rolling sabermetrics and statistics thread

where are u looking, on the politics thread?

also, why get excited about a middle manager?

damn.. some club just flat out bought the CHONE numbers and the guy who put em together. i kind of slightly liked those more than MARCEL.

http://www.baseballprojection.com/

damn, his projections were the most accurate the past few years of any of the public systems, free or subscription

ZIPS is the best one now i guess

I SO SAD

http://www.anditisliz.com/lusciousliz/wp-content/uploads/2008/08/figgins.jpg

if you'd like to attend a sabermetrics seminar at Harvard in May:

http://sabr.org/latest/sabermetrics-scouting-and-science-baseball-may-21-22

off praiseball, this is hilarious: http://en.wikipedia.org/wiki/Carson_Cistulli

particularly the wes anderson casting call photo

this seems cool:

http://www.fangraphs.com/blogs/index.php/shutdowns-meltdowns/

A reader comment from High Heat Stats that I think is quite good:

"Vida Blue's 1979 season is one of a thousand examples why W/L record is essentially meaningless." I think a better way to put it is "W/L record is often deceptive when used as a method of ranking individual pitcher performance". I looked at both the Wins and WAR of the 200 pitchers with the most career Wins over the 1962-2011 period, a group that takes you from Greg Maddux at 355 Wins to Dave Burba with 115 Wins. I ran CORREL in Excel and found the correlation between Wins and WAR among these 200 guys to be .85. (1.0 is a perfect correlation, -1 is a perfect negative correlation and 0 is random relationship with no correlation at all). .85 suggests a meaningful but not perfect correlation. In other words, Wins usually, but not always, tells a lot, but not everything, about a pitcher's WAR-type value. When I look at W-L records today, what automatically runs through my head is that I am looking at a set of numbers that combines individual performance with some team quality and some luck. That's not meaningless, just limited.

He's not saying anything startling, but I like that he took the trouble to actually investigate.

Neat chart from "Hey Bill": Power/Speed/Walks rating ("the harmonic mean" of three stats, an extension of his old Power/Speed rating). The Top 10 all-time, prompted by a letter about Abreu:

First Last HR SB BB P-S-W

Barry Bonds 762 514 2558 822.2
Rickey Henderson 297 1406 2190 661.5
Willie Mays 660 338 1463 581.7
Joe Morgan 268 689 1865 524.6
Alex Rodriguez 629 305 1166 523.9
Hank Aaron 755 240 1402 483.5
Bobby Bonds 332 461 914 478.1
Gary Sheffield 509 253 1475 454.9
Craig Biggio 291 414 1160 446.8
Bobby Abreu 284 393 1419 443.1

Thought I'd spaced that better.

you have to put everything in between CODE signs (see formatting help, below) if you use spaces to organize things into columns.

Por clemenza:

http://railsrx.files.wordpress.com/2011/08/james.png

Thanks! Such a great photo--I went to the site and tracked down the source, and it's from an early article about him in Sports Illustrated in May of '81. I've got four or five milk crates of SIs with baseball covers dating back to the early '70s, so I'll have to check if I've got that issue.

http://sabr.org/latest/unofficial-list-baseball-records-set-and-tied-2011

lol Uggla

Jo-Jo Reyes, Most consecutive winless starts, 28

this was actually news in Toronto. i went to the game he actually won to break the streak and it was mental!

ppl were rooting for a ND?

ha ha, no. the fans were shockingly pumped up, cheering for a W.

this Toronto, after all. the city of tempered expectations.

Bill James is contributing to Grantland! His first article: the 100 greatest pitchers duels of 2011:

http://www.grantland.com/story/_/id/7480753/bill-james-100-best-pitchers-duels-2011

Anyway, a great pitchers' duel implies that there is something at stake beyond fifth place, although you don't want to place too much emphasis on that criterion, or you wind up warbling on about Jack Morris in 1991, long after anybody cares.

I attended #82, which was the SABR convention's visit to Dodger Stadium.

Pretty sure the best pitching duel I ever saw live was this:

http://www.baseball-reference.com/boxes/TOR/TOR198506060.shtml

Key had a no-hitter through 8, until Tom Brookens led off the ninth with a single; he lasted 10 innings, gave up two hits, and left with the score 0-0. Petry also pitched 10 shutout innings, giving up 6 hits. The Jays won in the bottom of the 12th on a Buck Martinez home run.

It meets all of James's conditions: low-scoring, starters pitched well and deep into the game, pitchers of stature (Key was only 24, but in the midst of his first great year; Petry was having his fourth straight good-to-very-good season), and the game meant something (the best young team in baseball vs. the defending champs).

I've mentioned mine before -- also '85, Gooden v Tudor in September division race, scoreless til Orosco gives one up in 10th:

http://www.baseball-reference.com/boxes/NYN/NYN198509110.shtml

That's about as good as it gets, regular season; Gooden's year was historic, and Tudor would have won the Cy Young 19 out of 20 times. Combined WAR: 19.2.

Took a quick look through '65 and '66 to see if I could find a 1-0 Koufax-Marichal game, but couldn't. Such games undoubtedly occur more frequently in the post-season. Johnson beat Maddux 2-0 to lead off the 2001 NLCS, although Maddux only pitched seven.

where you can celebrate SABR Day tomorrow:

http://sabr.org/sabrday

Neyer linked to this article on Fangraphs that charts reliever usage over the past 30 years:

http://www.fangraphs.com/blogs/index.php/are-relievers-benefiting-from-pitching-less/

It's fairly short and to the point, basically it says that avg. number of batters faced per relief appearance has plummeted, but overall reliever effectiveness has stayed the same.

The data is persuasive, but you have to assume that 1) batters today wouldn't make the necessary adjustments if they faced the same relievers more, and 2) pitchers can be reconditioned to throw 100+ IP in relief (i.e pitchers throw a lot harder than they used to).

I was reading that last nIght - I think they had tacked on some analysis on performance in high-leverage situation. And what they found was slightly better performance in "clutch" scenarios and worse when it really didn't matter vs 30 years ago.

The Fangraphs iphone app is $1 today only. Pretty nice.

It's not too early for wild projections of meaningless metrics, is it? Good, I thought so.

If Halladay were to win the Cy unanimously this year, he'd move past Palmer, Seaver, Pedro, and Carlton into fourth place for Cy Young share on Baseball Reference. Only Clemens, Johnson, and Maddux would remain ahead of him.

That's pretty cool, actually. But he'll still move into fourth place eventually if he has a few more good years.

I think Halladay has led his league in CG's more times than any pitcher in MLB history.

Fourth seems like a good bet, and even Maddux is within reach (he's 1.42 shares behind--one solid win and a couple of seconds or thirds). Clemens and Johnson are out of range.

I think you're close to right about the CG's. I checked a handful of pitchers, and the only one I found who led the league more often was the first one I thought of: nine times for Spahn. I think Halladay's at seven. (Found out that Robin Roberts once completed 28 consecutive starts...science-fiction.)

Barring sudden collapse, I guess Halladay goes into the HOF in a Phillies cap. Too bad.

I swear to you, I didn't know there was "Cy Young share on Baseball Reference." I'm more familiar with Gingrich's delegate count.

I've always liked the Cy Young/MVP-share metric. If you think the two awards are compromised enough to be meaningless, then you'll think the metric is too. But the Top 10 in MVP share does a pretty decent job at ranking the greatest position players post-Ruth (when there often wasn't an MVP): Bonds, Musial, Pujols, Williams, Mays, Mantle, Aaron, Gehrig, DiMaggio, A-Rod. The Cy list is a little more erratic--Glavine sits ninth, and if Halladay does move into the top four, I'm not sure how many people would argue he's been one of the four greatest pitchers since 1955. (Maybe I'm being premature there.)

I can't really get worked up over the cap/HOF thing (nobody gets officially inducted as a member of any one team), but I can't see him in the HOF in anything other than a Jays cap. Gary Carter would be a good comp: started his career with the Expos, had most of his best years there, moved onto a higher profile gig and won a WS (which Halladay hasn't done yet, but he might), was inducted as an Expo.

so some friends and i accidentally, somewhat replicated RC.
except that instead of dividing the offensive numbers by PAs we used outs.

so the formula was this:
TB + BB + HBP + SB + SF + SH
––––––––––––––––––––––––
PA - (H + BB + HBP) + CS + GIDP

we liked it so much went and created a database for it!
http://www.dancochran.com/bapo

but then i discovered RC.

and yes - we called it BAPO. (bases advanced per out)

i think it's hands down the nerdiest thing i have ever done.

My legal firm and I are representing one William James; we'd like to arrange a meeting with you.

ha ha hahaa

i think one of the more interesting things we noticed with bapo is that even a good hitter is going to get out more than he advances the bases - which means sacrificing might not be such a terrible waste after all.

Of the dozen or so of my "Ask Bill"s that James has answered, the one I sent in on Saturday was the first time he responded with some impatience:

Bill -- A leap into the dreaded land of intangibles...Most everyone agrees now that the irreplaceability of a closer was vastly overstated for a number of years; teams move guys in and out of that role continually. Watching the Jays' struggles this year, though--five blown saves in 21 games--has hurled me back to the couple of years in the early '80s before they got Tom Henke, and the memory of how demoralizing a series of blown saves can be to an otherwise good team (to a fan, anyway). Question: even though the difference between a great position player and an adequate one is undoubtedly quantitatively larger (in terms of WAR, or Win Shares, or whatever) than the equivalent difference between a great/adequate closer, might there be an intangible psychological importance to the great closer that can't be measured?
Asked by: Phil Dellio

Answered: 4/29/2012
Implying that it isn't demoralizing to lose a game in other ways? If your offense is poor and you lose games because you can't score runs, this doesn't demoralize the pitchers? If your starting pitching is bad and you're playing from behind every day, this doesn't demoralize the rest of the team?

Of course it is POSSIBLE that there things we can't measure, in the same way it is possible that the world around us is full of creatures or beings of some nature of which we are unaware because none of our five senses will pick them up. The question is, why should one believe in any one of these things?

True enough about different kinds of losses being demoralizing in their own way--but I do think a blown save, where you were leading late into the game, has to be worse. (On the other hand, if you wanted evidence to the contrary, the 2001 World Series would do well.)

I feel like that's the kind of thing that's more demoralizing to the fans than the players...

I think you're probably right about that. (Not sure I picked the best counter-example in the 2001 Series; having Randy Johnson and Curt Schilling in reserve overcomes a lot of demoralization.)

I feel like that's the kind of thing that's more demoralizing to the fans than the players...

Yeah, I'm always amazed when your team loses in the bottom of the ninth in some brutal way, you'll be sitting at home stunned and the players are just calmly jogging off the field. Obviously they're upset about it but they don't get demoralized in the way that fans do. Which is why fans believe in stuff like "momentum", whereas players always claim they never do.

Cub broadcaster Len Kasper on the increased embrace of sabermetrics in the media/clubhouse environment:

http://www.baseballprospectus.com/article.php?articleid=18680

http://mlb.sbnation.com/2012/11/16/3651520/wins-above-replacement-sabermetrics-war-stats

Reminds me of James when he wrote that assists needed to be replaced by something less polite (he suggested "baserunner kills").

If you had been reading baseball annuals in 1984....

http://www.baseballprospectus.com/article.php?articleid=18993

interesting analysis of how the size of the strike zone changes according to the count. it aligns with experience watching games - when the count is 3-0 it seems like anything near the plate is suddenly a strike:

http://www.fangraphs.com/blogs/wp-content/uploads/2012/12/szBatR_02count.png