Anderson there)

Here's an interesting one - Dylan Cease has a total WPA this year of 0.348, compared with Anderson's 1.717. Fangraphs has Cease at 2.4 WAR, Anderson at 0.7.

Speaking of Flaherty, the maybe not so lousy team put up a five spot on him tonight, Phil. Seemed to take advantage of just about every mistake.

there is something awry with aaron judge’s defensive stats. fangraphs has him at -4.9 runs and soto at +1.7, and whether you believe soto is having a better overall defensive year than judge (even accounting for the fact that judge has played about a quarter of his time at DH), judge is not closer in defensive value to a full time DH than he is to soto. some of this may just be small-sample noise, but I wonder if there is some sort of systematic error here considering they play next to each other: there could be something about how they are positioned that is confusing the zone ratings. I don’t really have a deep enough knowledge of defensive stats these days to be confident, but it reminds me a bit of NBA adjusted plus-minus stats and how I was pretty convinced nikola jokic was being shorted and guys like kentavious caldwell-pope were being rewarded for playing so many minutes together

That definitely sounds wrong; Judge has been consistently good for his career defensively, Soto consistently mediocre.

Padres closer (and all star) Robert Suarez - bWAR is 1.8, fWAR is 0.6?

Padres closer (and all star) Robert Suarez - bWAR is 1.8, fWAR is 0.6?

― timellison, Monday, July 15, 2024 12:33 PM (nineteen minutes ago)

i haven't followed the changes in WAR calculations between these two sites that closely recently so i don't want to speak *too* confidently, but generally/historically speaking fangraphs pitcher WAR is more FIP/predictive stats based whereas BR's is based more on outcomes. in suarez's case his ERA is 1.67 but his xERA is 3.07 and his xFIP is 3.87 so i'd imagine that's where the discrepancy is coming from

yeah barring rare exceptions the bbref/fg differences in pitcher WAR almost always even out after couple more years worth of data

Lost track of Nate Silver for a long while, but the election led me to the Substack he does now. In a recent post on Kamala Harris--pros and cons--this, one of the pros, made me laugh:

12. It’s not that Biden can’t win — maybe the polls have been way off all along or there will be an alien invasion or something. But any election that Biden could win, any reasonable Democrat should be able win at this point. He is probably a below-replacement-level candidate.

https://www.natesilver.net/p/the-case-for-and-against-kamala-harris

Joe Biden, -0.3 WAR.

9/9, 9. First time I've used Gene Garber, I think, famous for ending Rose's 44-game hit streak in 1978. Vintage Rose: he complained after the game because Garber wasn't man enough to throw him a fastball.

https://www.youtube.com/watch?v=Y1sm1rOUfTg

https://i.postimg.cc/TP6qp28r/grid.jpg

"In Rose's final at bat in that game, in the ninth inning, he struck out against Garber, but Rose was critical for Garber not throwing more fastballs. 'He was pitching like it was the seventh game of the World Series,' said the Reds' 'Charlie Hustle.' Garber was not amused. 'I had an idea,' said the Atlanta reliever, 'that Rose was hitting like it was the final game of the World Series.'"

Meanwhile, Rose starts the AB by trying to lay down a bunt.

Wrong thread...I don't know how that happened, but I can't be bothered re-posting.

there is something awry with aaron judge’s defensive stats. fangraphs has him at -4.9 runs and soto at +1.7, and whether you believe soto is having a better overall defensive year than judge (even accounting for the fact that judge has played about a quarter of his time at DH), judge is not closer in defensive value to a full time DH than he is to soto. some of this may just be small-sample noise, but I wonder if there is some sort of systematic error here considering they play next to each other: there could be something about how they are positioned that is confusing the zone ratings. I don’t really have a deep enough knowledge of defensive stats these days to be confident, but it reminds me a bit of NBA adjusted plus-minus stats and how I was pretty convinced nikola jokic was being shorted and guys like kentavious caldwell-pope were being rewarded for playing so many minutes together

― brony james (k3vin k.), Saturday, July 13, 2024 7:02 AM (nine months ago) bookmarkflaglink

it’s early, but gonna go ahead and claim victory on this one

From Posnanski today: who is/was the Ultimate Replacement-Level Player?

But the right answer is Mario Guerrero. Guerrero played for four teams in his eight-year career from 1973 to 1980 and was always filling in for somebody or other. With the Red Sox, he was often playing for an aging Luis Aparicio. For the 1975 Cardinals, he played a lot in place of the less-tattooed Mike Tyson. He went to California, where he would spell Jerry Remy and Dave Chalk. Then he went to Oakland, where, for some reason, they played him every day before having him back up Rob Piccolo.

Through it all, he had a career WAR of exactly 0.0. And he was so remarkably consistent about it. He never had a 1.0 win season. And he never had a minus-1.0 win season. He just kept going, 70 OPS+, slightly below-average defense, slightly below-average baserunning, available when needed. There really is something beautiful about it.

Vaguely remember him.

Do defensive metrics differentiate between making a play easily and making a play but it was a lot closer than it should have been? I could ask the internet but asking ILB is usually more interesting.

Isn’t that the range factor?

Range Factor was James's primative defensive metric when there really wasn't much of anything beyond fielding pct. in the '80s. It was to measure how many balls a player actually got to, as opposed to having a good fielding pct. for many fewer plays. Range Factor = putouts + assists / innings played. It wasn't related to easy-play/difficult-play, though.

(Like a metric that elevated Ozzie Smith and downgraded Derek Jeter.)

the key data here is the plays made that, if they hadn’t been made, would still have been scored hits. lost to time.