FanGraphs Baseball

Comments

RSS feed for comments on this post.

  1. WOWY is right. I’ll literally be on the edge of my seat to read your conclusions!!!

    Comment by scstrato — January 7, 2009 @ 11:40 am

  2. Wow….this is quite an interesting article and I would love to hear more!

    Comment by Jim — January 7, 2009 @ 11:46 am

  3. Welcome to FanGraphs, Brian! This is a topic that I think holds a lot of promise for making DIPS theory less abstract and more usable for us armchair sabermetricians. I’ve looked at a lot of batted-ball data here and have grown increasingly uncomfortable with the +.12 magic trick. Keep fighting the good fight!

    Comment by Mark R. — January 7, 2009 @ 11:57 am

  4. Forgive my ignorance, but who is actually doing the coding on batted balls? Is it one person in each park? Or is it a crew? Team paid? MLB paid? Does it vary?

    Regardless, what is meant by “ball park variance” is actually variance in the humans who are doing the coding, right?

    Any idea if the Phils changed the person who coded hits when they changed stadiums? Their difference in coding is substantially larger than any of the other teams that changed sites over this period (while the sample size for Vets is small, it isn’t that much worse than the others).

    Or am I totally wrong?

    Comment by rwildernessr — January 7, 2009 @ 12:20 pm

  5. I have heard before that batted ball results had park factors. Some parks produce more GB, and some produce more LD. In fact, I think I learned this from Jeff Sullivan over at Lookout Landing.

    Comment by Evan — January 7, 2009 @ 12:25 pm

  6. Great article and very thought provoking. If a sabermetrician isn’t challenging fundamental assumptions, he’s not doing his job. Keep up the good work. We should be looking into these questions.

    Comment by NYRoyal — January 7, 2009 @ 12:38 pm

  7. Wow this is excellent Brian, thanks for the work. I’ve seen allusions to thinks like “weak line-drives” in the past which would mess up ebabip a lot, but to put everything in a numerical context makes it much easier to follow.

    Comment by Graham — January 7, 2009 @ 1:12 pm

  8. Welcome to the site! This is an interesting idea, though I’m initially apprehensive. I would be curious to see teams’ home/away splits, because the big factor you can’t control for in a ballpark is the home team. For example, with the exception of a couple of hitters in the middle of the lineup, I see the Twins as a fairly light-hitting team based more on speed and defense than on hitting powerful line drives up the middle. If a couple of those young pitchers who anchor their rotation also happen to be somewhat groundball- or flyball-inclined, the Metrodome might be deflating line drives simply because the Twins don’t have hitters who tend to hit them or pitchers who tend to give them up.

    If you were to compare each team’s tendancies to hit/give up line drives at home vs. their tendancies to hit/give up line drives away, and find a statistically significant difference, I will be convinced. I bet that there is somewhat of a factor–small ballparks deflate foul popups, thus inflating LD’s–but I don’t think it’s as big as you make it sound here.

    Comment by Jake — January 7, 2009 @ 1:19 pm

  9. “Houston and its opponents both have 18% fewer balls scored as liners in Houston than they do on the road.”

    Pay attention.

    Comment by Rex — January 7, 2009 @ 1:26 pm

  10. I’ve always wondered about the amount of unintentional human bias in the coding of these things as well. Similar to the way official scorers have their biases, umpires have their own “strikezones”, I’ve often felt the coders might as well.

    Do we know if different coders are used for a park over the course of a year or does it vary? Does one stat warehouse vary the coders while another does not?

    Comment by Doc | ballssticksstuff.com — January 7, 2009 @ 1:28 pm

  11. Giving one number means nothing. For all I know, teams usually vary 18% between their home park and their away park. To show that there’s a big difference, you have to show that there’s a significant home/away split year after year.

    Comment by Jake — January 7, 2009 @ 1:32 pm

  12. *sigh*

    Anyone else want to try?

    Comment by Rex — January 7, 2009 @ 1:34 pm

  13. Interesting stuff, I am currently trying to figure out the anomaly that is Ryan Ludwick. He has a LD% of 28.1% (highest in MLB last year) and GB% of 27.4% (lowest in baseball).

    I don’t know what to make of it. Was it just luck or skill? Is it repeatable? The best I can make of is that his power is very much for real.

    Comment by FlimtotheFlam — January 7, 2009 @ 2:21 pm

  14. Brian,

    I can’t tell if you tried to account for potential biases in the scoring codes for hit type (LD, GB, FB, PU) in your attempts to improve upon BABIP. You clearly notice them, and I don’t think it’d be that difficult to correct for the biases you note.

    Comment by Matt Mitchell — January 7, 2009 @ 2:27 pm

  15. Doc,

    I’m pretty sure there are multiple people who score games for each team. The authority on this would be Dave Smith, who can be contacted through Retrosheet’s website.

    Comment by Matt Mitchell — January 7, 2009 @ 2:28 pm

  16. Jake, it’s not a home road split that’s at issue. It’s the bias of the Retrosheet volunteer who scores the game so it can be coded into those wonderful play-by-play files that we all love.

    Rex, Jake’s right. One data point doesn’t mean much. In this case, it’s an example, not a conclusion. I believe Jake is looking for a conclusion from the data.

    Comment by Matt Mitchell — January 7, 2009 @ 2:32 pm

  17. I guess though, that I was looking at that data in a different way. It is possible that personal subjectivity plays a role, as was noted in yours as well as other comments. However, I was looking at it to mean that ballpark configuration can play a role in LD%. I guess it could be anything from mound height, batter’s eye, climate conditions, lighting, whatever. Maybe some parks just have a kickass underground batting cage, who knows. But it does suggest to me that some controllable conditions exist in a ballpark that a home team could use to its advantage.

    Comment by China Brown — January 7, 2009 @ 2:48 pm

  18. Interesting analysis. A quick way to check for scorer bias based on the outcome of the BIP would be to check the correlation of LD% with the FB out% (and GB out%) by park. I think you’ll find a strong correlation, suggesting that ambiguous BIP are coded as LDs if they aren’t caught, but FB or GB if they are caught. Alternatively, if BIP are being coded correctly, then there should be little or no correlation with the proportion of FBs and GBs that are turned into outs.

    Comment by Guy — January 7, 2009 @ 2:49 pm

  19. Any theories about what could cause that? The only thing that springs to my mind is that I would guess a park with a more difficult “background” (behind the pitcher) to hit against would produce fewer LDs. But are there any other physical characteristics of a park that could possibly effect how the ball comes off the bat??

    Comment by rwildernessr — January 7, 2009 @ 2:54 pm

  20. Nice article. Look forward to your future stuff.

    I would note that regarding DIPS theory, Tom Tippett showed that there are some pitchers who can control their BABIP and, if I recall right, Tango at The Book found that it takes about 7 years for a pitcher’s affect on BABIP to show up conclusively (or something like that).

    Also, what about the issue of Fliners? Plus, could the differences in the parks be related to human error in interpreting the difference between a liner and fly ball? Don’t know if that is true or how it is done, but just bringing up a thought that popped up.

    Comment by obsessivegiantscompulsive — January 7, 2009 @ 3:17 pm

  21. As far as the home/road split is concerned, I’m trying to control for the fact that every game played in the Metrodome features the Twins and every game played in Arlington features the Texas Rangers. Brian’s calculation of the LDf stat was ambiguous; if he compared line drives in the Metrodome to line drives by Twins/Twins opponents outside the Metrodome, then I have no objections and that is what I was talking about. If he compared line drives in the Metrodome to line drives ON AVERAGE, then he is not controlling for the fact that teams in the Metrodome are either the Twins or facing the Twins and this is why I want home road splits.

    Anyways, I imagine it’s not JUST the scorer bias, since there will be more pop-up and flyball outs in stadiums with more foul territory, and as a result slightly fewer batters will survive to hit a line drive. As far as I can tell, nothing else would really make much of a difference, except possibly issues concerning hitters’ vision such as the background and lighting. Also, it is entirely possible that hitters in the Metrodome are just really bummed out that they’re playing in such a crappy ballpark, and as a result make worse contact.

    My proposed solution to the scorer bias issue is simple: we have Bill James watch every play and make the ruling.

    Comment by Jake — January 7, 2009 @ 3:27 pm

  22. I hate to sound like both an idiot and a critic at the same time, but I had a bit of a hard time following everything in your analysis. What I did understand was extremely interesting, but if you made your writing a bit more concise and focused, I think it would be easier to follow for those of us who don’t grasp all of the baseball stats work as quickly or as intuitively as others.

    Thanks.

    Comment by NadavT — January 7, 2009 @ 3:33 pm

  23. One of the things that periodically makes me want to bang my head against the wall is when I see a good prospect tearing up the minors, and people worrying that his high BABIP means he’s been “lucky”. The evidence has always suggested that hitters have significantly more control over BABIP than pitchers. And the over emphasis on LD% ignores that some hitters also fare very well on fly balls (those with good XBH power) or even ground balls (those with exceptional speed).

    Especially when you see a guy with a high xbh%, you will see that these guys are often very productive on fly balls. Of course, some of the better hit fly balls will end up not being “in play”, as they are hit over the fence. But, many of the doubles and triples, while “in play”, aren’t really “playable” either. How much control do fielders really have over those?

    I suspect that sorting out what are really park effects from scorer effects would be difficult, but both will play a role. At least some effects will be due to the ballpark/environement, such as breaking balls not breaking as sharply at higher altitudes and in dryer air, or the batter’s eye, or how well the infield is maintained. But the scoring is likely to be unreliable as well.

    Now, it would really be interesting if we had something like PitchFx for batted balls? That would make for some interesting data, if you could have those computers tell you the speed and trajectory of the ball off the bat, and categorize batted balls based on that.

    Comment by acerimusdux — January 7, 2009 @ 3:57 pm

  24. Jake is right that in order for the effect to be considered systematic, it needs to be reproduced year after year in the same conditions. Otherwise, it’s an outlier. If we see the same strange things in Houston over three straight years, there’s something weird there.

    However, the evidence is already mighty suggestive that the effect will be there.

    Comment by Pizza Cutter — January 7, 2009 @ 4:19 pm

  25. Very much looking forward to the follow up articles on this.

    Comment by t ball — January 7, 2009 @ 4:34 pm

  26. “If you were to compare each team’s tendancies to hit/give up line drives at home vs. their tendancies to hit/give up line drives away, and find a statistically significant difference, I will be convinced.”

    Yes, that’s exactly what he’s saying he’s doing here. That the way park factors are normally done, usually for something like runs, rather than line drives.

    He didn’t spell the formula out exactly, but I think it should be assumed to be:

    ( ( home LD + home LDA ) / home G ) / ( ( road LD + road LDA ) / road G )

    where LD is the line drives hit by the team and LDA is line drives allowed by the team.

    One thing this won’t separate is the actual effects from the ballpark vs. the scorers in that park. You are really adjusting for both at the same time. You really don’t know why LDs were higher at a particular park, just that they were.

    Also, if both Huston and it’s opponents both had line drives reduced by 18%, that is likely the result of external factors (park, scorer, etc). But what if one was impact 36% and the other 0%? That would average to 18%, but might be less reliable an indicator. This does tend to be a problem in single season park factors, but using 6 years of data should have minimized that problem.

    Of course there is still a possibility that there are *some* team related impacts. I would bet you would see this though most in the more unusual environments, like Coors, where teams might be tempted to build the team partly to suit their own stadium. It might be interesting to see where there was the greatest discrepancy between home team and visiting team effects in the sample.

    Comment by acerimusdux — January 7, 2009 @ 4:41 pm

  27. great work, Brian. i look forward to ‘results’ — whenever they arrive.

    want me to proofread future articles for you? proofreading for stat guys is my ‘niche’ in the sabermetric world.

    Comment by robert j. — January 7, 2009 @ 5:45 pm

  28. There are certainly biases with the Retrosheet scorers, since a different set of scorers are used in each city.

    But I thought most of the people doing this type of analysis were at The Hardball Times, and they use BIS data I believe not Retrosheet, and the BIS data is input by a central team in Pennsylvania, so IF there are biases in the Retrosheet scoring, it should be easy to see by comparing it to the BIS data.

    Comment by KJOK — January 7, 2009 @ 5:59 pm

  29. “If you were to compare each team’s tendancies to hit/give up line drives at home vs. their tendancies to hit/give up line drives away, and find a statistically significant difference, I will be convinced”

    That’s how matched pairs work. I took how the Twins and the Yankess did in the Metrodome in bucket one, how the same two teams did in Yankee Stadium in bucket two. The the Twins and the Tigers, etc, repeating for every every combination of teams and ballparks. Then the two buckets are summed, grouping by ballpark, and the expected (road) totals from bucket two ar compared to the observed (home) totals from bucket one.

    “I have heard before that batted ball results had park factors. Some parks produce more GB, ”

    I think GB and FB rates by ballpark are greatly influenced by the pitchers. This will need a WOWY/matched pairs approach to control for the pitchers. I doubt that I’ll find a real difference, I think that’s something that comes from the batters and pitchers, but you never know.

    I have also calculated foul fly factors, and I will be looking into how they affect the other rates. For example, if two parks have the same percent of fly balls that go for homeruns, but one has a much higher rae of foul flies than the other, how much does that alter the HR park factor?

    Comment by Brian Cartwright — January 7, 2009 @ 6:08 pm

  30. Part of the point I am trying to make here is that deciding what’s a line drive or a “soft line drive” (sounds like an oxymoron, but it appears in GameDay) or a fly ball is subjective, which is something I like to avoid when possible. If it’s all you have, learn how to measure any biases.

    Comment by Brian Cartwright — January 7, 2009 @ 6:15 pm

  31. I’d venture there’s a fair share of people doing analysis off both Gameday and Retorsheet data along with the BIS data. For those who don’t know, all the batted ball data on FanGraphs is BIS data and they are further broken out into Fliners, which come in the Fly Ball and Line Drive variety.

    Fliners are considered the edge cases.

    Comment by David Appelman — January 7, 2009 @ 6:28 pm

  32. “if he compared line drives in the Metrodome to line drives by Twins/Twins opponents outside the Metrodome, then I have no objections and that is what I was talking about”

    Yes, that;s what I did.

    “My proposed solution to the scorer bias issue is simple: we have Bill James watch every play and make the ruling”

    Or tell us where the ball landed and how long it took to get there (hang time). From that we can calculate speed off bat (see hittrackeronline.com). Nice objctive measures – except BIS and Stats can seem to agree on where the same balls land.

    Comment by Brian Cartwright — January 7, 2009 @ 6:29 pm

  33. “He didn’t spell the formula out exactly, but I think it should be assumed to be:

    ( ( home LD + home LDA ) / home G ) / ( ( road LD + road LDA ) / road G )”

    I measure it more precisely. Home and Road are grouped by opponent and road ballpark, and the denominator in this case is (FB+LD)
    ((homeLD)/(homeLD+homeFB)) / ((roadLD)/(roadLD+roadFB))

    The result can be expressed as a ratio, a difference, or as a rate scaled to a mean.

    Comment by Brian Cartwright — January 7, 2009 @ 6:51 pm

  34. The last guy I know of to post an insane line like that was Ryan Howard, 2006. He has not repeated that performance, and he’s not been the same without it.

    Comment by philosofool — January 7, 2009 @ 7:14 pm

  35. From 2007 to 2008 (major league data only), Ludwick dropped his GB% from .38 to .27 (avg .46) whil raising his LD% from .19 to .29 (avg .20). Marcel projects LD% .25 GB% .30.

    It looks like he took about 40 grounders and centered them better, turining them into liners, and thus boosting his BA. Including minor league data, I project Ludwick for 269/336/513, a 314 BABIP.

    He should consistently hit 25-30 HRs, but I think he can’t be counted on for a BA above .270.

    Comment by Brian Cartwright — January 7, 2009 @ 7:21 pm

  36. Where does retrosheet get the batted ball type data? Do they get it from one of the stat providers or score it themselves?

    Comment by Sean — January 7, 2009 @ 8:53 pm

  37. “For all the batters from 2003-2008, in non-bunt plate appearances, I added up the base hits, line drives, ground ball, fly balls and popups. I compared the predicted BABIP to the observed one in each season, which showed a root mean square (RMS) error of .045. Then I compared each years predicted value to the next years observed, and the RMS was .048 – slightly larger. For pitchers, the RMS was .039 in the same season, .039 in the next. I don’t see the evidence of future regression.”

    This is meaningless, isn’t it? If these are aggregate stats, you’re not going to see regression because each year’s aggregated stats will include all underperformers and all overperformers. As a group, then, their regression (or progression) back to their established tendencies will produce no net difference for the aggregate stats. What am I missing here?

    Comment by Dudeski — January 7, 2009 @ 10:18 pm

  38. Atmospheric conditions are one possibility-I’ve looked at 5-year data of Coors and found similar results, probably at least in part because pitches don’t move as much and are easier to center. Thick air in places like San Diego would have the opposite result. The hitters’ eye is another likely cause. The mound could also make a difference, because every mound has its own unique feel on the landing spot (for example, they could have slightly different slopes from the rubber to the landing area, different holes dug into the rubber and into the landing area, etc), and some are easier for pitchers to get a feel for their pitches on. Lighting variances could be an issue, especially if the factors are exaggerated in night games. For some parks, the configuration or turf could influence hitters’ and pitchers’ approaches as well-like trying to hit more or induce fewer fly balls in a home run park or hit more or induce fewer ground balls on a fast surface-that affect batted ball rates. That would be complex and difficult to quantify, but I wouldn’t doubt that is part of the effect showing up here.

    Comment by Kincaid — January 7, 2009 @ 11:06 pm

  39. I was trying to see if the error rate in year+1 was smaller than the error rate in year0.

    How about this – there were 57 pitcher-seasons with 300 or more Balls in play from 2003-2008 where the pitcher had a BABIP more than 50 points better than the rate predicted by his LD%. These are the pitchers having “unusually” good luck. 50 points is three times higher than the six season rms of .016.

    In the next season, 25 of those 57 still overperformed their previous estimate by at least 50 points. Another 24 were still higher than .016.

    Comment by Brian Cartwright — January 7, 2009 @ 11:23 pm

  40. My point is you’ve got a small group (2 or 3 guys maybe?) at the BIS office marking via TV or Videotape all of the MLB games, vs. Retrosheet with 30 separate small groups who only mark games in the same city. Looking at the Hardball Times annual, all teams fall within .18 – .22 line drive rate.

    Houston batters were .19 in The Hardball Times annual, just slightly off the 0.20 MLB average, while Brian is geting a 0.82 LD park factor from Minute Maid Park, which would indicate a much larger deviation from the norm in teh Retrosheet data for Houston. However, Hunter Pence did have the lowest LD% even in the BIS data.

    Comment by KJOK — January 8, 2009 @ 12:50 am

  41. Very interesting. I look forward to future analysis.

    Comment by Kyle Boddy — January 8, 2009 @ 1:56 am

  42. you might look at the rate of LD to battery experience, especially catchers. ;-)

    Comment by Yousif — January 9, 2009 @ 12:38 am

  43. To make a point of the obvious, the pitchers mentioned who outperform BABIP estimates are mostly good pitchers, the underperformers seem shaky. I think that this result (under/overperformance of estimate) is evidence of a certain skill set rather than the cause of them being good or bad. I guess in a way this is logically obvious, but I think that studying the outliers will yield insights. It seems unlikely that the performance of Maddux, Zambrono, or Webb could be explained as largely a park factor- am I wrong?

    Comment by Thejeffg — January 9, 2009 @ 5:15 pm

  44. [...] What I Hate About Line Drives | FanGraphs Baseball When is a line drive not a line drive? [...]

    Pingback by links for 2009-01-10 — January 10, 2009 @ 9:04 am

  45. I agree with your obersvation, those who outperform their unadjusted expected values are better pitchers, and those who underperform are worse (and given fewer opportunities).

    The next step is to try to isolate the pitcher’s contribution, controlling for ballpark factors, and also the defense behind each pitcher and the batters that he has pitched to. What remains should be each pitcher’s ability to make his GB, LD & FB easier or harder to catch.

    Comment by Brian Cartwright — January 10, 2009 @ 10:57 am

  46. I remember seeing some research last year- sorry I can’t be any more specific- that seemed to demonstrate that certain pitchers had the ability to reach back for something extra in high-leverage situations. Pitchers that have that extra something might reach back for the K and get a weak popup or grounder instead as the batter flails-barely making contact to avoid the strikeout.

    I think this dominance factor could be at the root of the performance of these outliers. Imagine, if you will, Big Fat Sydney Ponson reaching back for something extra and coming up with a ham sandwich and mug of beer.

    Comment by Thejeffg — January 10, 2009 @ 11:24 am

  47. [...] however, a similar concern about subjectivity of the classification applies. There has been speculation that, given similar trajectories, the observer might be more likely to classify a batted ball as a [...]

    Pingback by Confessions of a DIPS apostate | MLB This Week — March 4, 2009 @ 9:35 am

  48. “Big Fat Sydney Ponson reaching back for something extra and coming up with a ham sandwich and mug of beer”……now wiping coffee from my monitor. thanks for that, great to start the day with a hearty guffaw.

    Comment by Jay in BMore — March 5, 2009 @ 7:01 am

  49. [...] answer lies in how a ball put into play is classified. Fellow Fan Graphs author Brian Cartwright wrote a very interesting piece on line drive rates by stadium, noting that there is a wide disparity between parks in terms of how [...]

    Pingback by Howie Kendrick And Weird BABIP Splits | FanGraphs Fantasy Baseball — March 18, 2009 @ 2:15 pm

  50. [...] and the average line-drive rate (sans Young) was 20.5% vs. 19.9% league wide. Furthermore, in a study at Fangraphs, Brian Cartwright determined that “a batter is 18% more likely to have a batted ball coded as [...]

    Pingback by The Case of Michael Young and Line Drive Rates | Baseball - NBL Daily News Network — March 20, 2009 @ 9:40 am

  51. [...] distinction is a subjective one made by the official scorer (and the rate at which liners are coded varies greatly by stadium). That being said, B.J.’s liner rate has fallen from over 19 percent from 2007-2008 to 14.1 [...]

    Pingback by What’s With B.J. Upton? | FanGraphs Fantasy Baseball — September 20, 2009 @ 7:11 am

  52. [...] distinction is a subjective one made by the official scorer (and the rate at which liners are coded varies greatly by stadium). That being said, B.J.’s liner rate has fallen from over 19 percent from 2007-2008 to 14.1 [...]

    Pingback by What’s With B.J. Upton? — September 20, 2009 @ 7:30 am

  53. [...] just “exist”-they’re a subjective judgment by the official scorer. And as Brian Cartwright displayed last off-season, the rate at which line drives are coded can vary dramatically by stadium. It wouldn’t seem [...]

    Pingback by Nothing Wrong with Nolasco | FanGraphs Fantasy Baseball — October 3, 2009 @ 7:51 pm

  54. Did Brian ever run this study anywhere that he mentioned towards the end of this post???

    Comment by Dan Budreika — December 29, 2009 @ 3:00 pm

  55. wats6@mail.ru
    - ???????????? ?????? ?? 30$, ???????? ?? 10$
    - ???????? ??????? ?? 30000 ????????????? ??????? – 20$
    - ???????? ??????? ?? 1000 ????????????? ???????? ????? – 7$
    - ??????????? ? 6000 ????????????? ????????? – 7$
    - ???????? ??????? ?? 30000 ???????????? ??????? – 20$
    -???????? ??????? ?? ???????????? ????? (????????,????? ? ??.) – ???? ????
    + ???????? 60$. ????????? ???????? 10-20$.???????? ???? ????? ???.
    -?????? ???????? ??????? ?? ????? ????? ????????? 0.2$ ?? ?????? ??????????? (?? ?????? 200 ???????????)
    -???????? ?? ???????? 100$ -?? ???? ??? ?????? ????? ???????? ?? ????? ????? 150$ ?????? 50$ ?????? ????? ???????? ?????????
    -??????????? ????????? ?????? 1000$ ??????? + 100$ ? ????? ?? ???????? ??? ?? ?????????????.
    ??????????? ????????? ????????????? ?? ??????? 1000 ?????????? ? ?????,????????.
    -E-mai ???????? (?????? ?? ????? ????? ???????????) 20$ ?? 150000 ???????. ????? ?? ??????? ?????? ?? ?? ?????? 10$
    -???? ??? ????? ?????? ??? ???? ??????????.
    ???????????? ??????, ????????-????????? ??? ?? ??????? ???????? ??? ? ?? ???????? “? ????”.
    ?? ???? ???? ????????? ?????,?????,????????,????? ?????,???????????,???????????,?????,
    ????????,???,????? ? ?????? ??????.????? ? ??????? ?????????????? ? ???.????? ???????? ???????
    ????? 1 – 5 ??????? ?????? ? ????????? 10$-30$ ???? ????? 100 ??????? ????????.
    - ????? ???????????? ???????? (??????).
    wats6@mail.ru

    Comment by reldaks — January 19, 2010 @ 7:19 am

  56. gfd23@mail.ru
    - ???????????? ?????? ?? 30$, ???????? ?? 10$
    - ???????? ??????? ?? 30000 ????????????? ??????? – 20$
    - ???????? ??????? ?? 1000 ????????????? ???????? ????? – 7$
    - ??????????? ? 6000 ????????????? ????????? – 7$
    - ???????? ??????? ?? 30000 ???????????? ??????? – 20$
    -???????? ??????? ?? ???????????? ????? (????????,????? ? ??.) – ???? ????
    + ???????? 60$. ????????? ???????? 10-20$.???????? ???? ????? ???.
    -?????? ???????? ??????? ?? ????? ????? ????????? 0.2$ ?? ?????? ??????????? (??????? 200 ???????????)
    -???????? ?? ???????? 100$ -?? ???? ??? ?????? ????? ???????? ?? ????? ????? 150$ ?????? 50$ ?????? ????? ???????? ?????????
    -??????????? ????????? ?????? 1000$ ??????? + 100$ ? ????? ?? ???????? ??? ?? ?????????????.
    ??????????? ????????? ????????????? ?? ??????? 1000 ?????????? ? ?????,????????.
    -???????? ?? E-mail (?????? ?? ????? ????? ???????????) 20$ ?? 150000 ???????. ????? ?? ??????? ?????? ?? ?? ?????? 10$
    -???? ??? ????? ?????? ??? ???? ??????????.
    ???????????? ??????, ????????-????????? ??? ?? ??????? ???????? ??? ? ?? ???????? “? ????”.
    ?? ???? ????? ????????? ?????,?????,????????,?????,???????????,???????????,?????,
    ????????,???,????? ? ?????? ??????????.????? ? ??????? ?????????????? ? ???.????? ???????? ???????
    ????? 1 – 5 ??????? ?????? ? ????????? 10$-30$ ???? ????? 100 ??????? ????????.
    - ????? ???????????? ???????? (??????).
    gfd23@mail.ru

    Comment by kerbets — January 26, 2010 @ 3:26 am

  57. I read an article (I think it was at Beyond the Boxscore, but I’m not positive) that looked at the differences between line drives and fly balls among ballparks, and attempted to correlate the ratio to press box height. The idea was that some press boxes are close to field level, and some are really high, and the coder’s perspective could change their observation. That article had more information on the people who do the coding, but IIRC there are one or two people responsible for batted ball descriptions at each game.

    I wonder if the hit f/x data can eventually be used to help characterize line drives vs. fly balls. The speed and angle off the bat are the primary factors determining a ball’s trajectory, and those could sorted into bins corresponding to line drive, fly ball, etc. and possibly additional modifiers, such as “weak line drive” or “high fly ball”.

    Comment by Jeremiah — January 26, 2010 @ 3:44 pm

  58. I’m pretty sure that’s part of the plan for Hit F/X, whenever it is finally and fully implemented. The raw data should be available in the same way Pitch F/X data is, so it should be possible to categorize hits correctly regardless of how they’re “officially” coded.

    Comment by joser — January 26, 2010 @ 10:30 pm

  59. [...] put in play. Point to the line drive rate if you’d like, but one person’s line drive can be another’s fly [...]

    Pingback by Don’t Give Up On…Nolan Reimold | FanGraphs Fantasy Baseball — May 6, 2010 @ 12:47 pm

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>


Current day month ye@r *

Close this window.

0.104 Powered by WordPress