FanGraphs Baseball


RSS feed for comments on this post.

  1. great post.

    is the -.79 wins noise, i am curious as to what the expected range/standard deviation would be for a team “average” at these things over 120ish games.

    Comment by TheUnrepentantGunner — August 17, 2010 @ 4:03 pm

  2. Loved the article. Glad you can avoid listening to the scounts (and ESPN) and quantify the “little things” in some way other than just saying that they move the runner over and blah blah blah that we’ve all been hearing are necessary to win games since little league.

    Comment by Jim Lahey — August 17, 2010 @ 4:40 pm

  3. Thanks… I’m not trying to dismiss scouts, obviously, they’re an essential part of pro baseball and stuff. I was just using the comments (which are just his subjective impression, nothing wrong with that) as an occasion to see if whatever they are doing is actually “more efficient” in terms of getting wins. Doesn’t look like it. There are, of course, times when sacrifices and bunts and stuff ARE good ideas, and maybe the Padres are trying to do this. But the numbers say if they are, it ain’t workin’!

    But it’s more interesting than just saying: “wow, the Padres are playing much better than I thought they would,” which is my (boring) answer… I thought they’d win in the 70s.

    Comment by Matt Klaassen — August 17, 2010 @ 4:43 pm

  4. Okay, so, I’m a Giants fan, and yes, the Padres’ success is driving me nuts. I can accept their roster was better than I realized, and I can recognize that their offensive components are better than they look at first glance. Adrian Gonzalez is awesome, Venable actually contributes quite a bit, Torrealba has been surprisingly good, and Headley isn’t too bad either. All this is true, considering the environment they play in.

    Here’s the part that is really driving me nuts. Padres OPS this year to date is .704. Giants OPS this year to date is .732. And yet, the Padres offense has generated as many runs as the Giants so far, while playing 2 fewer games. This has nothing to do with how many runs the offense WOULD score or SHOULD score in any given environment or whatever. It’s everything to do with how many hits and walks and home runs they had in comparison to the Giants, and how the Giants are putting more men on base and hitting balls farther – and STILL the Padres are keeping pace with the Giants.

    Admittedly, this may well have to do more with the Giants hitting into a league leading 118 double plays than the Padres somehow being ultra-special, but still – AUGH! WHAT DOES IT TAKE?!

    Comment by Ott — August 17, 2010 @ 5:27 pm

  5. After having read some of these “Little things” pieces, I do kind of wonder if they’re not precise enough to have any useful information pop out from noise involved. Forgive me if I’m wrong, but I basically see this stat as doing the following:

    1. Taking the expected wins from a non-uniformly random ordering of events that happened (since linear weights takes the average of what people actually do, rather than just shuffling).
    2. Taking the expected wins from the actual ordering of events.
    3. Finding the difference.between these.

    Firstly, as has been seen with lineup effects- any such analysis of this nature tends to display only modest impacts. Secondly, the only “context” it adjusts for is the order that the events occurred. This seems to me like a pretty poor proxy for context. Isn’t a major part of the context going to be pitching-hitting matchups?

    Let’s assume you have a man on 1rst with 1 out and an average hitter at the plate. This is a pretty flexible situation. Sometimes you want to bunt, sometimes you want to try to draw a walk, and sometimes you want to swing away. If the next guy(s) in the lineup is a 3 true outcomes guy or hits for tons of power, you’re probably wasting your time to bunt. (Not captured here). If it’s a good contact hitter without much extra base power, you might very much want to bunt (assuming you only need one run).

    Basically, I think that a better test for this sort of thing might be more like the lineup analysis techniques where you compare the value of what happened versus some metric which actually takes into account who is on the team. With all due respect, the Padres do not have a lot of power- even discounting their ballpark. This could make something like bunting more useful than for the average team. By using standard linear weights, one is assuming that the Padres lineup is somehow comparable to other lineups.

    To me, that doesn’t sound like “little things.” Little things would involve harnessing more production out of your events by synergizing them with your skill set. It would not be trying to maximize the expectation for each event compared to if any random team took over your at bats for the rest of the game.

    To be quite honest, until the “little things” stat can account for the skill sets of a team’s line up- I think you’re going to end up with a lot more signal than noise with this stat anyways.

    Comment by B N — August 17, 2010 @ 5:41 pm

  6. Interesting I actually think the pitching is something that falls more in line with something a scout should notice. Watching some of the pitchers (especially Garland) it seems as if they are inclined to allow base runners in certain situations with a comfort that the defense/park would lessen the chance that hitter would score. Almost as if the expectancy of a runner scoring when on base was less given the defense and park and as opposed to giving the hitter a pitch that would potentially result in a XBH they will just walk him. Seems like with good defense and ballpark the value of slugging percentage increases over OBP.

    Comment by Christian — August 17, 2010 @ 6:01 pm

  7. I freely admit that this sort of analysis has its limits, although it is worth noting as an aside that the limits you mentioned are the same ones shared by WPA/LI (and to a certain extent, WPA, from which WPA/LI is derived). Some of the things you mention are siimlar to the complaint that “WPA doesn’t take into account relative team strength/player ability.” But those are acknowledged limits to the context.

    Aside from the separate issue (not broached here) of whether or not this represents a repeatable skill, I still believe that WPA/LI is a useful derivative of the “story-stat” WPA, but more useful because it is “unleveraged. ” No, it doesn’t add in complete team-and-player specific context, but it does add more context. More germane to the issue at hand, if we buy that it at least tells a significant part of the game story beyond traditional linear weights, it does reconstruct what happened in terms of “swings,” if not the reasons for making the decisionmaking. And relative to average, the Padres are not winning more games than the traditional linear weights would suggest, which was all I was getting at within the acknowleged limits of these numbers (which also apply to traditional linear weights as well!).

    I imagine that’s not satisfying as more than an acknowledgement, and too wordy at that, but thanks for your comment anyway.

    Comment by Matt Klaassen — August 17, 2010 @ 6:10 pm

  8. Oh, and Jim, are you going to finish a bottle of rye by yourself tonight?

    Comment by Matt Klaassen — August 17, 2010 @ 6:11 pm

  9. The Padres lead all of MLB in the Elias stat “Productive outs” at 39% (league average 32% in both AL and NL). This is defined as “the percentage of outs when either a) successful sac by pitcher with 1 out b) advancing any runner with none out or c) driving in a baserunner with the second out of the inning.

    Comment by Baggins — August 17, 2010 @ 6:19 pm

  10. Those Friars, they ‘play the game right’. And they’ve got a bunch of ‘gamers’ like Eckstein. Pitching, check; situational hitting, check; team speed, check; young and cheap, double check.

    Comment by Fergie348 — August 17, 2010 @ 6:30 pm

  11. Can you tell I’m bitter. Yes, I am a Giants fan..

    Comment by Fergie348 — August 17, 2010 @ 6:31 pm

  12. Interesting, as the numbers above idicate that it isn’t helping them win more games than their traditional linear weighs say they should.

    Comment by Matt Klaassen — August 17, 2010 @ 6:54 pm

  13. The lower the average RPG, the less valuable outs re.

    Comment by Matt Klaassen — August 17, 2010 @ 6:55 pm

  14. Actually, if it’s any consolation to you as a Giants fan, the Giants’ “Little Things” score is .86 — about a win-and-a-half better than the Padres.

    Comment by Matt Klaassen — August 17, 2010 @ 6:58 pm

  15. Doesn’t WPA discredit advancing the runner in most situations? That is my general impression.

    Comment by James — August 17, 2010 @ 7:37 pm

  16. Also, I don’t believe the leverage stats don’t take into account the context of the opponent you are playing or the team that you are playing for – in other words, the “leverage”/”clutch” stats don’t account for the subjectiveness of the situation. For example, doing the little things on a competing team (e.g., the Padres) to improve win probability against a competing division rival (e.g., the Giants – who the Padres happen to be dominating this year) should have greater weight than doing those things against non-divisional or otherwise mediocre opponents.

    The gist of this comment is that playing well in the clutch/high leverage situations for a good team in meaningful games is different than doing so for a poor team in relatively meaningless games (is there such thing as a high leverage situation in a Royals-Orioles game), and I don’t believe the stats capture that.

    Comment by James — August 17, 2010 @ 7:49 pm

  17. Best post I’ve read in a week.

    Comment by fangraphsfan — August 17, 2010 @ 8:28 pm

  18. “In addition, runs above/below average is baselined against all of the MLB, and includes pitchers hitting.”

    I thought you guys excluded pitcher hitting in your LWTS/wOBA calculations?

    Comment by J.T. Jordan — August 17, 2010 @ 8:44 pm

  19. Padres fan here. I don’t think you are going to find any numbers that explain how the team is better than the sum of their parts. That’s what’s going on here. That and a bullpen for the ages. The three headed monster of Gregerson, Adams, and Bell, well I can’t think of any bullpen that has three such dominating pitchers as these. They have one of the best road records of any team in either league, so it’s not all park factors.

    Comment by jirish — August 18, 2010 @ 12:02 am

  20. I think it has to do with the Padres just being better on the basepaths than the Giants. According to the Baseball Prospectus numbers the Padres have been 7.8 runs better in terms of EQBRR.

    Comment by Drakos — August 18, 2010 @ 12:48 am

  21. The figures from the “team value pages” look strange to me. They say the Padres have an above-average offense, but also a below-average pitching staff. In fact, their starting rotation is ranked 3rd-lowest, with only the Brewers and Pirates being worse. This is completely different than what you’d expect from wOBA (13th of 16) or FIP (1st). Is this all because of park adjustments? It’s hard to believe Petco is that big of a factor.

    Comment by neurofarm — August 18, 2010 @ 1:09 am

  22. It’s partly Petco, which has a truly huge effect on both the hitters and pitcher. Most of the pitchers are also doing well on the road, but the park adjustments don’t make that distinction.

    Also, the pitching staff’s value is being dragged down by the starting staff, where three of the five pitchers have lousy FIP, but good ERA. The bullpen has been so good that not even the Petco adjustment can keep them from being the best in baseball, value-wise.

    Comment by maestro876 — August 18, 2010 @ 1:56 am

  23. I got 2 reasons for you: Chris Denorfia!!! DENO RULES!!!!

    Comment by DonCoburleone — August 18, 2010 @ 2:34 am

  24. One thing the Padres do have though is a legitimate Ace – Matt Latos. I know his innings limit (and pitches per start limits) hurt his value, but the innings he gives them are elite.

    Comment by DonCoburleone — August 18, 2010 @ 2:40 am

  25. Deno! Deno! Deno!!!

    Comment by DonCoburleone — August 18, 2010 @ 2:41 am

Leave a comment

Line and paragraph breaks automatic, e-mail address never displayed, HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Close this window.

0.349 Powered by WordPress