rolling sabermetrics and statistics thread

Sorry, my phone autocorrected you to the rocket

Or wait, it’s adding for walks, and then subtracting 250

That's supposed to be 2 x SO - 1/2 x BB; strikeouts looks like 50. An ILX'or I'm Facebook friends with figured out how to make the numbers work out...Mike Trout comes in at 43-something, just below Mantle. The guy just strikes out too damn often, and I'm surprised the Angels don't cut him loose.

give me ed delahanty over any of these bozos

Weren’t you afraid you missed some players?

I also found pages of the final totals (not sure what to call it...CI, the Clemenza Index) for an alphabetized list of anybody who had a good career. I was working from the '74 edition of the Mac, not the first '69 edition; Aaron was still active, so I'm not sure what my cut-off date was (more likely, I had a minimum AB requirement).

Ed definitely makes the all-drinking team, from what I remember.

in 1895, delahanty scored 149 runs in 116 games. he hit a healthy .404 but came in second in the batting race behind jesse burkett at .409

it's a really cool list - awesome that you kept it!

despite my misreading of the equation, it is also very very neatly printed.

Would have generated a few posts on the ILB police blotter thread:

"Delahanty died when he was swept over Niagara Falls in early July 1903. He was apparently kicked off a train by the train's conductor for being drunk and disorderly. The conductor said Delahanty was brandishing a straight razor and threatening passengers after he consumed five whiskies. After being kicked off the train, Delahanty started his way across the International Railway Bridge connecting Buffalo, New York with Fort Erie (near Niagara Falls) and fell or jumped off the bridge (some accounts say Ed was yelling about death that night). Whether "Big Ed" died from his plunge over the Falls or drowned on the way to the Falls is uncertain. His body was found at the bottom of Niagara Falls two weeks after his death."

don't ask why i felt compelled to make this, but might as well share. this is the top 5 wRC+ for each year of the 1950's. i wanted to see which player seasons were way ahead of their peers, relatively. williams in '54 and williams and mantle in '57 both stand out

https://i.imgur.com/5tSykHj.png

also, i have to share this as well because it is crazy how the charts in Excel wannabe clone Numbers are hard to customize and offers all the wrong options...unless you go into 3D graph mode, in which case everything is possible

https://i.imgur.com/Sr0KzWP.png

Surprised there's such a difference between Mantle's '56 and '57--his bWAR is the same for the two seasons. I know it's relative to the league, but still.

heh, the chart i made is confusing as all hell. it's so confusing even i can't figure out wtf is going on!

but basically, the top line represents the first place finisher in wRC+ for the year. it happens to be williams in both '54 and '57, which makes it look like the entire top line is williams. but in '55, for example, the first place wRC+ was mantle, at 179.

...i know, that's really confusing! i only did it that way because i wanted to see the discrepancy between the top 5 finishers each year, to look for outliers.

mantle's '56 was actually very close to '57, by wRC+. it was 202 in 1956 (twice as good as the average hitter!!), which was first place in the league, and 217 in 1957, second only to williams.

*fingers crossed behind back*

i will never make another ill-advised graph again

C'mon, the 3D one even casts a shadow--worth the price of admission alone.

not exactly sabermetrics, but: the details behind pitch classification

https://technology.mlblogs.com/mlb-pitch-classification-64a1e32ee079

this seems like a wild stab in the dark to me, but

We created a dashboard for NCAA WAR over the past four seasons. Calculate WAR based on school, year, name, and more: https://t.co/ZI6q74BApb

You can read an introduction to cWAR and what went into the dashboard here: https://t.co/PvFB5KGiKV pic.twitter.com/ftkYyfEsW1

— Driveline Baseball (@DrivelineBB) June 10, 2020

this is great

https://www.theringer.com/mlb/2020/7/17/21328094/analytics-60-game-season-war-stats

i'm with John Thorn

Dallas Keuchel is 6-2 w/ a 2.19 ERA. He has 1.7 bWAR & 1.5 fWAR.

Rick Porcello is 1-5 with a 6.06 ERA. He has 0.0 bWAR.& 1.5 fWAR.

— David Laurila (@DavidLaurilaQA) September 17, 2020

thing is, both bWAR and fWAR are useful, for pitchers, they just measure different things. bWAR lines up with what people think they're getting out of WAR, and is based on results. fWAR for pitchers is based on FIP, so it's more of a measure of how they "should" be doing. which is confusing, agreed on that. wish that both b-r and fangraphs would offer both versions - give me both a results-based WAR and a peripheral based WAR, and just keep them separate imo

i like having them both because i can look at porcello's line and immediately know that he's been unlucky in terms of his results (likely he has been giving up a ton of home runs, or his LOB% is way lower than it normally would be because of bad sequencing luck).

actually porcello's given up very few homers and his babip's a nightmare, which is no doubt why fangraphs likes him

i'm willing to grant that he's been unlucky with balls in play and LOB (the mets defense probably doesn't help). but i suspect his 5.7% HR/FB this season (12% for career) is luck too. meanwhile his WHIP is over 1.5, he doesn't have great strikeout numbers for this day and age, and he goes less than five innings per start.

and keuchel's been *insanely* lucky in every facet! so i guess i feel like they should both have lower fWAR

(this is not to argue with your larger point, which is otm)

it also seems likely that i have an outdated idea in my mind of what one should expect from a (good/average/replacement-level) major league starter

Starters? pfffft. Give me a listicle of the top 500 openers in the game (globally, including little league) and I'll be projecting HOF eligibility by sundown.

there's always been this...WAR (sorry)... between BR and fangraphs with the WAR stats. honestly, i think fangraphs needs to change, because their WAR philosophy inconsistent between the pitchers and hitters. for hitters it seems more (though not entirely) results-based, like BR. but for pitchers it's more peripheral. it's just kind of weird, especially when you're trying to compare value between pitchers and hitters.

Do you really want to get me started about how flawed their defensive metrics* are and how that number is quietly folded into each site's "offensive" WAR and just published as QED gospel?

*just about every other season there is some radical makeover of defensive stats causing some huge ripple in the lists that appeal to whatever you call the devotees of "MOJO-magazine equivalents of baseball"-type content.

oh yeah, i feel you on that for sure. i've never understood why it's not just clearly bifurcated as offense WAR and defense WAR.

it is! At least on bbref it is, there's oWAR and dWAR (which doesn't actually add up to WAR because it includes the position adjustment bonus in each)

oh d'doh, i just mean on fangraphs. i still hate the baseball-ref style/format, and rarely visit it for that reason alone. you know how there are chrome extensions? and particularly for some websites (like discogs) that reorganize the data on the fly? i wish someone would do that for either baseball-ref or fangraphs.

Good Posnanski piece that definitely spotlights a major flaw in bWAR.

https://joeposnanski.substack.com/p/burnes-baby-burnes?utm_campaign=post&utm_medium=email&utm_source=copy

(That's his "share this piece" link, so should work.)

the fog of glove again

great article, thanks for taking the time to share!

however, i ran across a similar anecdote the other day and removed it from my memories, and i am afraid i am about t delete this again for similar reasons:

*By the way, do you know who IS playing in front of the 1970s Baltimore Orioles defense? Adam Wainwright. DRS does show the Cardinals being a terrific defensive team, but even that underrates how good they’ve been behind Wainwright; the Cards are 22 Outs Above Average when Waino is on the mound. That is more than double anyone else in baseball.p

i don't know what the previous series of deleted words was about, but one thing i know is that adam wainwright is on his way to be being the best pitcher in baseball again, and it's definitely related to luck!

i prefer fWAR to bWAR a lot for pitching, though i find that bWAR defenders tend to be way more defensive about it than fWAR defenders. people really just don't like the concept of FIP. but the thing tango said about halving the value of DRS should apply to hitters too... it results in massive outlier seasons because DRS is so springy when UZR is more conservative. it really seems to impact MVP conversations too even though no one really trusts defensive stats. you get these wild 10+ WAR hitter seasons because someone put up a DRS of 8 million (and a UZR of 15)

should also point out that fWAR isn't just FIP, there are also adjustments for IFFBs and catcher framing (though that's controversial for some people too)

that said i do appreciate that bWAR updates its position adjustment values season by season, fangraphs just keeps pretending that playing shortstop is as valuable now as it was 20 years ago despite teams now feeling comfortable parking their trucks there

yeah it is a fluid situation for sure

trying to figure out why there's such a huge disparity between the 2015 phillies' UZR (-15) and DRS (-98) - one thing i didn't know was that DRS also calculates pitcher defense while UZR doesn't even attempt to. a lot of the difference between the two is just DRS giving a lower rating to their defenders across the board, but there's also 22!!! runs lost by the pitching staff. 5 pitchers have a DRS of 1 and the rest have a combined DRS of -27. there are 3 pitchers worth -4 runs and 2 of them are relievers. refuse to believe that ken giles, who had 1 error all year, somehow cost his team 4 defensive runs in 70 innings of work

i generally get that, but then i remember jon lester playing his position a few years ago, when he forgot how to throw first base for a while

those would generally register as errors though wouldn't they? so what's the knock against giles, he wasn't rangy enough? hard to figure out what the individual components are on bbref pages. lester did have a pretty bad season (-7 DRS) though he never had more than 3 errors in a season - i think that's mostly because he allowed a lot of SBs that year.

thing is, i don't think pitcher bWAR includes individual defense, but i think individual pitcher DRS contributes to the overall defensive adjustment. so hypothetically a pitcher could have a really bad DRS which wouldn't lower their WAR at all, but might actually raise their WAR by contributing to a bad team defense adjustment. but i'm not sure of that.

lester allowed a lot of SBs that year because he was yips-unable to throw to first and keep runners honest. so, not an error, but a definite problem on defense.

that *was* kind of an outlier though, and the ken giles example seems unreasonable

Have statskeepers ever tried to account for walkoff singles that would have been doubles or triples if they'd played out in a non-walkoff situation? Like last night, Austin Riley's walkoff would have been a double in any other inning.

no idea, but I’d guess that these are so rare that the difference would be negligible

Sounds like irresistable bait to statheads to me.

well there are def stats out there that rate hitters on hit probability based on predicted batted ball outcomes measured by contact speed off the bat, launch angle etc that would prob see such a hit as something more than a single (would not even know that it was recorded as a single) and thus reward austin riley for it based on that

yeah that’s a good point, his xwOBA doesn’t take a hit because he’s only credited with a single

xp interesting, thx

As EW pod

-ask

From Robert Christgau's monthly reader-questions column (it has a terrible name...):

Professional baseball is rapidly changing. Are you familiar with sabermetrics baseball and its implications? Or is this just too nerdy a thing to ask? — KBW, South Korea

I was reading sabermetrics pioneer Bill James as early as the ‘70s, I think--long ago, anyway. Thought all of his analysis was fascinating and a lot of it worth incorporating into the game. It really changed pitching, although not as much as the revised strength training stratagems that have generated so many near-100 fast balls. But if I remember correctly, even then I didn’t like how down he was on stolen bases--they’re too much fun (I loved how much the Yankees stole late in the past season). And when I watch the game with its radical shifts these days I sometimes get nostalgic for the old days, as well as wishing more players would settle for singles by exploiting shifts. In particular I still prefer human umpires calling balls and strikes even though what was clearly a bad call on a held-up swing prematurely ended the Dodgers-Giants championship game.

xxxxp i thought official scorers were supposed to use their judgment on walk-off hits . . . and i was wrong. this seems unnecessarily complicated, especially with the ground-rule double possibility:

2019 OBR rule 9.06(f) Subject to the provisions of Rule 9.06(g), when a batter ends a game with a safe hit that drives in as many runs as are necessary to put his team in the lead, the Official Scorer shall credit such batter with only as many bases on his hit as are advanced by the runner who scores the winning run, and then only if the batter runs out his hit for as many bases as are advanced by the runner who scores the winning run.

Rule 9.06(f) Comment: The Official Scorer shall apply this rule even when the batter is theoretically entitled to more bases because of being awarded an “automatic” extra-base hit under various provisions of Rules 5.05 and 5.06(b)(4)…

When interpreted literally, does WAR really work with an extreme closer like Josh Hader?

Hader has pitched 28.2 dominant innings and is 1.3 WAR on Baseball Reference. He's saved 26 games out of 27 save opportunities. If you actually did swap him for a replacement-level closer, wouldn't that guy likely blow at least three or four of those games?

Are you suggesting blowing a save should earn you -1 WAR? There's a whole 8 innings of player performance that happened beforehand

I don't even know what I'm suggesting...it just doesn't seem to jibe. The way a poor inning can be directly translated into a loss at that stage in the game, and the way all of Hader's innings fall into that category, throws me. But I guess he's no different than any closer since the advent of save-only usage.