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

Monday, August 20, 2018

How are the Diamondbacks 200 runs better than the Phillies on defense? | The BIS Blog

I received a tweet recently that asked if I could demonstrate the difference between the team with the most Defensive Runs Saved (the Diamondbacks with 120) and the team with the fewest (the Phillies with -102).

It seemed a reasonable question to wonder – can the difference between the two be more than 200 runs? We wouldn’t question the best and worst teams being that far apart in terms of their offense or their pitching. But for defense, we don’t have as good a feel for the value of certain statistics.

It’s important to note that when we give a team’s DRS total, that represents the sum of their players’ DRS. It’s not a value solely assigned to the team, but rather the accumulation of many values.

With that in mind, let’s go through a few notable positions around the diamond.

Jim Furtado Posted: August 20, 2018 at 02:08 PM | 34 comment(s) Login to Bookmark
  Tags: diamondbacks, drs, fielding, phillies

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. snapper (history's 42nd greatest monster) Posted: August 20, 2018 at 02:23 PM (#5730281)
Good article. Gets into a lot of component stats underneath DRS, e.g.

The Diamondbacks have turned 74 percent of ground balls hit between second base and two-thirds of the way to third base into outs in an unshifted defense (mostly by Ahmed). Within that same area, the Phillies have converted 68 percent.

For the Diamondbacks, the differential is nearly 30 extra plays made.


I would love to see more breakdowns like this.
   2. Man o' Schwar Posted: August 20, 2018 at 02:45 PM (#5730294)
The Phillies are kind of hamstrung by the signing of Carlos Santana. They've got at least one guy in the outfield (Hoskins) who should be a first baseman. His problems aren't even touched on in the article. Ditto Nick Williams, who seems to have defensive issues as well. But with Santana taking 1B every day for the near future (through 2021), it's hard to see at least the OF parts getting better.

By BBREF, their position player (non-P) wins above average is worst in the NL by a large margin. They're at -12.1, with COL second worst at -7.6. Their pitching is keeping them in the race.

(FWIW, BBREF sees the difference between Arizona and Philly as "only" 184 runs on defense.)
   3. perros Posted: August 20, 2018 at 03:18 PM (#5730316)
The Phillies are notably terrible by the naked eye, too.
   4. McCoy Posted: August 20, 2018 at 03:21 PM (#5730317)
Hard to imagine the Phillies would have a better record than the Red Sox if they had the diamondbacks defense
   5. Dog on the sidewalk Posted: August 20, 2018 at 03:23 PM (#5730319)
I want Hoskins and Santana in the OF corners and Bour at first. Keep things interesting.
   6. jmurph Posted: August 20, 2018 at 03:27 PM (#5730324)
The Phillies are kind of hamstrung by the signing of Carlos Santana.

People (here and elsewhere) were weirdly convinced he was going to remain really good in perpetuity this past offseason (I remember many giving the Red Sox grief for not going after him). Despite the fact that he was about to turn 32.
   7. Dog on the sidewalk Posted: August 20, 2018 at 03:36 PM (#5730331)
He hasn't been that bad. And while he's never been a BABIP master, his .222 is the worst of his career by a healthy chunk. I wouldn't write him off yet.
   8. snapper (history's 42nd greatest monster) Posted: August 20, 2018 at 03:38 PM (#5730335)

People (here and elsewhere) were weirdly convinced he was going to remain really good in perpetuity this past offseason (I remember many giving the Red Sox grief for not going after him). Despite the fact that he was about to turn 32.


His "decline" is pretty much all BABIP. His xwOBA by Statcast is .365 vs. an actual of .329, and a .350 last year. He's likely fine.

The issue is the Phillies already had a 1B.
   9. jmurph Posted: August 20, 2018 at 03:44 PM (#5730340)
His "decline" is pretty much all BABIP. His xwOBA by Statcast is .365 vs. an actual of .329, and a .350 last year. He's likely fine.

His ISO is down again, too. I'm not saying he's terrible, but I'd certainly rather not have him than have him. And that will be more true next year at 33, and more true the year after.
   10. Dog on the sidewalk Posted: August 20, 2018 at 03:48 PM (#5730344)
It's not the most efficient use of resources, but the Phillies are understandable happy to live with Hoskins in LF if it means having Santana's bat in the lineup, given their alternatives.

The signing seemed weird, given that they weren't expected to content, but it's good for them at the moment.

Edit: who would you prefer be getting his PAs, jmurph?
   11. jmurph Posted: August 20, 2018 at 04:00 PM (#5730350)
Edit: who would you prefer be getting his PAs, jmurph?

Oh I'm not a Phillies fan, I didn't really mean to distract the conversation. Just a couple times this year when it's occurred to me I've checked in on him and seen him not playing well, which stands out to me because I thought at the time (in the offseason) expectations for him were a little too high.
   12. asinwreck Posted: August 20, 2018 at 04:03 PM (#5730353)
The Phillies bolstered their defense at midseason by acquiring Asdrubal Cabrera to play shortstop. In 2018.
   13. Never Give an Inge (Dave) Posted: August 20, 2018 at 04:34 PM (#5730373)
Hard to imagine the Phillies would have a better record than the Red Sox if they had the diamondbacks defense

The Phils allow 4.24 runs per game. If you believe the numbers, with an average defense, they would be allowing 3.42 R/G. And with the D-backs defense they'd be allowing 2.45 R/G? Is that right? I'll actually gross that up to 2.66 R/G so that we aren't giving any credit for the superior fielding of the D-backs pitchers. All of this in a stadium that is playing as a pretty extreme hitters' park.

There have only been three teams in MLB history that have allowed fewer than 2.7 R/G (the Cubs of 1906, 1907 and 1909). The 1906 Cubs did that in a league that was scoring 3.56 R/G, compared to 4.38 in the NL this year.

You have to go back to 1972 to even find a team that had allowed fewer than 3.0 (Baltimore and Oakland both did it that year). There were also two such teams in 1968 (the Cards and Tigers) and before that you'd have to go back to 1919.

It's possible that Philly really has the best pitching staff in the league, and that if you put them in front of the D-backs' fielders they would be historically great at run prevention -- as good as the best teams of all time despite a much higher run scoring environment...while I want to believe these metrics, or at least to have a set of fielding statistics that I can believe, I am a bit skeptical here. Either that, or I'm just bad at math -- someone please correct me if any of the above is wrong.
   14. What did Billy Ripken have against ElRoy Face? Posted: August 20, 2018 at 04:36 PM (#5730376)
I want Hoskins and Santana in the OF corners and Bour at first. Keep things interesting.
Hell, move Santana back to third or catcher.
   15. Bote Man the walk-off king Posted: August 20, 2018 at 05:09 PM (#5730394)
The Phillies are notably terrible by the naked eye, too.

Even an eye cloaked in rose color would find it difficult to disagree. There were at least 3 obvious boo-boos last night by the Phillies fielders.
   16. drdr Posted: August 20, 2018 at 05:47 PM (#5730414)
Problem with extreme numbers: remember that each game has 27 outs (more with extra innings, 24 if losing on the road after the top of the ninth). If ARI fielders were transported to PHI, they would have less opportunities, so less of the positive score. Also, it's questionable if some of the results can be the same with different team. ARI seems to have mastered positioning, in PHI it may be different. Passed balls depend on the catcher, but also on the pitcher. Wild pitchers can be uncatchable even for the best catchers. Not every ball in dirt is the same.

As for PHI, they signed 1B when they had young 1B and hole in OF. For not too different money, they could have had two year younger OF (bad defensive OF, but still better defensively than 1B masquerading as OF) with better recent hitting stats.
   17. Walt Davis Posted: August 20, 2018 at 06:10 PM (#5730426)
For the Diamondbacks, the differential is nearly 30 extra plays made.

That's good stuff. But this is also just (give or take) 24 runs. To get to 200, you've got to have this sort of gap at every position on average.

The BABIP gap is just 17 points. That's about 50 hits in the Phillies' BIP count. The RoE gap is 22.** Those gaps can't explain more than 60-70 runs. There's a long way to go to get to 200.

By trad stats, the DBacks do have nearly 150 more assists but they also have about 140 more chances. Rtot puts the gap at just 76 runs which seems more reasonable.

And neither team is much of an outlier by DRS (at b-r). The DBacks are at +91, closely followed by the Brewers at +84; the Phils are at -93, closely followed by the Mets at -86.

By BABIP, the Cubs and Braves are nearly as good as the Brewers/DBacks; by TZ, the Cubs are the best by 22 runs over the Brewers; by DRS, the Braves/Cubs are 40-50 runs behind the DBacks/Brewers; by UZR the Brewers are on top with +28, 57 runs better than the Phils/Mets. Based on Inside Edge (OF play only, at fangraphs), the Phils OF is only about 22 plays worse than the DBacks OF. (baseballsavant doesn't have team statcast stats near as I can tell so we can't comp those -- still haven't done anything with IF defense anyway. Not that I've done a thorough look but IE and statcast usually seem pretty close.)

So it's still a mess. The stats can't quite agree on ordering -- DBacks #1 by DRS (at b-r), Brewers #1 by UZR, Cubs #1 by TZ. Only DRS puts the top/bottom gap at anything close to 200. There is reasonable agreement that those 3 teams are the best and Mets/Phils are the worst so it's not random crazytown. But none of the other stats put the gap at anything remotely close to the 200 run gap by DRS.

** It seems a lot of the Phils' errors are giving extra bases, given the error gap is 46. An extra base is worth about 1/3 of a run if I remember right so that's still probably not much more than 10 runs.
   18. Sunday silence Posted: August 20, 2018 at 07:16 PM (#5730443)
DRS seems to be more liberal in awarding runs for defensive efforts than the other systems but i think it reflects reality better.

We were arguing just a few weeks ago about how many runs a good fielder could save your team and I suggest that a very good central defender at say CF could save you 30 runs vs an average OF'er if he's able to excel in range, arm and whatever else it takes to hold base runners (quickness, release, etc.). To see the Phillies under performing the avg LF by 20 or the DBacks outperforming the avg CF by 15 I think supports my position.

We were also debating at that time, just how many fly balls to the OF are really more than routine and I think Walter and CFB suggested something like 55. the article talks about what amounts to 170 balls/year in the gaps between CF and lf and CF and rf. there's also balls hit down the line and also ones directl7 to an OF albeit a good deal behind him.

Still how do they get 38 runs more on shifts and 26 more at P.
   19. Never Give an Inge (Dave) Posted: August 20, 2018 at 07:39 PM (#5730451)
If ARI fielders were transported to PHI, they would have less opportunities, so less of the positive score.

But not by that much. Philly strikes out more hitters, but the two teams rank #2 and #3 in the league in Ks. If Arizona had Philly's K rate, they'd have 26 more Ks, or 26 fewer outs to be made in the field. That is only about 1% of their outs in the field, probably not enough to move the needle very much.

ARI seems to have mastered positioning, in PHI it may be different.

Yes, but positioning is part of defense -- unless you think that the Phillies' opponents, pitching staff and park don't afford a similar opportunity to benefit from good positioning as Arizona's. I'm sure it's not identical, but I would be skeptical if the opportunity was that materially different between two teams in the same league with similar component stats.

A few points worth noting, although I'm not sure what they mean in this analysis: (1) Arizona is the most extreme GB staff in the league at 0.91 GB/FB, while Philly is basically in a tie for third at 0.84. (2) Philly has the most at-bats with the platoon advantage in the league with 54%, whereas Arizona is tied for 6th at 47%. (3) Philly has actually faced more LHB this season (46% vs. 40%), which would seem to present more opportunities for shifting.

Wild pitchers can be uncatchable even for the best catchers. Not every ball in dirt is the same.

This one is a bigger factor -- Arizona's catchers throw a lot more balls in the dirt than Philly's, based on the article. And about half the difference in catchers' defensive value betweeen the two teams looks to be attributable to pitch framing, which is not completely independent of the pitchers. So it's possible that the defensive stats are over-attributing value to the catchers.

EDIT: Still, the catchers are only 30 runs worth of difference here. Even if you exclude them entirely, you'd still be saying that Philly with Arizona's defense would have allowed 2.90 R/G. The only team better than that since 1919 would have been the 1972 Orioles.
   20. there isn't anything to do in buffalo but 57i66135 Posted: August 20, 2018 at 08:08 PM (#5730462)
who cares about defense. the phillies' everyday lineup now has 7 guys with 17+ HRs, plus cesar hernandez (starting 2B, 10) and justin bour (19, bench). the lineup also has hitters with the 4th, 8th and 12th most walks in MLB (plus justin bour off the bench, 11th).

the team isn't good, but they've gotten this far without any unsustainably high performances. noone's having a breakout year, noone's achieving their 95th percentile outcome. this team has not clicked on all cylinders at any point this season, but they're still on pace to win 89 games.

in fact, they kind of embody the idea that if you finish 3 games over .500 every month, you'll win 90 games. they were +4 in april, +4 in may, (-1 in june), +4 in july, +1 in august (so far).
   21. Eric J can SABER all he wants to Posted: August 20, 2018 at 08:48 PM (#5730482)
His ISO is down again, too.

FWIW, the ISO decline is all doubles (he's had 1 HR per 29 PA each of the last two years), which means it's at least partly related to his low BABIP.
   22. Walt Davis Posted: August 21, 2018 at 12:29 AM (#5730539)
I think Walter and CFB suggested something like 55. the article talks about what amounts to 170 balls/year in the gaps between CF and lf and CF and rf. there's also balls hit down the line and also ones directl7 to an OF albeit a good deal behind him.

Most of those balls to the gap and down the line are impossible to catch by anybody. The number of possible plays that aren't routine ends up being something like 55 (or whatever).

Using Inside Edge (at fangraphs) -- the best option we have for team data AFAIK -- AZ vs Philly, number of catches/opps per probability category ... my earlier numbers might be off, the IE table I looked at originally was not restricted to the OF, may bad. This is ... differences are pro-rated

0% 0/196; 0/161
1-10%: 2/41; 2/25 (Philly 1 catch better)
10-40%: 10/28; 3/17 (Philly 3 catches worse)
40-60%: 10/17; 11/27 (Philly 5 catches worse)
60-90%: 32/40; 32/47 (Phil 6 catches worse)
90-100%: 661/663; 653/657 (Phil 2 catches worse)

So that's 15 fewer plays by the Phils, maybe 20 runs across the entire OF (that's assuming they average about 2 bases). This puts AZ 51 OF balls ahead (including the impossible) or just 16 if we limit it to the possible. For whatever reason, b-r credits each OF with a couple more plays made than IE but the gap in actual putouts is just 14 there too. And, according to UZR, the Phils' OF gains back 4 runs on arm.

Note UZR is at least closer to DRS here -- they give the DBacks a 33-run advantage in OF range. They're credited with about 15 more in-zone plays and 41 more OOZ plays. But here we get a massive disconnect between IE (and statcast) vs UZR (presumably even worse to DRS). According to UZR, the DBacks OF have made 236 plays OOZ ... according to IE, the DBacks have made only 54 plays with a probability under 90% (and most of the ones in the 90-100% category are obviously clustered close to 100% given how few are missed by anybody). Just how small are these UZR zones? How much run credit are they giving for "OOZ"?

If you look at BABIP by hit trajectory ... it's about 56 more hits by GB, 4-5 fewer hits by LD, 0 extra hits by FB and maybe 1 hit by bunt. That's about 45 runs by GB and maybe 6 runs by LD. Add 22 extra RoE.

A 200 run difference on defense has to be the result of something like 200-250 extra hits/RoE on BIP ... or 150-200 of those and like 100 extra bases taken. They simply don't exist.
   23. PreservedFish Posted: August 21, 2018 at 08:55 AM (#5730573)
Those Inside Edge numbers never seem to make sense to me.
   24. Rally Posted: August 21, 2018 at 09:14 AM (#5730584)
A 200 run difference on defense has to be the result of something like 200-250 extra hits/RoE on BIP ... or 150-200 of those and like 100 extra bases taken. They simply don't exist.


Phillies have allowed a .302 BABIP. Take 100 hits off (what they should have with an average defense) and you are down to .268. This would be the best in baseball, currently Oakland leads at .275.

Take 200-250 hits away (give them the AZ defense) and we are down between .219-.236. The lowest team BABIP against I can find on BBref are from 1968 - .244 for the Indians, .245 for the Orioles.

It's hard to believe this.
   25. Der-K: at 10% emotional investment Posted: August 21, 2018 at 09:32 AM (#5730589)
I'm also reminded of the recent Sam Miller piece on shifting, which suggests that we ought to not treat the more visible impact of shifts (more balls converted by players into outs; more extra base hits on ball into suddenly larger gaps) as independent from pitch outcome (pitchers seem to throw more balls when the shift is on).
   26. PreservedFish Posted: August 21, 2018 at 09:32 AM (#5730590)
Am I reading this correctly that runs saved on shifts are not in any way apportioned to individual positions? Seems like a double-counting opportunity.

Also, I tried double-checking the shift data on Fangraphs, and found that Fangraphs appears to not record the outcome of a PA during the shift as "shifted" if it ends in a HR, BB or K. That seems stupid.
   27. Der-K: at 10% emotional investment Posted: August 21, 2018 at 09:48 AM (#5730599)
Threw me as well
   28. PreservedFish Posted: August 21, 2018 at 09:55 AM (#5730606)
Did anyone notice that the Phillies' two top starters, Nola and Arrieta, are actually outpitching their FIP?
   29. PreservedFish Posted: August 21, 2018 at 10:12 AM (#5730614)
More on the double-counting. Fangraphs says that each team has shifted about 1,200 times in PAs that end in a BIP. Both teams have allowed about 3,170 balls in play total. So more than 33% of BIP happen with the shift on (is that credible?). If you're throwing out all of the individual defensive performances during those ABs, and considering that they've only played 75% of the season, each individual fielder only has about 50% of the opportunity to generate defensive statistics in comparison to the season-long norms we established just half a decade ago before teams went batty with the shifts. Like, if you think of Ozzie Smith or Brooks Robinson as a +20 fielder, in this 75% of a season, playing shifts 33% of the time, he would only be +10.

That means that if I tell you Nick Ahmed is 32 runs better than Scott Kingery, what I'm really telling you is that according to the historical standards you probably have in your brain, Nick Ahmed is in fact 64 runs better than Scott Kingery. Shenanigans.
   30. snapper (history's 42nd greatest monster) Posted: August 21, 2018 at 10:13 AM (#5730616)
Did anyone notice that the Phillies' two top starters, Nola and Arrieta, are actually outpitching their FIP?

Nola's go an 81% strand rate, so that one makes sense.

But, the low BABIPs don't scream "terribel D!"
   31. Rally Posted: August 21, 2018 at 10:23 AM (#5730622)
Pivetta's making up for them, striking out 11 batters per 9 and still having more hits than innings. A .350 Babip.
   32. snapper (history's 42nd greatest monster) Posted: August 21, 2018 at 10:29 AM (#5730627)
Pivetta's making up for them, striking out 11 batters per 9 and still having more hits than innings. A .350 Babip.

Yes, he's getting hosed. 3.43 FIP, 4.66 ERA.

Of course one of the criticism of applying defensive stats to pitchers has always been that pitchers don't get the same defensive support on the same team. It's very possible that Pivetta has been disproportionately victimized by poor D.
   33. Never Give an Inge (Dave) Posted: August 21, 2018 at 12:01 PM (#5730703)


Also, I tried double-checking the shift data on Fangraphs, and found that Fangraphs appears to not record the outcome of a PA during the shift as "shifted" if it ends in a HR, BB or K. That seems stupid.


Yep, that was one of the key points of the Sam Miller piece -- that while shifts result in fewer hits on BIP, they also appear to result in more BBs as pitchers seem less comfortable pitching in front of the shift, and this negates much of the positive impact.
   34. Infinite Yost (Voxter) Posted: August 21, 2018 at 02:28 PM (#5730835)
How are the Diamondbacks 200 runs better than the Phillies on defense? | The BIS Blog


They aren't. Defensive metrics still don't work.

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
HowardMegdal
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

Gonfalon CubsThe Final Push
(134 - 1:55pm, Sep 20)
Last: What did Billy Ripken have against ElRoy Face?

Hall of MeritMost Meritorious Player: 1947 Discussion
(10 - 1:54pm, Sep 20)
Last: MrC.

NewsblogOTP 2018 September 17: How Brett Kavanaugh explains his baseball ticket debt
(1379 - 1:50pm, Sep 20)
Last: Gonfalon Bubble

NewsblogWho needs Justice when you have OMNICHATTER! for Sept. 20, 2018
(1 - 1:50pm, Sep 20)
Last: Los Angeles El Hombre of Anaheim

NewsblogDodgers bash their way to MLB history with record-tying seven 20-homer hitters
(9 - 1:36pm, Sep 20)
Last: Blastin

NewsblogSnell wins 20th, ties Rays record in 4-0 victory over Texas
(28 - 1:36pm, Sep 20)
Last: snapper (history's 42nd greatest monster)

NewsblogIs Red Sox' slump temporary, or did they come back to earth?
(6 - 1:25pm, Sep 20)
Last: TomH

NewsblogOrioles match club record with 107th loss, 5-0 to Blue Jays before tiny home crowd
(54 - 1:24pm, Sep 20)
Last: PreservedFish

NewsblogOT - Catch-All Pop Culture Extravaganza (September 2018)
(355 - 1:17pm, Sep 20)
Last: jmurph

NewsblogOT - 2018 NBA Thread (Pre-Season Edition)
(424 - 1:01pm, Sep 20)
Last: TFTIO is Lounging from the flat one

NewsblogOT: Soccer Thread (2018-19 season begins!)
(805 - 12:58pm, Sep 20)
Last: AuntBea calls himself Sky Panther

NewsblogMinor league baseball players exploring union as they continue to battle low wages
(13 - 11:34am, Sep 20)
Last: Pat Rapper's Delight (as quoted on MLB Network)

Sox Therapy11 Days Later
(74 - 11:19am, Sep 20)
Last: ERROR---Jolly Old St. Nick

NewsblogRed Sox division title banner falls off truck; its return may cost them - The Boston Globe
(31 - 11:12am, Sep 20)
Last: Never Give an Inge (Dave)

Sox Therapy6,036 Days
(12 - 10:50am, Sep 20)
Last: villageidiom

Page rendered in 0.3806 seconds
46 querie(s) executed