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, May 28, 2013

Neyer: Run differentials tyrannize standings

Really, it’s fairly amazing how quickly the standings and the run differentials line up. You’d think, or at least some people probably think, that the relationship’s not really that strong. That the blowout wins and losses just don’t even out, or might even out eventually but certainly not before the calendar turns from May to June.

Except, look at the standings.

bobm Posted: May 28, 2013 at 11:47 PM | 33 comment(s) Login to Bookmark
  Tags: pythag

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. Robert in Manhattan Beach Posted: May 29, 2013 at 05:49 AM (#4454367)
You’d think, or at least some people probably think, that the relationship’s not really that strong.

Put me down as someone who thinks that the correlation between runs and wins is strong. I know, call me crazy.
   2. Ivan Grushenko of Hong Kong Posted: May 29, 2013 at 08:19 AM (#4454388)
How many times has Neyer made this point over the past 15 years? Who the hell is he arguing against?
   3. AROM Posted: May 29, 2013 at 09:19 AM (#4454408)
Oriole fans from last season? It works. Orioles are now within 1 game of their pythag, their run differential has normalized to their W-L record.
   4. RoyalsRetro (AG#1F) Posted: May 29, 2013 at 09:55 AM (#4454435)

How many times has Neyer made this point over the past 15 years? Who the hell is he arguing against?


I'm guessing Harold Reynolds and Orioles fans.
   5. SoSH U at work Posted: May 29, 2013 at 10:18 AM (#4454451)
Orioles are now within 1 game of their pythag, their run differential has normalized to their W-L record.


Looked at another way, their run differential has normalized to their W-L record. They're about halfway between their pythag and actual W-L record last year.

   6. Pat Rapper's Delight Posted: May 29, 2013 at 10:33 AM (#4454467)
I'm guessing Harold Reynolds and Orioles fans.

Mitch Williams has to be in this group too, doesn't he?
   7. Jose Is The Most Absurd Thing on the Site Posted: May 29, 2013 at 10:39 AM (#4454471)
Oriole fans are the recent example but fans of every team that overperforms pythag go through all manner of machinations to demonstrate why their team is different. Generally those teams all regress to their PR sooner than later. The 2007-2008 Diamondbacks were a really good example of a team that overperformed one year then put up basically the same year the next but won ten fewer games.
   8. SouthSideRyan Posted: May 29, 2013 at 11:17 AM (#4454505)
How many times has Neyer made this point over the past 15 years? Who the hell is he arguing against?


Seemingly everyone who responded to the most recent Cubs thread on the topic.
   9. Yeaarrgghhhh Posted: May 29, 2013 at 11:38 AM (#4454527)
That's not really fair to most Orioles fans. I think many fans argued last year that (a) wins are wins, and it's silly to say that they weren't *really* a 90 win team because they outperformed pythag, (b) the team was very different in August and September, and that team didn't outperform its pythag by much, and (c) any predictions for 2013 should be based more on the August/September team than the season as a whole. So far, the last point has proved to be basically correct.
   10. PerroX Posted: May 29, 2013 at 12:15 PM (#4454558)
Neyer waited a few days until the Giants fell out of first.
   11. PerroX Posted: May 29, 2013 at 12:17 PM (#4454563)
I'd be more interested in an explanation of the outliers rather than merely dismissing them as outliers.
   12. Matt Welch Posted: May 29, 2013 at 12:18 PM (#4454564)
Angels had a decent four-year stretch (2006-09) of strongly outperforming their pythag, leading to some unreasonable expectations among certain people. Cough.
   13. McCoy Posted: May 29, 2013 at 12:18 PM (#4454565)
Seemingly everyone who responded to the most recent Cubs thread on the topic.

I think it is far more likely for the Cubs to have their run differential catch up to their record than have their record catch up to their run differential.
   14. AROM Posted: May 29, 2013 at 12:57 PM (#4454602)
(c) any predictions for 2013 should be based more on the August/September team than the season as a whole. So far, the last point has proved to be basically correct.


Especially since that's when they added the closest thing to 1997-1999 A-Rod since A-Rod himself. As for the kid's defense, just compare the team DER up to the callup, and from there to the end of the season.
   15. Jose Is The Most Absurd Thing on the Site Posted: May 29, 2013 at 01:05 PM (#4454606)
any predictions for 2013 should be based more on the August/September team than the season as a whole. So far, the last point has proved to be basically correct.


Sample sizes can move this pretty quickly but;

August/September Actual - 37-18 (107-55 full season)
August/September Pyth - 35-20 (102-60)
2012 Actual - 93-69
2012 Pyth - 82-80
2013 So Far - 28-24 (87-75)
2013 Pyth - 27-25

So far they are certainly a lot closer to 2012 PR than to any kind of August/September combination and are very very very slightly closer to 2012 PR than to 2012 Actual.

It's 52 games. A 4 game winning streak, a 4 game losing streak and this all looks vastly different. What I think is more interesting is that they are 8-7 in one run games so far, a far cry from last year's 29-9.
   16. Steve Treder Posted: May 29, 2013 at 01:17 PM (#4454620)
I'd be more interested in an explanation of the outliers rather than merely dismissing them as outliers.

Agreed. Intuitively, I would think that one structural dynamic that would seem likely to impact variance from PR would be the relative strength of the starting staff vs. the bullpen. That is, a team with exceptionally strong starters but a very poor bullpen would likely win the blowouts (determined by big early innings) but lose the squeakers (determined in the late innings), and thus underperform against PR. And that team's mirror image, with weak starters and a great bullpen, would seem likely to do just the opposite, and overperform against PR.

But that's just a theory that I've never tested. Perhaps someone else has done so, and debunked it?
   17. BDC Posted: May 29, 2013 at 02:07 PM (#4454658)
It sounds like a great theory, Steve, but the distribution of the run support across those various situations might drown out the effect of starters being stronger or weaker than the bullpen. For instance:

The 2011 Rangers were unusual in that their starters (98 tOPS+ against) were stronger than their bullpen (105). They indeed did well in blowouts (34-21) and poorly in one-run games (19-24), and underperformed PR: by all of two games.

The 2007 Rangers had the weakest starters relative to the strongest bullpen of recent Texas teams (113 tOPS+ against vs. 81). They did poorly in blowouts (19-24) and well in one-runs (26-18), yet they underperformed PR by even more (four games). True, that was the year they beat Baltimore 30-3. But if a single game can have that much influence, then it'll be very hard to perceive the effect you note over even 162 games.

Brought to you by the Department of Hasty Conclusions from Two Data Points :)
   18. Ron J2 Posted: May 29, 2013 at 02:08 PM (#4454660)
#16 There were a couple of articles on beating pythags at Hardball Times. Quick summary. It's mostly record in close games. The only clearly identifiable factor is strength of bullpen, and record in one run games doesn't seem to carry forward from year to year.


   19. Steve Treder Posted: May 29, 2013 at 02:10 PM (#4454664)
Thanks!
   20. zenbitz Posted: May 29, 2013 at 02:39 PM (#4454693)
Isn't the relationship to bullpen strength rather weak as well? There are 30 teams, every year one or two will be off by an SD or so. Which is basically the observation.
   21. snapper (history's 42nd greatest monster) Posted: May 29, 2013 at 02:42 PM (#4454699)
Isn't the relationship to bullpen strength rather weak as well? There are 30 teams, every year one or two will be off by an SD or so. Which is basically the observation.

We should expect 1-2 to be +/- 2 SDs or more, and about 10 to be +/- 1 SD or more.
   22. madvillain Posted: May 29, 2013 at 02:55 PM (#4454730)
It's almost like nobody has ever studied this.

After July actual record becomes a better predictor going forward than Pythagorean record (I can't find this HBTs study after 5 minutes of googling but I know they did it in like 2010).

Keep fugging that Chicken Neyer.
   23. SouthSideRyan Posted: May 29, 2013 at 05:18 PM (#4454968)
I think it is far more likely for the Cubs to have their run differential catch up to their record than have their record catch up to their run differential.


Well it's sitting at +1 right now.
   24. valuearbitrageur Posted: May 29, 2013 at 11:10 PM (#4455345)
The 2007 DBacks had some amazing pitching, both starters and relievers, ended up averaging a 115 ERA+, their top 5 relievers ended up between at 146 and 179 ERA+. Their offense was execrable, 86 OPS+. They only had one hitter over a 109 OPS+, he hit .333/.349/.683/1.033 for a 152 OPS+, but was only 4th on the team in offensive WAR due to only having 60 ABs. That was their 4th starter, Micah Owings, who also threw 152 innings with a 111 ERA+.


Some of the next year was reversion to mean. Most of those relievers and starters weren't really so good (Micah Owings).

When I looked at 2007, it seemed like the real reason they outperformed is they rarely blew out the opposition, but whenever down by a good margin Melvin was happy to go to the dregs of his bullpen and turn it into a rout. Their worst hurlers were godawful, and helped them lose a few by double digits. I credit that very smart strategy to having no offense and wanting to rest the arms of his studs, as tomorrow was always another day.
   25. McCoy Posted: May 29, 2013 at 11:33 PM (#4455361)
Yes the Cubs are definitely going to keep on scoring 2 runs a game for around 5 or 6 games and then score 7 while having their pitchers throw a shut out. That is definitely going to keep on happening.
   26. McCoy Posted: May 30, 2013 at 01:03 PM (#4455848)
Take out the blowout games and basic BRef pyth says the Cubs should have won 18 games out of the 43 remaining games. Which is 3 more than they actually did. Put the blowouts back in and you have the Cubs with 24 wins instead of 21 wins.
   27. Chris Fluit Posted: May 30, 2013 at 10:58 PM (#4456483)
Speaking for O's fans, the attitude about winning percentage vs. run differential was more like:

April: Good start, but it won't last
May: This can't last, but let's enjoy it while it does
June: Wow, I can't believe they're still doing it
July: They've done it for four months, why not five?
August: They've done it for five months, why not six?
September: Who cares if Keith Law says we're a bad team, let's get into the playoffs and see what happens

The objections of O's fans was more about
1: those who insisted that the run differential/winning percentage would even out before the end of the year as if the O's would finish 0-20 to make up for their earlier good fortune
and, more importantly,
2: those who insisted that the O's would stink this year because they had a negative run differential last year without realizing that the O's made it into positive territory before the end of the season and the team that finished the season was significantly different than the team that started it.

So far, this season's results have been what we expected as O's fans: they haven't been as dominant in extra inning or one-run games but they've been a better team overall and they're over .500 and hanging around on the periphery of the playoff race.

But the comments about trying to convince Harold Reynolds are spot on.
   28. DKDC Posted: May 30, 2013 at 11:45 PM (#4456544)
Might as well OsJack this. The Orioles are 30-24, which puts them on pace for another 90 win season (although they do get to play the Astros in the other two thirds of the season)

They were 30-24 at this point last year too. Their run differential last year at this point was zero, this year it is +23.

The middle third of the season is where the 2012 Orioles really made their pythag fame. They went 27-27 despite being outscored by over a run per game on average.
   29. McCoy Posted: May 31, 2013 at 12:10 AM (#4456556)
In the 7 blowout wins the Cub pitchers have had this line, .389/.389/.722 with 4 runs and 10 RBI. The rest of the time the pitchers have hit .134/.155/.317 with 7 runs and 9 RBI.
   30. McCoy Posted: June 02, 2013 at 09:11 AM (#4458203)
So now in June the blowouts start off going to the other way for the Cubs.
   31. McCoy Posted: June 02, 2013 at 09:50 PM (#4458732)
Missed another blowout by 1 run.
   32. LargeBill Posted: June 02, 2013 at 10:48 PM (#4458761)
Another potential explanation for the outliers would be a team with so-so offense, very good front end of rotation and very very bad fourth and fifth starters. #1, 2, & 3 pitchers win lot of low scoring games and back end of rotation lose really ugly.
   33. Sunday silence Posted: June 03, 2013 at 09:52 AM (#4458906)
could we go through again why we think pythagorean data is useful for predicting? or whatever exactly it is good for? I know we went through it before and what I took from that is there is something like about 10% of games are more or less random in outcome; but I cant remember what reasoning I got to that pt.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Brian
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogOT: Politics, October 2014: Sunshine, Baseball, and Etch A Sketch: How Politicians Use Analogies
(2852 - 2:48pm, Oct 21)
Last: Bitter Mouse

NewsblogOT: Monthly NBA Thread - October 2014
(284 - 2:45pm, Oct 21)
Last: madvillain

NewsblogSo You’re About to Pitch to Pablo Sandoval | FanGraphs Baseball
(4 - 2:45pm, Oct 21)
Last: deputydrew

NewsblogBrisbee: The 5 worst commercials of the MLB postseason
(167 - 2:45pm, Oct 21)
Last: Jolly Old St. Nick Is A Jolly Old St. Crip

Newsblog2014 WORLD SERIES GAME 1 OMNICHATTER
(23 - 2:36pm, Oct 21)
Last: Hey - you and Tommy Lasorda...I hate Tommy Lasorda

NewsblogCardinals proud of fourth straight NLCS appearance | cardinals.com
(50 - 2:36pm, Oct 21)
Last: McCoy

NewsblogSielski: A friend fights for ex-Phillie Dick Allen's Hall of Fame induction
(91 - 2:36pm, Oct 21)
Last: Ron J2

NewsblogMorosi: Could Cain’s story make baseball king of sports world again?
(106 - 2:35pm, Oct 21)
Last: Jarrod HypnerotomachiaPoliphili(Teddy F. Ballgame)

NewsblogBaseball's hardest throwing bullpen - Beyond the Box Score
(5 - 2:14pm, Oct 21)
Last: Misirlou's been working for the drug squad

NewsblogRoyals’ James Shields passed kidney stone during ALCS but is ready for World Series | The Kansas City Star
(28 - 2:10pm, Oct 21)
Last: Jolly Old St. Nick Is A Jolly Old St. Crip

NewsblogBaseball Prospectus | Pebble Hunting: An Illustrated Guide to the People of Kauffman Stadium
(7 - 2:07pm, Oct 21)
Last: Tulo's Fishy Mullet (mrams)

NewsblogCalcaterra: So, if you’re not a fan of the Royals or Giants, who ya got?
(100 - 1:35pm, Oct 21)
Last: Belfry Bob

NewsblogFan Returns Home Run Ball to Ishikawa; Receives World Series tickets
(47 - 1:27pm, Oct 21)
Last: Lassus

NewsblogOT:  October 2014 - College Football thread
(445 - 12:59pm, Oct 21)
Last: andrewberg

NewsblogDealing or dueling – what’s a manager to do? | MGL on Baseball
(18 - 11:29am, Oct 21)
Last: GuyM

Page rendered in 0.3835 seconds
54 querie(s) executed