Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Baseball Newsstand > Baseball Primer Newsblog > Discussion
Baseball Primer Newsblog
— The Best News Links from the Baseball Newsstand

Tuesday, September 03, 2019

One advanced metric says he’s baseball’s top pitcher. Mike Minor isn’t buying it

NEW YORK — Mike Minor is in a class of his own.

At least according to “Baseball-Reference WAR,” which ranks Minor as the top pitcher in all of major-league baseball this season (7.0).

Just don’t tell Minor that he’s No. 1.

Because even after pitching 7 1/3 innings of shutout ball against the mighty New York Yankees on Monday, the Texas Rangers veteran southpaw wasn’t buying it.

To be fair to Minor, pitcher WAR can get contested in ways not as common with that for position players…..

 

QLE Posted: September 03, 2019 at 12:22 AM | 28 comment(s) Login to Bookmark
  Tags: mike minor, war

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. bfan Posted: September 03, 2019 at 02:59 PM (#5876475)
Didn't Aaron Nola have a crazy high WAR last year (10.5) that was only defensible on the basis that the Phillies fielded one way when he was pitching, and another way for all of the other Philly pitchers?
   2. What did Billy Ripken have against ElRoy Face? Posted: September 03, 2019 at 03:34 PM (#5876489)
“You look at the guys over there in Houston — Gerrit Cole and Justin Verlander — and I feel like those guys dominate every game. And I feel like they look better than me.”
We're not selling jeans here, Mike.
   3. Ziggy is done with Dominican discotheques Posted: September 03, 2019 at 03:42 PM (#5876493)
the Texas Rangers veteran southpaw wasn’t buying it


I'm not either Mike, I'm not either.

Really, what gives here? Verlander has him beat on both IP and ERA+. You'd think it would be UER, but Minor has three of those and Verlander (in more innings, remember) has only two. Is the Rangers' D is terrible but not error prone?
   4. RJ in TO Posted: September 03, 2019 at 03:55 PM (#5876498)
Really, what gives here? Verlander has him beat on both IP and ERA+. You'd think it would be UER, but Minor has three of those and Verlander (in more innings, remember) has only two. Is the Rangers' D is terrible but not error prone?


B-R has the Rangers' D costing minor 0.23 runs per nine, while Verlander is benefiting by 0.22 runs per nine. Just taking a quick look at their defenses, I can absolutely believe the Rangers' D would be that bad, and the Astros that good. The Rangers also have a multi year park factor of 111 (106 for this year), while the Astros have a multi year and single year park factor for pitchers of 98.

Minor has a WAR of 7.5, while Verlander is at 6.7. Given all the adjustments, I can certainly believe Minor has a better WAR for this year, while also believing I'd still rather have Verlander for the remainder of this year as he's got a track record to support pitching at this level, while Minor seems to be way above his head.

Verlander has a chance this year to allow more homers than walks, which would be a nifty and fairly rare trick for a starting pitcher.
   5. ERROR---Jolly Old St. Nick Posted: September 03, 2019 at 04:19 PM (#5876505)
So who else in the past year has shut down the Yankees like Minor did yesterday? And he did it after being smacked in the back with a line drive by the first batter of the game.

As to whether he's the best pitcher in the game, maybe he is and probably he isn't, but in hindsight I sure would've given up Clint Frazier to get him at the end of July.
   6. Itchy Row Posted: September 03, 2019 at 04:59 PM (#5876513)
Verlander has a chance this year to allow more homers than walks, which would be a nifty and fairly rare trick for a starting pitcher.
It looks like ERA qualifiers have done that 20 times, with 17 of those since 1998. Mike Leake (35 HR and 24 BB) will probably do it too. Miles Mikolas (23 HR and 25 BB) has a chance.
   7. Kiko Sakata Posted: September 03, 2019 at 05:35 PM (#5876519)
Didn't Aaron Nola have a crazy high WAR last year (10.5) that was only defensible on the basis that the Phillies fielded one way when he was pitching, and another way for all of the other Philly pitchers?


Actually, kind of the reverse. The fielding adjustment in BB-Ref assumes that a team fields the same for all of the team's pitchers. The problem, in Nola's case was that he had an exceptionally low BABIP (.254) but the Phillies had a bad defense in general, so BB-Ref said that Nola's BABIP should have been even lower than that but for the bad defense. Which is unlikely - much more likely that the Phillies defense, for whatever reason, was actually pretty good (or, at least, not bad) when Nola was pitching.

Incidentally, with respect to Minor this Twitter thread does a really nice job of working through the math of how Minor's bWAR gets so high.
   8. Walt Davis Posted: September 03, 2019 at 05:49 PM (#5876522)
Yu Darvish in his last 14 starts (86 IP): 20 HR, 8 BB

Didn't Aaron Nola have a crazy high WAR last year (10.5) that was only defensible on the basis that the Phillies fielded one way when he was pitching, and another way for all of the other Philly pitchers?

Actually that's the primary (only?) argument in the attack which, while probably true, is a pretty weak argument. He had a very low BABIP in front of a poor defense -- bWAR gives him credit for that. The "theory" (don't recall any facts presented) is that it couldn't possibly be that low (or anything close to it) in front of a lousy defense so they must have turned into gold glovers when he was on the mound and so he doesn't deserve credit.

On Verlander vs Minor (post 4), we don't really have to wonder how it works, the numbers are (nearly) all right there.

Verlander gives up 0.61 fewer runs per 9, a big edge.
But Minor's opponenents are slightly tougher at 5.06 vs 4.93 so he gets back 0.13 runs there.**

As mentioned, Minor's defense is (usually) much worse than Verlander's, the 0.45 run gap so now we're basically even. Basically, with the Astros' +0.22 defense, Verlander's average opponent should score 4.70 against an average pitcher and he's held them to 2.66 so just over 2 runs better; Minor's opponents should score a whopping 5.29 so his 3.27 is just over 2 runs better too.

They get the same boost as SP (0.23 R/9 added to their opp's "average" scoring) and then you get the weighted park factors of nearly 110 for Minor vs 99 for Verlander which is a huge difference.

This is all summarized in RA9avg which is how much their mix of opponents should score per 9 against an average pitcher in front of their defenses and in their mix of parks. This comes out to a whopping 6.06 for Minor vs 4.87 for Verlander. So Verlander's RA9 "should" be 1.19 runs better than Minor's but it's only 0.61.

Minor has 182 innings so let's just call that 20 9-inning stints and across those common innings, at 0.58 runs better per 9, that's 11.6 runs or something along the lines of 1.1 to 1.2 wins. Verlander then claws a few runs back with his extra 12 innings to bring the RAA gap back to 9.

Then we get some slighty mysteries -- how RAA get converted to WAA and how replacement value is calculated for pitchers but 12 innings isn't going to make a lot of difference in replacement value.

The defenses matter here but aren't a huge factor. Put them both at 0 and Minor loses about 4.5 RAA and Verlander gains about 4.5 RAA which makes them dead even. The park factors are the key to the swing. j

I wonder how much of the difference in quality of opponents comes down to Verlander gets to face the Rangers (87 OPS+, 4.95 R/G in a hitter's park) while Minor has to face the Astros (119 OPS+, 5.55 R/G in a neutral park).

Anyway, unless you're going to throw out park factors, Minor's results have been (give or take) as impressive as Verlander's ... whether they're repeatable or not.

** That doesn't adjust for the actual lineup faced as far as I know, just the weighted team average runs scored.
   9. Eric J can SABER all he wants to Posted: September 03, 2019 at 06:06 PM (#5876528)
As mentioned, Minor's defense is (usually) much worse than Verlander's, the 0.45 run gap so now we're basically even.

The Rangers' defense has apparently been better behind Minor than they usually are; his BABIP is .287.

On the other hand, Verlander's BABIP allowed this year is .203. TWO OH THREE. So... the defensive adjustment is still probably a good thing to throw in there.
   10. Walt Davis Posted: September 03, 2019 at 06:09 PM (#5876530)
And FWIW, the one with the crazy BABIP is Verlander (203) not Minor (287). Minor's BABIP is 290 for his career and it was better than 287 the last two seasons. Verlander's career is 282 and it was in the 270s the last two years (about the same as Minor's). Verlander is out-pitching his FIP by about 0.85 runs this year, Minor by 0.77. If anything, "luck" appears to be on Verlander's side.

On ERA+ ... we get into additive vs. multiplicative. The "lgavg" ERA gap is about 0.6 (4.53 vs 5.12) while the ERA gap is just 0.56. In additive terms, that's slightly to Minor's favor but in multiplicative (or divisive terms) it's to Verlander's (4.53/2.56 > 5.12/3.12). WAR is essentially additive -- 10 runs better is (give or take) 1 win better whether that's the difference between 0 RAA and 10 RAA or 40 RAA and 50 RAA. But Verlander gives up 56.5% as many earned runs as park-adjusted average while Minor gives up 60.9%. (Also pretty sure that ERA+ uses the generic park factors while WAR uses a weighted park factor based on the parks they actually pitched in this year.) Note also they look much closer by ERA- (the percentages I gave) than by ERA+ (the inverse which is a 13 point gap).
   11. cardsfanboy Posted: September 03, 2019 at 06:16 PM (#5876532)
Note also they look much closer by ERA- (the percentages I gave) than by ERA+ (the inverse which is a 13 point gap).


Which is why I like ERA- more, it's not as deceptive as ERA+...both tell the same story, one just does a better job at doing it (in my opinion)
   12. Walt Davis Posted: September 03, 2019 at 06:28 PM (#5876535)
Going back to Nola ... even if you ding him back to average defense last year, he only loses 15 RAA which brings him down to 9 WAR. His 254 BABIP is still very good -- not as crazy good as Verlander's 203 of course. His BABIP in front of a supposedly average defense this year is 287 ... give him a 287 BABIP last year and he loses probably about another 15 RAA bringing him down to 7.5. At this point we're basically assuming that (a) he actually played in front of an average defense and giving him no credit for a below-average BABIP and we're still way ahead of fangraphs' 5.4 fWAR. But sure, the first adjustment puts him behind deGrom (assuming we don't have to make any adjustmens to deGrom) and the second adjustment would put him behind Scherzer and Freeland in bWAR.
   13. the Hugh Jorgan returns Posted: September 03, 2019 at 06:52 PM (#5876546)
one with the crazy BABIP is Verlander (203)


But of course that doesn't include homers. 33 given up so far. Make half of those doubles and BABIP sits somewhere around normal.

Is BABIP down overall this year due to the outrageously high number of home runs?
   14. Eric J can SABER all he wants to Posted: September 03, 2019 at 07:03 PM (#5876550)
But of course that doesn't include homers. 33 given up so far. Make half of those doubles and BABIP sits somewhere around normal.

Is BABIP down overall this year due to the outrageously high number of home runs?


I don't see why it would be; BABIP isn't penalized for HR, it's entirely unaffected.

BABIP for 2019 so far is .299; for the four previous years it was .296, .300, .300, .299.
   15. Walt Davis Posted: September 03, 2019 at 07:04 PM (#5876551)
Overly fascinated by this but ...

Verlander's K-rate this year is basically the same (amazing) rate that he's had since the move to Houston; his (amazing) walk rate is actually up a smidgen and of course his HR rate is up. This leads to his FIP being 0.64 runs higher than last year's but his ERA is up just 0.04.

Minor's (more pedestrian) K-rate is up from last year and a bit above his career average but his walks are up too ... but his HR rate is down. The FIP gap between Verlander and Minor is slightly smaller than the ERA gap. By FIP+ they are both at 132 (so also the same by FIP-).

Aaron Nola was a very good pitcher pitching better than he ever has and he probably got lucky too. Mike Minor is a solid pitcher having the season of his career but probably not being overly lucky. Verlander is a great pitcher possibly having a slightly worse season than his recent (outstanding) seasons who is probably having massive good luck too. If you're gonna whack anybody this season for being too lucky, it's Verlander, not Minor. How fangraphs gets to a 1-WAR advantage for Verlander I don't know.

It looks like it's bullpen support -- which I don't get in this case. Minor has bequeathed 14 runners of whom 6 have scored -- that seems a reasonably high rate. Verlander has bequeathed jut 5 (it's nice to have an insanely low BABIP) and 2 have scored (a slightly lower rate than Minor). Minor's left 6 runners with 2 outs, 5 with 1 out and 3 with 0 outs. None of those runners were on third base. I don't know what the expected runs would be there ... but how in the world could that possibly add up to 1.1 wins above average? His bullpen can't possibly have saved him 10-11 runs.
   16. Walt Davis Posted: September 03, 2019 at 07:05 PM (#5876552)
Yes, I know I owe several cokes. Not refreshing often enough.
   17. . Posted: September 03, 2019 at 07:14 PM (#5876557)
The Rangers' defense has apparently been better behind Minor than they usually are; his BABIP is .287.

On the other hand, Verlander's BABIP allowed this year is .203. TWO OH THREE. So... the defensive adjustment is still probably a good thing to throw in there.


Begs the question entirely. Indeed, question begging is the entire problem with the idea of "credit." (Note here that I'm using the term not in the lazy sense of "raises a question" but instead the proper sense of ignoring a question under the false assumption or belief that it's been definitively answered.)
   18. Never Give an Inge (Dave) Posted: September 03, 2019 at 08:30 PM (#5876571)
The "theory" (don't recall any facts presented) is that it couldn't possibly be that low (or anything close to it) in front of a lousy defense so they must have turned into gold glovers when he was on the mound and so he doesn't deserve credit.

Come on, Walt. There were no facts presented by those who claimed the Phillies defense performed poorly for Nola, either. It's an assumption that's obviously false -- that the defense performed equally good / bad for every pitcher on the roster. In fact, there are plenty of reasons to think that a defense might perform better for certain pitchers on the roster than others -- fly ball vs. ground ball tendencies, handedness of the pitchers, etc.

Just based on a quick look at the numbers -- Nola and Arrieta, who are more ground ball pitchers, outperformed their FIP while the remaining starters, who are more fly ball pitchers, underperformed their FIP. So maybe there's something there.

Or you can continue to believe that Aaron Nola was really a 1.72 ERA pitcher who was just hurt by his defense last year.
   19. Walt Davis Posted: September 03, 2019 at 08:59 PM (#5876584)
I don't see why it would be; BABIP isn't penalized for HR, it's entirely unaffected.

Yes and no. The notion that Hugh is raising is that flyballs are going farther this year so that doubles are becoming HRs. There is of course one immediate problem with Hugh's formulation which is that some of those are flyouts becoming HRs. He has to this point only given up 6 fewer doubles this year than last so it would be a stretch to adjust his BABIP by 16 doubles.

But sure, to the extent that HRs "replace" doubles, we'd expect (on average, all else equal) a HR to lead to a drop in the numerator and the denominator of BABIP, reducing BABIP. To the extent that HRs "replace" flyouts, we'd expect (on average, all else equal) a drop in the denominator of BABIP, leading to a big jump. If we were to approach it that way though, we'd expect a mix of doubles/triples and flyouts and flyballs have low BABIP and double/triple rates are quite low ... but some of these new HRs might be more on "deep line drives" which would be more likely to result in hits if they stayed in the park. Anyway, without the jump in HR, we'd at best expect a mix of extra IP outs and extra IP hits that, on a leaguewide basis, would almost certainly balance out pretty well and have at most a tiny impact on BABIP.

But it's even trickier than that in our pretend universe because of course time in baseball is measured in outs. A HR does not necessarily "replace" a BIP. Unless it ends the game, it creates a new PA. That PA of course won't always result in a BIP but, on average, it will about 2/3 of the time. In a less HR-happy world, let's say the Astros give their opps an average of 39.5 PA per game with 38.5 of those being non-HR PAs that produce all 27 outs they eventually get. In the HR-happy world, they give their opps 40 PAs with 38.5 of those being non-HR PAs which are still the ones that produce all 27 outs. In that simple scenario, there's no reason to expect any change in the number of BIPs or the number of hits on BIPs. That's pretty close to the league-wide scenario.

Suppose we were in a world closer to 20/20 or ODI cricket where, say, each team was given a max of 40 PAs per game (timed by PAs, not outs). Then each HR would reduce the number of BIPs (or at least the number of opportunities to produce a BIP). But that just puts us back in the first scenario -- unless those HRs would have been BIPs that produced hits at a substantially higher or lower rate than a random BIP, they won't move the needle on BABIP.

That scenario is closer though to what it is like for a hitter. A guy who starts about 150 games a year will end up around 650 PA and there's very little he can do about that. When he hits a HR, he has almost no impact on his number of PAs in that game -- he delays the 27 outs for his team but it's random chance if that extra PA ends up in his lap. So an "extra" HR for a batter does decrease the number of opportunities to create a BIP across his season. But it's still the case that it will only impact BABIP if those PAs were much less/more likely to result in a BIP hit.

Obviously that extends to individual pitchers as well given they will generally face a pretty stable number of total batters from year to year (assuming the same health/durability/etc) ... but there's probably still some residual desire to have Verlander go 7 innings as opposed to 27 batters so a HR allowed possibly does increase his batters faced by a bit. That is, pitchers used to be mostly timed by outs as well; that is shifting towards timing by PA.

Then there's also the fact that if you turn 16 HRs into doubles (and BIPs) that only brings his BABIP up to 232 which is still a very low number. How low? In the 2000s, his current 203 is easily the lowest among qualified seasons -- Marco Estrada 2015 is 2nd at 217. A BABIP of 232 would tie him with Greinke 2015 for 6th.

Note Marco Estrada 2016 is 7th at 234. Those helped him to ERA+s of 131 and 122 -- which were preceded by a 87 and followed by a 90 and, released by the A's on Aug 20, nobody seems to have picked him up, this might be it -- 81 ERA+ over his last 353 innings and turning 36. But he missed half the season with injury so maybe he just needs to get healthy.
   20. Walt Davis Posted: September 04, 2019 at 04:50 AM (#5876647)
ome on, Walt. There were no facts presented by those who claimed the Phillies defense performed poorly for Nola, either.

The statement you're referencing was referring to Nola. But sure, as I made clear in other posts, it's likely Nola wasn't THAT good. But that's still not based on any evidence relating to how the Phils played behind him, it's based on assumption that BABIP couldn't be that low in front of a defense that bad. "The Phils had a terrible defense so it was probably terrible behind Nola too" is essentially the flip side of the assumption "that event was very unlikely so it must not have happened that way."

I agree the assumption that the defense has the same impact for each pitcher is clearly untrue. It's probably closer to the truth though than the notion that, when we disagree with it, a defense transforms from -.65 R/9 (worse actually) to one that's +.65 R/9. That's not impossible but it's bloody unlikely.

I agree that Rdef should at least be adjusted for G/F ratio but this isn't likely to make a huge difference. Pivetta (for example) had a G/F of 0.78 which is league average; that means he gives up a GB on 44% of his BIP. Nola's G/F was 1.02 which means he gives up GBs 50% of the time. Pivetta had 431 BIP by my calculation so that 6-7% difference is let's say 28 FBs replaced by 28 GBs. Now things get tricky because "FB" here is a mix of FB and LD. But Pivetta's LD rate was a smidgen below league-average. Looks like 11-12 of those extra FB might have been LDs -- that's probably 6-7 hits there; the other FBs will produce only another 2-3, let's call it 9 hits. But Nola should have given up about 6 hits on his 28 GBs. A 3-hit difference over 150 innings doesn't matter.

The more likely explanation is that Pivetta is just a crappy pitcher. His career BABIP is 331 (just 385 innings) and he had a 333 that year. If anything, the Phils' defense may have played really good for him too. Here are his ERAs and FIPs to date:

6.02 4.87
4.77 3.79
5.38 5.42

That 3.79 FIP is the number that stands out like a sore thumb.

Velasquez had a better history than Pivetta but much the same argument holds. He had a 322 BABIP against a 315 career average -- 2019 is his first season below league-average. For the record, Nola's BABIP was poor in 2016-17 too and he's only a bit better than average this year.

There are different types of "luck." When Brady Anderson went from 16 to 50 HRs, nobody expected him to repeat it but nobody could deny it happened. Now some of that might have been "luck luck" -- maybe he faced more RHP, maybe he faced worse pitchers, maybe he managed to just squeak 5-10 over the wall or just down the line. But there's really no doubting that for unknown reasons, Brady Anderson started hitting the ball farther and/or higher than he ever had before (or would after). And it wasn't "doubles turning into HRs" cuz he had 33 doubles the year before, 37 that year and 39 the year after. The Ks and BBs were about the same, the PAs were about the same, the number of hits was the same that year and the year after (he had many fewer the year before) -- the man pretty much literally turned 30-35 singles into HRs.

If Brady can turn into a 50-HR hitter then clearly it's possible that Nola turned into a guy who nailed the corners for a year. Or maybe he had tons of "luck luck". OK, sure, like all ballplayers having a good season, he had some of both. Or maybe a terrible fielding team became an average or even excellent fielding team.

The chances of that were near zero but there's no doubt it happened. The chances of Nola becoming a dominant pitcher nobody could hit for one season were near zero ... but there's no doubt that in fact he gave up very few hits ... we just don't believe it.

But fine, as I mentioned ... we'll give him full credit for the hits he didn't give up but we won't add the "bonus" that it was done in front of a generally terrible defense. As I mentioned earlier, that gets him down to 9 WAR. (It of course doesn't change his ERA or BABIP one iota.) There was still very little chance that a 4.5 WAR pitcher would have a 9-WAR season (but it's obviously higher than having a 10.5 WAR season).

Weird stuff happens in baseball. In Brady's case, it was easy to count the weirdness and nobody could deny the evidence. In Nola's case ... well, it's easy to count the weirdness of a nice BABIP and nobody can deny it happened ... but since we don't have easy, undeniable measures of the defense it's easy to decide that didn't happen ... which doesn't really change our conclusions dramatically.

In Nola starts, there were 7 ROE, a career high but of course it was a career high BF. But by rate, it was much worse than 2017 and 2019. I have no idea what the relationship between ROE and UER is but as a rookie he had 1 ROE and 0 UER; in 2017 he had 4 ROE and 1 UER; this year 4 ROE and 4 UER ... and in 2016, it was 6 ROE and 9 UER. So last year he didn't let the runners put on by his defense score -- maybe he "should" have given up an extra 3-4 UER ... or maybe he picked up his crappy defense.

FWIW, Arrieta led all Phils starters with 11 ROE ... but he got whacked for 17 UER. Zach Eflin was probably 2nd worst by rate, the Pivetta had 8 ROE in fewer many fewer PA but gave up just 4 UER. Then Velasquez had ROE at the same rate as Nola. By ROE, Nola had it a little better than most Phils' starters but not by a huge margin except Arrieta. (Arrieta is more a lesson in ERA vs RA than he is BABIP, FIP, defense, etc.)

deGrom also played on a team with lousy defense, he also had 7 ROE (giving up 7 UER), he also had a nice (but nowhere near as nice) BABIP. He had an "unbelievable" low HR/9, a crazy low ERA and, thanks to that HR rate, a crazy low FIP. The main difference between his and Nola's WARs are the park factors, not the defensive adjustment.

Scherzer also played in front of a below-average defense (but not nearly as bad as the other two) and had just 4 ROE and gave up 4 UER. His BABIP was 267, not much higher than Nola's. The gap between him and Nola was heavily dependent on that defensive difference.
   21. Never Give an Inge (Dave) Posted: September 04, 2019 at 07:57 AM (#5876648)
My view is that we have the data, we should be using it the same way we do for defensive statistics. Rather than making a broad assumption that defense plays equally for every pitcher on the team, actually use the available PBP/zone data to estimate how many runs the defense cost/saved each pitcher. It won’t be perfect but it will at least make use of all the data we have. Half-measures like what BB-Ref currently does are just knowingly putting false information out there and damaging the credibility of the advanced metrics.
   22. snapper (history's 42nd greatest monster) Posted: September 04, 2019 at 09:13 AM (#5876663)
My view is that we have the data, we should be using it the same way we do for defensive statistics. Rather than making a broad assumption that defense plays equally for every pitcher on the team, actually use the available PBP/zone data to estimate how many runs the defense cost/saved each pitcher.

Those metrics aren't nearly accurate enough to be used in such small sample sizes. That sort of false precision in statistics is worse than the blanket adjustments BRef uses.
   23. Never Give an Inge (Dave) Posted: September 04, 2019 at 09:55 AM (#5876676)

Those metrics aren't nearly accurate enough to be used in such small sample sizes. That sort of false precision in statistics is worse than the blanket adjustments BRef uses.

BB-Ref uses that same data for fielders with no regard for sample size. The sample sizes are roughly the same for starting pitchers and everyday position players (and actually much higher than for outfielders). Aaron Nola allowed 515 BIP last season.
Meanwhile, Matt Chapman led MLB in WAR fielding runs last season, with only 484 chances. Mookie Betts had 19 WAR fielding runs in only 278 chances. Heck, Aaron Judge has 15 WAR fielding runs this season in only 152 chances.
   24. snapper (history's 42nd greatest monster) Posted: September 04, 2019 at 10:17 AM (#5876688)
BB-Ref uses that same data for fielders with no regard for sample size. The sample sizes are roughly the same for starting pitchers and everyday position players (and actually much higher than for outfielders). Aaron Nola allowed 515 BIP last season.
Meanwhile, Matt Chapman led MLB in WAR fielding runs last season, with only 484 chances. Mookie Betts had 19 WAR fielding runs in only 278 chances. Heck, Aaron Judge has 15 WAR fielding runs this season in only 152 chances.


But now you're dividing each of those defenders into 7 or 8 slices. The sample size issue is not just total BIP, it's BIP in a particular zone. 500 BIP spread all over the field is less reliable than 250 BIP to one position. With a small number of balls in each zone, the odds that they average out to a normal degree of difficulty is low.
   25. Never Give an Inge (Dave) Posted: September 04, 2019 at 10:24 AM (#5876692)

Maybe (I don't see why you still wouldn't expect things to average out to normal over the full season, even if the likelihood of that is a bit less than if all the chances were to the same zone).

However, your point also seems like good argument for why the current BB-Ref assumption -- that all pitchers on a team get the same level of defensive support -- should be discarded.
   26. snapper (history's 42nd greatest monster) Posted: September 04, 2019 at 10:32 AM (#5876696)
However, your point also seems like good argument for why the current BB-Ref assumption -- that all pitchers on a team get the same level of defensive support -- should be discarded.

It's not a great assumption, but in the absence of better data, it's the safest assumption. We know which teams are good and bad defensively with a fair degree of certitude, we don't want to throw that out.

The issue is people need to stop thinking about WAR as a stat that tells them what happened, like ERA or HR. It's a value estimate, which still has fairly wide error bars, and likely always will.
   27. Never Give an Inge (Dave) Posted: September 04, 2019 at 10:58 AM (#5876705)

It's not a great assumption, but in the absence of better data, it's the safest assumption.

So you think it's more likely that Nola was really a 1.72 ERA pitcher who was hurt by his defense last season, rather than that he was really a 2.37-3.01 ERA pitcher who got a bit lucky?
   28. snapper (history's 42nd greatest monster) Posted: September 04, 2019 at 11:18 AM (#5876711)
So you think it's more likely that Nola was really a 1.72 ERA pitcher who was hurt by his defense last season, rather than that he was really a 2.37-3.01 ERA pitcher who got a bit lucky?

No. I think Nola was an outlier, and we were able to figure that out by looking into the numbers a bit.

But for the large majority of pitchers I think adjusting for D makes the results more accurate, i.e. the Minor-Verlander example we started with.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

News

All News | Prime News

Old-School Newsstand


BBTF Partner

Dynasty League Baseball

Support BBTF

donate

Thanks to
The Piehole of David Wells
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogKEN HARRELSON NAMED 2020 FORD C. FRICK AWARD WINNER
(37 - 9:42pm, Dec 11)
Last: Walt Davis

NewsblogMiller, Simmons elected to HOF on Modern Era ballot
(66 - 9:25pm, Dec 11)
Last: yest

NewsblogOT - NBA Thread, Start of the 2019-2020 Season
(1422 - 9:20pm, Dec 11)
Last: jmurph

NewsblogOT - Catch-All Pop Culture Extravaganza (December 2019)
(97 - 9:01pm, Dec 11)
Last: vortex of dissipation

NewsblogYankees, Gerrit Cole agree to record 9-year, $324M deal
(95 - 9:00pm, Dec 11)
Last: The Yankee Clapper

NewsblogPhillies to sign Didi Gregorius | MLB.com
(22 - 8:53pm, Dec 11)
Last: PreservedFish

NewsblogOT- Soccer Thread- October 2019
(921 - 8:20pm, Dec 11)
Last: Richard

NewsblogPrimer Dugout (and link of the day) 12-11-2019
(16 - 8:16pm, Dec 11)
Last: snapper (history's 42nd greatest monster)

NewsblogLou Whitaker snubbed from the Hall of Fame again
(150 - 7:59pm, Dec 11)
Last: blueshaker

Gonfalon CubsRegrets
(153 - 7:20pm, Dec 11)
Last: Zonk Has Two Faces, Both Laughing

NewsblogSome of baseball’s biggest stars are ready to change the game’s free-agency system – The Athletic
(18 - 7:04pm, Dec 11)
Last: Walt Davis

NewsblogRyan Thibs’ Hall of Fame Tracker
(398 - 6:01pm, Dec 11)
Last: Eric J can SABER all he wants to

NewsblogWhite Sox trade for Nomar Mazara | MLB.com
(32 - 5:52pm, Dec 11)
Last: bbmck

NewsblogHefner becomes Mets pitching coach, Meulens bench coach
(2 - 4:47pm, Dec 11)
Last: Lassus

NewsblogMLB to stop testing minor leaguers for marijuana as part of new drug agreement, report says
(29 - 3:00pm, Dec 11)
Last: Nasty Nate

Page rendered in 0.4473 seconds
46 querie(s) executed