The LoHud Yankees Blog

A New York Yankees blog by Chad Jennings and the staff of The Journal News


Pinch hitting: Yair Rosenberg

Posted by: Chad Jennings - Posted in Misc on Jan 21, 2010 Print This Post Print This Post | Email This Post Email This Post

Now batting in the Pinch Hitters series is Yair Rosenberg, presenting a skeptical but not dismissive view of advanced statistics in baseball.

Yair is a junior at Harvard College, where he’s an Arts and Culture editor at The Crimson. Yair wrote that he’s been enjoying Yankees life in the middle of Red Sox country, “from the recent World Series to watching the Giants beat the Patriots in the Super Bowl during my freshman year.”

———

To hear many a sabermetrics buff tell it, Derek “Intangibles” Jeter brings nothing but a sound shortstop skill set to the table. A.J. Burnett is actually a rather consistent pitcher. And Mariano Rivera has a phenomenal cutter, but not some magical “closer’s makeup” that helps him succeed when the stakes are highest. Clutch hitting? Likely a myth. Some pitchers displaying the ability to “win games?” Random chance. Given a sufficient sample size, we are told players’ hot streaks and cold streaks emerge as mere coincidence.

This argument against overestimating an individual player’s ability to influence a game has much to recommend it, and I won’t rehash the evidence here. My intent is instead to point out an unfortunate consequence of our Moneyball-influenced era: The tendency of statistical measures to unintentionally obscure the human side of baseball.

The more statistically-minded baseball community has often adopted the following implicit assumption: Players are essentially machines, largely unaffected by clubhouse atmosphere, personal psychological factors or the day-to-day effects of real life. Columnists who refer to “team cohesion” or a player’s “mindset” (think Alex Rodriguez) as factors in performance are treated with indifference, if not derision, and are considered a product of a bygone era where intuition trumped hard data. By contrast, a modern talent evaluator like Billy Beane looks at advanced metrics and finds the right players to draft without ever observing them in person. A player’s performance can thus be predicted, fantasy baseball style, without reference to anything but the numbers.

But this emphasis on statistics demonstrates a remarkable failure of imagination. What other occupation in the world do we evaluate in a similar fashion? Aren’t there days when one phones it in at the office, when one’s focus is diverted due to a family concern, or when recent failure throws off one’s confidence and productivity? Don’t some people thrive under pressure, while others collapse? Why should baseball players be any different? Could it be that some players only give a half-hearted effort on some days, while others bring their all with consistency? That some are better at coping with failure or unexpected bad breaks than others?

Certainly we frequently observe that an error or freak play (e.g. Johnny Damon’s double stolen base) can cause a previously flawless pitcher to unravel. Mentally, this makes sense. We also find that most elite closers seem far more human come playoff time, even against teams they’ve handled easily in the regular season (Joe Nathan and Billy Wagner, anyone?). Of course, these players could just be the victims of a small sample size — but they could also be buckling under the pressure of the postseason. After all, how likely is it that so many closers who have converted 95% of their career save opportunities somehow all choose the playoffs to have some of their worst — and statistically unlikely — outings? Something seems to be up.

And is it such a stretch to consider the possibility that some pitchers might do a better job — on average — of holding a lead than others? That they might focus on their craft better when pushed into a corner, where less successful starters might often freeze up or overthink their pitches? Consider the college student who falls asleep and wakes up with four hours to complete a mostly unfinished term paper. Does that student suddenly lock in to the task at hand, or become unable to concentrate with the deadline looming?

While we may not yet have adequate metrics to discern the impact of such mental and personal factors on baseball performance — to distinguish statistical noise from psychological poise — that does not mean the factors themselves do not have an impact. Indeed, intuitively, if we view baseball as a real-world job like any other, we should come to the very opposite conclusion.

Ultimately, the numbers don’t lie, but sometimes people exaggerate their significance. While statistical measures are good at tabulating a player’s value and production, they don’t tell us nearly as much about how he got there. Until we come up with better formulas which can isolate and take into account the human aspect of the sport, we still need old fashioned, pre-Moneyball baseball men and scouts to discern player makeup, and yes, intangibles.

 
 

Advertisement

108 Responses to “Pinch hitting: Yair Rosenberg”

  1. ArtieA January 21st, 2010 at 9:45 am

    Yair, good points and well said. What we really should value is players who are “clutch” and who are “winners”. Reminds me of Billy Martin’s tenure as yankee player and why he was so loved by Casey during those great early 50s years.

  2. Mike Axisa January 21st, 2010 at 9:48 am

    This post completely misrepresents the statistical community. No one ever said “clutch” or “intangibles” or stuff like that didn’t exist. It does, it obviously does. However the problem is that those factors are grossly overstated in the grand scheme of things.

    Baseball teams shouldn’t be built on what we imagine them to be. There’s more data out there than we know what to do with, and if you ignore that just because you don’t agree with it, then you’re already behind the 8-ball. Do you think Fortune-500 companies make decisions on a whim?

    Stats and track records are just market research. They shouldn’t be ignored.

  3. bru January 21st, 2010 at 9:48 am

    there is no way damon would be happy taking that much of a paycut considering the role he would play as the # 2 hitter when the mitres,gaudins,etc… are making 1.5-2 million

    the only way i see damon coming back is if there are absolutely no other offers

    if the yankees offer him 4 million & another team 3.5 or 3 million i think he goes too another team

    pride & egos are a tough nut too crack

  4. Erin January 21st, 2010 at 9:49 am

    Very nice job Yair! :)

  5. m January 21st, 2010 at 9:49 am

    Agreed. I understand the value, but don’t always understand the values. :)

    Stats are important in that they’re recorded history. They can be plugged into computer models to spit out whatever you want it to to be interpreted however you want.

    And truth be told, how boring would a game be (closing in on 5 hours for some Red Sox/Yankee classics) if the screen or scoreboard didn’t flash statistics? What else will Coney talk about :( ?

    Numbers were a big pull for me as I came to appreciate baseball as a fan.

    But at some point those numbers are just that. Numbers. When evaluating a player, I think scouting and video will tell the story better. The numbers don’t tell you about hustle or athleticism. Not to say that stats are unimportant, but they only tell part of the story.

  6. Stan January 21st, 2010 at 9:51 am

    This entry is a great example of someone who doesn’t understand stats talking about people who talk about stats.

  7. m January 21st, 2010 at 9:53 am

    There are no other offers, therefore Damon will be back.

    But like I said before, if Damon’s value drops into the $4-5M range he may look more attractive to other teams.

    Even the offense-starved Red Sox. Unless Bill James recommends that shy away. :P

  8. Shame Spencer January 21st, 2010 at 9:53 am

    Wow, what a great post. Well said and completely accurate. Too often, everything in baseball relies on numbers (Joe Girardi, this means you!) and we forget that inside the uniform is a real person.

    Maybe we like to say the stats aren’t as important because we are Yankee fans and see on a daily basis what guys like Jeter and Mariano can do. The idea that a bunch of pinheads think Jeter is overrated will never stop making me mad. I almost get irrate, even when the comment is coming from some lowly Bosox fan. But I guess at the end of the day, all that matters are the right wins at the right moments.

  9. Erin January 21st, 2010 at 9:54 am

    m
    January 21st, 2010 at 9:49 am

    What else will Coney talk about?

    ***********************
    Sadly, he won’t be talking about anything-he won’t be back at YES this year. :(

  10. SJ44 January 21st, 2010 at 9:55 am

    Mike,

    My opinion is BOTH the stat and “clutch” communities overstate their positions.

    Statistical analysis does have an important role in the game.

    If it didn’t, teams wouldn’t be spending between 250K and 1 million dollars to create and operate proprietary software measuring statistical analysis for their own use.

    However, its when people use it (and misuse it) as the SOLE judge of player evaluation it loses appeal.

    Same can be said for those who use “eyeball” or “clutch” arguments when evaluating player performance.

    They are guilty of the same misuses the stat community are when misapplied. Which is done as often as those who misuse stats, IMO.

    Bottom line is, proper player evaluation involves both eyeball readings as well as statistical analysis.

    The better organizations balance both and are not wedded to one way or the other.

  11. Matt January 21st, 2010 at 9:56 am

    Re-post from previous thread …..

    It was proven the Sabathia and Teixeira were difference makers to compliment other talent the Yankees had in 2009.
    Adding Curtis Granderson, Nick Johnson, and Javy Vasquez were good and correct moves going forward to this season.
    Johnny Damon would not be a difference maker in 2010 and just 1/25th of the team if he were signed.
    Brian Cashman knew this once the budget restraints were in place. He also has to keep an eye toward 2011 when Derek Jeter’s contract is due along with possible year to year deals with Mariano among other moves.
    Scott Boras couldn’t care less about any team’s budget concerns. All he wants is the most money for his clients which equates to more commissions for him.

  12. blake January 21st, 2010 at 9:56 am

    SJ,
    Its a paycut in a sense (as far as Damon is concerned it probably is) but its not the same thing as me going into a job and them reducing my pay while I’m on an existing contract.

    He has no contract at the moment so he has no salary to reduce..

  13. Patrick January 21st, 2010 at 9:58 am

    Good thing no teams rely solely on statistics like you imply the A’s do. Every team in baseball take into account statistics and scouting when evaluating a player.

    One other point, you say,

    “After all, how likely is it that so many closers who have converted 95% of their career save opportunities somehow all choose the playoffs to have some of their worst — and statistically unlikely — outings? Something seems to be up.”

    That something is that those closers are facing the best teams every day in the playoffs. I think it’s pretty obvious why their % of blown saves goes up in those situations.

    Only the best players in AAA make it to the majors and only the best teams make it to the playoffs. Yet it’s no mystery when really great AAA pitchers blow up when they get to the majors.

  14. bru January 21st, 2010 at 9:59 am

    girrardi just won the ws & on several occasions he has talked about heart to heart conversations that he has had with numerous players

    we do not know that girrardi is only about numbers & just because he is a numbers freak means absolutely nothing

    if he was only about numbers in 2009 & delivers the same results i’ll take it every time but i seriously doubt that is the case

  15. SJ44 January 21st, 2010 at 10:01 am

    Blake,

    Let’s be real here.

    Johnny Damon will see it as a paycut. So will any logical thinking person.

    He would have the same role on the same team. Oh by the way, he had a great year last year.

    2009: He made 13 million. 2010: You may make 2-4 million.

    If you don’t think he will see that (or anybody else for that matter) as a paycut from the Yankees, you don’t get ballplayers.

    Its a paycut and a MASSIVE one. Often, too massive for guys to take and stay with their current team.

  16. m January 21st, 2010 at 10:01 am

    Girardi likes to consult his “book”, but I think as he gains more experience he’ll rely more on gut instincts.

    Sometimes his moves worked, sometimes they didn’t.

    Just like sometimes a manager’s gut instinct is right, sometimes it isn’t.

  17. Y's Guy January 21st, 2010 at 10:02 am

    there are 2 parts to statistical analysis. statistics don’t lie but the analysis can often be badly flawed. statheads tend to forget the second part of that.

  18. Yair January 21st, 2010 at 10:03 am

    Mike,

    “No one ever said “clutch” or “intangibles” or stuff like that didn’t exist. It does, it obviously does. However the problem is that those factors are grossly overstated in the grand scheme of things.”

    Many have said such factors don’t exist. And certainly many on both sides of the question have far overstated their cases. Some sabermetrics folks have begun to try to quantify “clutch” (Nate Silver has a great article in Baseball Prospectus). My point was that that school of thought is on the right track, and it makes intuitive sense that “clutch” should exist in some form.

    I’m interested in explaining a plausible narrative to WHY there might be “clutch” hidden within statistical noise. WHY we should continue looking. Why someone like Silver might be right – that there are a select few clutch players.

    I’m not denying that it’s very hard to figure out, nor that we’ll probably find very few players who fit the category. I do not think current statistics are wrong in any way. I’m saying they sometimes mislead people into thinking we know all there is to know, OBSCURING the human factor out our considerations as to what stats to investigate next.

    I have no beef with sabermetrics – I just wanted to weigh in on the side which says “clutch” (and performance under pressure) might very well exist.

    Sorry if I was unclear!

  19. hjcho January 21st, 2010 at 10:05 am

    Remember that episode of classic Trek where the two warring countries don’t bother to actually attack each other, they just run computer simulations and the “casualties” report to killing stations to be dispensed? Sometimes it seems like the stat-heads are trying to reduce baseball to the same level. It’s what the stats DON’T predict that make the game fun. At least to watch.

  20. Y's Guy January 21st, 2010 at 10:05 am

    “Do you think Fortune-500 companies make decisions on a whim?”

    you’re joking right? considering the performance of fortune 500 companies over the past couple of years, you might want to rethink holding thier statistical analysis up as a beacon of the right way to do things.

  21. Doreen - Ain't it Just "Grand"? January 21st, 2010 at 10:05 am

    Well-written Yair.

    I think it is most fair to say that each without the other is incomplete when evaluating players.

    Erica -

    Are you here?

    I’ll do it.

  22. blake January 21st, 2010 at 10:09 am

    SJ,
    I’m don’t disagree with what you are saying but when the Yankees agreed to pay Johnny 13 million for 2009 he was 4 years younger. If he had been a free agent last year he wouldn’t have got 13 million in 2009 either.

    When you you have no other options don’t you have to adjust your expectations a bit. Totally agree with what you’ve been saying about it being more difficult for a player to take a lesser salary from their old team. I’m just hoping Damon decides he can accept it and they can work something out where they both will be happy.

  23. Erica - always OPPC - Bring Back Johnny!!!! January 21st, 2010 at 10:12 am

    Doreen-

    I am here- semi-paying attenion.

    Are you sure???

    It may be best for us to get in touch off the blog. Is there a way I can reach out to you? Like facebook or something if you don’t want to put your e-mail in here

  24. Y's Guy January 21st, 2010 at 10:14 am

    if statistical analysis applied to sports is so accurate, then why doesn’t some stat geek work out a foolproof program to pick the winners or horse races or football games?

    because the analysis side of statistical analysis is nowhere near as accurate as the statheads would have you think.

  25. tex's friend January 21st, 2010 at 10:20 am

    Its a paycut and a MASSIVE one. Often, too massive for guys to take and stay with their current team.

    ____

    Johnny got that $52M in a better economy with more competition for his services. Abreu had a good year too before he went from $16M to $5M and now he has a good 2 year contract again. $13M was way too much to begin with so bringing it back to 4-5 M is where it should be.

    Now on the flip side, he knows Jeter will be making 15-20M/year on his next contract so his pride will play a part.

    But 2,3,4,5 Million is still a lot of money. Beats the $0 he is working with now.

  26. Captain Chaos January 21st, 2010 at 10:21 am

    Nicely done!!!! I agree statistics only provide the central tendencies of the player, team… in a linear fashion. However, the world certainly doesn’t work in a linear fashion (enter the tipping point). Certainly the “intangibles,” maybe someone can define that term for me (Chad???) influence, or maybe that definition is held like a trade secret by the top flight talent scouts. Finally, baseball is statistically defined (era, BA, WP…), why is it a reach that some take this fascination to the extreme…Maybe that is why it is America’s favorite pass time…Chaos!!!

  27. Doreen - Ain't it Just "Grand"? January 21st, 2010 at 10:21 am

    Erica -

    I am sure.

    I played with a couple of lineups from last season. I think I have a system down that will work for me, but you are the master, and any hints would be appreciated. :)

    I am on facebook: facebook.com/dorgal218

    It’ll be fun. ROFL

  28. ko January 21st, 2010 at 10:22 am

    If you look at statistics as a starting point for a conversation rather than the end all/be all, they will be a useful tool. If you look at them as a conversation ender, then you are the tool. All you have to do is look at weather forecasts to get a good idea of how inaccurate our statistical models are. They are approximations, in their best case. In baseball, fielding is the most foolish area to use statistics as any true measurement of what’s going on. You need to believe your eyes on that one. By the time you’re done with your simplifying assumptions to put a statistical model together on fielding, you’ve assumed away the measurement. Bill James (who works for the Red Sox, by the way) has been hammering away at Jeter for years. Guess what? – Boston would sign Jeter in a heartbeat if they could. A relevant Yankee example of misleading statistics are the Nick Swisher and Curtis Granderson stats. Altho’ their overall stats appear to make them solid big league hitters who you could stick into the middle of a lineup – you can’t. Swisher can’t hit a breaking pitch and Granderson can’t hit lefties. You put them in the middle of a lineup and you’ve got two automatic outs in close games in the late innings because all you need to do with Swisher is bring in a breaking ball pitcher who can throw them for strikes or a lefthander with Granderson. Like I said, statistics are a converstation starter, not ender

  29. Joe from Long Island January 21st, 2010 at 10:23 am

    1. Very nice post by Yair. I think the stats have to be balanced by the particulars in play in any one instance, but they can certainly open your eyes to some things.

    2. Very interesting discussion since last night about Johnny Damon. Personally, I think at this point he represents just one of the options being considered for OF help. At this point, him coming back will depend if they have any better options available, either thru trade or FA. That is, if Damon is willing to deal with the baggage of coming back to Yanks for the big cut in pay.

    Johnny and Boras missed the boat on this one – very badly. I am surprised. I thought Boras was smarter than that. There certainly seemed to be ample warning signs, going back to Abreu’s situation last year.

  30. Erica - always OPPC - Bring Back Johnny!!!! January 21st, 2010 at 10:24 am

    Doreen – Ain’t it Just “Grand”?
    January 21st, 2010 at 10:21 am
    Erica -

    I am sure.

    I played with a couple of lineups from last season. I think I have a system down that will work for me, but you are the master, and any hints would be appreciated.

    I am on facebook: facebook.com/dorgal218

    It’ll be fun. ROFL

    ************

    Awesome. I will get in touch with you tonight then. I can’t gte on facebook at work, so I will request you later!

  31. SJ44 January 21st, 2010 at 10:24 am

    Doesn’t beat the 0 when you don’t have to work if you choose not to.

    You can analogize athletes and “regular” human beings in the workplace.

    Johnny Damon, if he chooses to, can say, “I’ll stay home with the kids” and call it a day if he chooses not to take a 60+% paycut to play this year.

    He’s already earned 100 million in the game. Just bought another new house in an all cash sale in Orlando.

    He has no pressing financial needs. Meaning, he will either play for a price and team he is comfortable with or he won’t.

    Its different from the real world workplace where most people have to take what’s offered to put food on the table.

    Johnny has no such pressures to do that in his present situation.

  32. m January 21st, 2010 at 10:27 am

    This has all the makings for a good Disney Channel movie. Caveman goes to Japan to play baseball for a year and ends up being a big hero!

  33. Jerkface January 21st, 2010 at 10:27 am

    if statistical analysis applied to sports is so accurate, then why doesn’t some stat geek work out a foolproof program to pick the winners or horse races or football games?

    This is silly.

  34. Jerkface January 21st, 2010 at 10:28 am

    Last I checked, UZR for horses was still in BETA…

  35. Wave Your Hat January 21st, 2010 at 10:29 am

    I have to go with Mike Axisa’s criticism of the guest post.

    On the Damon issue, I still think the Yanks ought to try to sign him if his price falls below $6MM. I’m perfectly happy if the Yanks get him for less – I don’t really care how much they pay him.

    At that point it’s up to Damon, he could sign with the Yanks, sign with another team for less, or retire. The Yanks can’t control for that.

    But I still say, if he does sign a one year deal for less than $6MM, and it’s not with the Yanks and the Yanks didn’t make him an offer in that range, the Yanks will have screwed up. If the budget is so tight, trade Gaudin to dump some salary.

    But Damon is still a very good player, his defense is likely not to be as bad as many peoples seem to think, and between the risk of Granderson not hitting lefties and/or Gardner not hitting well enough at all, and/or the general risk that the Yanks don’t play as well in 2010 as the projections seem to think, or the Sox or Rays or both play better, the Yanks can use him.

  36. lordbyron January 21st, 2010 at 10:30 am

    I couldn’t agree more – well said Yair.

  37. Frank January 21st, 2010 at 10:32 am

    Harvard, eh? Couldn’t get into BC? 8)

    Nice piece. I lean more to the stat geek side of the argument so far as determining a player’s value, but see the importance in the “what makes this guy tick?” factor as well.

  38. Doreen - Ain't it Just "Grand"? January 21st, 2010 at 10:32 am

    Great Erica. I look forward to hearing from you later. :)

    Joe from LI

    About Boras and not being able to read this market. I am surprised as well. He has not changed his definition of “the best deal.” Also, I think he wasn’t listening to the same things most of us were hearing, apparently – it seemed like every time I listened to the sportstalk on XM and they got around to talking about Damon, the conversation pretty much focused on how perfect a match Damon was for the Yankees and vice/versa AND it seemed that other teams thought this as well. I think rightly or wrongly an awful lot of people out there seemed to equate Damon’s successful year with his fit to Yankee Stadium and possibly had questions about how he’d adjust to other parks, especially larger parks. Or maybe a few of them simply figured he’d end up with the Yankees anyway, so they focused their attention elsewhere.

    Clearly, no one bought into Boras’ utterly ridiculous statement about Johnny’s superior genetics. ;)

  39. David January 21st, 2010 at 10:32 am

    Yair, while I agree with you completely that sabermetricians ignore aspects of the game that obviously exist, I think you present a few false contrasts. Specifically:

    “a modern talent evaluator like Billy Beane looks at advanced metrics and finds the right players to draft without ever observing them in person. A player’s performance can thus be predicted, fantasy baseball style, without reference to anything but the numbers.”

    I’m not sure whether this is just an issue of semantics, but the “modern talent evaluator” is still a scout, not a statistician. The view of MLB that Moneyball presented was never the reality, even in the A’s office. Teams make decisions on both scouting and stats, and if they ignore either they do so at their own peril.

    Part of being a good statistician is knowing what a particular number doesn’t say, not just what it does say. Sabermetricians who overstep their boundaries are doing so either to prove a point (e.g. saying that pitchers don’t have any control over defense, just to prove that they in fact have limited control) or simply because they are wrong.

    But I commend you on making an intelligent point that is too often made unintelligibly. Most people just say that “stats don’t tell the whole story” because they hold a belief that they’re too stubborn to change. This article was a nice change of pace from that.

  40. Erin January 21st, 2010 at 10:32 am

    m
    January 21st, 2010 at 10:27 am
    This has all the makings for a good Disney Channel movie. Caveman goes to Japan to play baseball for a year and ends up being a big hero!

    **************************
    LMAO

  41. Jerkface January 21st, 2010 at 10:33 am

    Football doesnt have the same kind of stat impact as baseball. Football has far more factors going into any given play than the pitcher-batter matchup. Also there is tackling.

    The statistical analysis side of football is more about figuring out who the best DBs are (and its almost always accurate, Nmadi is always #1 as he should be) and the different conditions wide receivers catch balls or whatever.

    BASKETBALL has really awesome statistical analysis, they track everything and you can get a detailed report on a players scoring. How he scores going to the left, to the right, off the dribble, off a pass, shots from everywhere on the arc, and if you have a player who is smart enough it will work very effectively.

    There is a really good article about Shane Battier who is one of the only defenders smart enough to get all the data (most just get basics). He used it to force Kobe into his least efficient shooting lanes and Kobe scored a lot of points, but did it on far more attempts than normal (which means he missed more and took more shots from his teammate, a victory in a game like basketball where its hard to stop teams from scoring completely).

  42. stanzy January 21st, 2010 at 10:33 am

    Good post, Yair. I don’t really get why there has to be so much tension between the (supposed) two camps. It would seem to me that both things, a player’s metrics and his makeup/clutchness/whatever, are different ways of seeing the same thing: how good a player is. If a player is prone to “phoning it in” or being too sensitive to pressure, it’s going to show up in his stats. Sure there are players who will perform better in low pressure games/situations than in higher pressure ones. But usually the higher pressure comes against better teams or better players. Those who can perform equally as well under both sets of circumstances are “clutch” or have “the right makeup,” which will also show up in their stats.

    Some people have a better natural (or developed from experience) ability to identify good ballplayers or know in their gut who to use in what situation. Those people are likely unconsciously incorporating much of what the metrics would tell them, when they make their decisions. The stats give those with less natural ability a way to be better at evaluating talent or making tactical decisions. It’s similar to how some people can become excellent piano players without years of formal training, while others can also become excellent players but only after learning and understanding music from a more technical perspective and then applying what they’ve learned through training.

  43. m January 21st, 2010 at 10:33 am

    What’s this tipping point everyone speaks of?

    How about a tipping point for the Yankees? How low does Damon have to go?

  44. Jerkface January 21st, 2010 at 10:36 am

    By contrast, a modern talent evaluator like Billy Beane looks at advanced metrics and finds the right players to draft without ever observing them in person. A player’s performance can thus be predicted, fantasy baseball style, without reference to anything but the numbers.

    GOing to backup David on this one. Billy Beane uses stats to find players to then scout. His scouts in the book all said the players he wanted ‘didn’t look like baseball players’, and he then drafted a few of them anyways.

    They still scouted them. No one picks a guy to spend a million dollars on without seeing them. Thats just silly.

  45. Patrick January 21st, 2010 at 10:42 am

    Sabermetrics are dumb!

    There’s no stat for grits, bozos

  46. gregori January 21st, 2010 at 10:43 am

    always have had the same opinion regarding sabermetrics.
    statistics are wonderful tools but should never be taken out of their context and should always be viewed with the perspective of their origins.
    all too often investigators tend to overstate statistical meanings when applied to biological systems.
    statistics are governed by mathematical rules rules that more easily explain inanimate systems.
    physics, chemistry and the like are realms in which mathematics rule as the ‘rules’ are indisputable.
    in biological systems their are no hard and fast rules.
    a lethal dose of a particular poison may vary by hundreds of percent when acting on a particular individual of a species.
    but statistics will pick a number, the LD50, that will be lethal to 50% of individuals.
    this is useful for research purposes but not as useful in predicting or evaluating individual reactions to this poison.
    this is analogous to base ball and the statistics therein.
    sabermetrics are research fine tools in recognizing trends of large numbers of players to reveal what may be winning, losing, average or above average tendencies.
    sabermetrics however are inaccurate and occasionally misleading in evaluating individual performance, tendencies and predictions.
    as with everything in life, there is never any one encompassing theory that explains everything.
    rather one must take the best from all the competing theories and blend them into something more accurate and useful.
    of course the difficulties arise when we attempt to decide what are the best aspects each theory.
    that’s why we continue to discuss and research.

  47. Jerkface January 21st, 2010 at 10:43 am

    Sabermetrics are dumb!

    There’s no stat for grits, bozos

    How does anyone buy sand paper?!

  48. m January 21st, 2010 at 10:44 am

    Erin,

    Titled “From $0 to Hero”

    Just kidding. Even though I shouldn’t feel sorry for a ball player that’s earned more than the GDP of some countries and made some rather ill-advised demands, I do. A little.

    He’s one of the better players on the market, but he’s still the last kid to be picked.

    All in all, though, this comes down to Damon’s true value than anything he or Boras did.

    His age, questions of his durability, NYS splits, and his defense are costing him.

  49. Patrick January 21st, 2010 at 10:45 am

    “How does anyone buy sand paper?!”

    Well personally I go to the hardware store and scout the different brands making sure to mark down any intangibles I see.

  50. David January 21st, 2010 at 10:48 am

    One more thing to add, more tangential to the post than an argument against it, but I think it’s kind of cool:

    Think about a projection system like PECOTA that relies entirely on statistics. It uses the career paths of past comparable players to make its projections. But doesn’t that mean it includes a player’s psychological makeup, to a degree?

    Consider a prospect that has started out really well statistically, but has caused doubts as to whether he has the makeup to succeed in the majors with more pressure on him. PECOTA projects a variety of possible career paths for this player. Some of these paths are probably based on players who turned out not to have the makeup, and so he’ll be penalized for that. This isn’t just hypothetical: this is part of the reason that projections for prospects diverge so much from their 90th percentile to their 10th.

    Someone well-versed in both statistics and scouting could identify their own opinion of the player’s makeup, and then have a better idea of which projected career path the player would follow.

  51. Yair January 21st, 2010 at 10:50 am

    David and JF,

    You’re both right. I intended that “fantasy baseball style” draft portrait as a caricature given from the perspective of the person overly convinced of our current stats’ ability to utterly totalize a player. Billy Beane didn’t ACTUALLY do that (I certainly hope not!). Yet another point I should have been more clear on.

    And PECOTA is just the sort of statistical measure I think we could use more of.

  52. Benny January 21st, 2010 at 10:55 am

    If push comes to shove, Damon can grow a beard again and go to work for Geico doing commercials as a Neanderthal man.
    He doesn’t need Boras to communicate with Madison Avenue.

  53. Erin January 21st, 2010 at 10:56 am

    m
    January 21st, 2010 at 10:44 am
    Erin,

    Titled “From $0 to Hero”

    **********************
    LOL

    You forgot to add :The Johnny Damon Story. ;)

  54. Guest January 21st, 2010 at 10:58 am

    I appreciate the thorough post, Yair, but I have to disagree.

    Of course a player’s ability to deal with stress will deeply impact his performance. Of course there are some players who will mail it in the dog days of August and others who will give 100% in every inning of every game of their entire careers. Of course there are some players who react better to adversity during the course of the game than others.

    BUT here’s the main point: the numbers a player puts up is the product of all of these factors. If a guy would have been a .350 hitter if he brought it every day, but he doesn’t so he’s only a .300 hitter, well, he’s just a .300 hitter. It doesn’t matter why he’s a .300 hitter, the point is, he’s a .300 hitter. And the numbers tell us that.

    More importantly, if we have a large enough sample size showing that, for whatever reason, he’s a .300 hitter who is not likely to decline due to age or injury; we will likely be correct to predict that he will hit approximately .300 the next year. (Better numbers, like BABIP, OPS+, Park-adjusted numbers, etc, tell us even more information and are more useful in determining likely future performance).

    Contrary to popular opinion, many numbers people do actually believe in the existence of intangibles. The whole point is that numbers are a better predictive tool than anything else because the numbers are a product of a players talents AND intangibles. The effect of the intangibles are already factored in. Sometimes things become cliche because they are true, like the fact that numbers don’t lie.

  55. Yair January 21st, 2010 at 11:03 am

    I don’t think we really disagree. I’m simply arguing there’s still benefit to keeping the concepts of “clutch” and “intangibles” in mind and drafting, if at all possible, in light of them as well as current statistical measures.

    Because wouldn’t one have an advantage if one could spot – either through experienced scouting or ideal metrics – a closer who won’t break down (or will do better than his usual) in the playoffs? If one could separate the Joe Nathans and the Billy Wagners from the Mariano Riveras? If you don’t admit “clutch” exists or have given up on quantifying it, you lose this human – and important – aspect of player selection.

  56. Wave Your Hat January 21st, 2010 at 11:07 am

    “Because wouldn’t one have an advantage if one could spot – either through experienced scouting or ideal metrics – a closer who won’t break down (or will do better than his usual) in the playoffs? If one could separate the Joe Nathans and the Billy Wagners from the Mariano Riveras? If you don’t admit “clutch” exists or have given up on quantifying it, you lose this human – and important – aspect of player selection.”

    You are begging the question. You don’t know whether the ability you are presuming to spot through character analysis even exists, much less whether if it exists it is capable of being spotted by the means you would like to use.

    If you assume away the issues, the answers do get easier, I admit.

  57. raymagnetic January 21st, 2010 at 11:08 am

    “That something is that those closers are facing the best teams every day in the playoffs. I think it’s pretty obvious why their % of blown saves goes up in those situations.”

    What’s Mariano’s excuse for being better in the postseason than he is in the regular season? You’re crazy if you think nerves don’t play a HUGE role when it comes to closers in the playoffs.

    Or is Mariano simply an outlier, performing with the hammer of God in the postseason?

    In my opinion some players do perform better when the pressure is on and some shrink under the pressure.

  58. upstate kate January 21st, 2010 at 11:09 am

    If I am understanding some of the comments correctly, then intangibles and clutchiness are reflected in statistics.
    If a player chooses to watch the ball he hits instead of running right out of the box, and it turns out not to be a HR, then he gets a single instead of a double. If a player steals a second base b/c no one is covering, he adds to his SB total. If a player runs hard on every play, he has the opportunity to score on a dropped ball, or a throwing error.

  59. Yair January 21st, 2010 at 11:17 am

    Many of these human factors ARE reflected in the statistics we have. We just don’t know how to ISOLATE them (as I write in my final paragraphs). Once you admit to “clutch”, my argument is simply that it’s in the team’s best interest to try their best to identify the players who perform best under pressure.

    Wave Your Hat – As I wrote earlier, I’m with the school of sabermetrics that thinks clutch does exist (e.g. Nate Silver), and am offering a reason for why they’re on the right track while those who deny clutch or think it’s not worth identifying (since our stats already tell us all we need to know) are missing out on an important – if small – aspect of the game.

  60. Jonas January 21st, 2010 at 11:17 am

    you should read “The Book.” This will address a lot of these intangible questions.

  61. Tom Swift January 21st, 2010 at 11:18 am

    This post vacillates between saying that there are factors for which there are currently no metrics but perhaps one day there could be and that there are factors that are inherently unmeasurable. The former position would surely be consistent with the statistical analysis community’s approach; they are constantly looking for new things to measure. As for the latter position, who could argue with that? The problem is that if there is no metric, it is difficult to come up with a rational basis for decision-making. But we all know that there is something special about guys like Jeter, apart from the stats, that makes them good for the team and makes the team a winner.

  62. mick January 21st, 2010 at 11:19 am

    Most baseball fans could care less about the fancy statistics that have become a cottage industry. Feel free to pursue this scientific endeavour but know that it will never become an integral part of the game.

  63. Jerkface January 21st, 2010 at 11:26 am

    but know that it will never become an integral part of the game.

    People with this attitude are in for a surprise

  64. Wave Your Hat January 21st, 2010 at 11:26 am

    Yair-

    Without statistical analysis, it would be impossible to determine whether such a thing as “clutch” exists.

    That’s because if it exists, it ought to be measurable, i.e., not intangible which by strict definition is not measurable. And if it is measurable, then it is worthwhile putting some effort into discovering the best way to measure it, in the hopes of being able to predict its future path.

    IF it exists, and IF some scouts can detect it and predict its future path, and IF they really have detected it and predicted it and aren’t just lucky sometimes (which is a possibility), then “clutchiness” should be susceptible to measurement, ergo to statistical analysis which would to a greater or lesser degree eliminate much of the human error and bias that all human judges bring to what they are judging.

    And that would be a good thing, wouldn’t it?

  65. pete January 21st, 2010 at 11:26 am

    I agree with Mike Axisa. Not that your piece is wholly, or even partially unreadable or anything like that, and it attempts to hold both alleged “sides” of this alleged “argument” to equal reputation. But the piece is woefully misrepresentative and likely misinformed as well. No team dismisses scouting, and obviously Billy Beane nor any other GM has the time to look at every draft prospect. But nobody in the history of the draft has been drafted without being looked at, at least a little, by at least a couple of major league scouts, and anybody taken in the first 15 rounds or so has most likely had several scouts at every game he’s played the last couple years.

    Furthermore, sabermetrics do not dismiss things like intangibles or clutchiness. They simply don’t try to predict them, and their followers try to avoid being trapped by media-driven narratives that find most of their roots in “anonymous sources” or simply writers’ opinions. While I personally doubt that anyone can truly improve upon their abilities in clutch situations (unless they are woefully lazy or simply too bored with the game of baseball to concentrate in any but the most exciting of situations, which would likely weed them out in the early minors), but I certainly believe that players can be negatively affected by pressure, especially pitchers. I also believe, though, that in general, a player who makes it all the way to the majors, and succeeds at the major league level enough for expectations to be leveled upon him, will, in a large enough sample, play at roughly the same level in high-stress situations as regular situations.

    The reigning argument in the anti-sabermetrics camp seems to be that sabermetricians and advanced statistics advocates are trying to “change the game of baseball” and trying to eliminate the “human” element. This is simply not true. For one thing, sabermetrics and advanced statistics began changing the game, slowly but surely, about 20 years ago. There are still a few holdouts – the Reds come to mind as a prominent example, but for the most part, teams have realized that their are more financially efficient ways to brew success than the traditional norms that depend on spending on guys who hit for high averages or pitchers with high win totals. There is a constant push between teams that have their own statisticians whose work is probably a few years ahead of anything we’ve got here in the blogosphere, even with the superb work over at fangraphs. As data and technology accumulates, front offices are able to apply algorithms to the evaluation process that can better suggest how much players contribute, and thus are able to underspend for production. This has been the convention of MLB teams since the late 80s at the very least, so please don’t try to sell the “they’re ruining baseball argument”.

    What sabermetricians are, in fact, trying to do, is change the way fans view the game, and, perhaps more importantly, the way writers write about it. The thing about baseball (along with all sports, really) is that wayyy more teams exist than can actually compete for a title every year. What keeps fans coming back to teams that lose is the unique attachment to players that baseball allows for. In baseball, unlike most other sports, one player cannot turn a bad team into a good team, no matter how good that player is. But in baseball, unlike other sports, it is easy to follow a single player, rooting for his success in the wake of his team’s failure. Sabermetrics really only expand that, and in reality, will eventually allow teams to pursue success with “boring” players who walk a lot and don’t run the bases well, or who defend really well but can’t hit too well, et cetera, et cetera, because while it’s unlikely that fans are ever going to be indifferent to their favorite player hitting .350, or hitting 40 hrs, or driving in 150, or winning 20 games, or whatever other statistical goals fans may have for their favorite players now, in the future more fans will get excited over a wRC+ of 140, regardless of how that number is approached. Fans will see a 10-10, 4.50 pitcher’s 2.95 FIP and get excited for the next season, get excited about being the guy who predicted a breakout year. Because that, above all, is the mission of sabermetrics – to find more accurate and more production-weighted methods of projection for the future. There are many ways to observe the past, which is essentially what every single statistic does, but some are much more reliable for predicting the future than others.

    Now here is where I hear the “human” side of the argument – if we get to the point where we can reasonably predict a team’s win total, barring extreme aberration or extraordinarily bad luck, to within a few games (which of course we can’t right now, but it isn’t out of the question for the near future), then why bother watching? Well it’s simple – the season is so long, that all of the random statistical anomalies that occur during the season (and there are plenty) provide enough fluctuation that the game does indeed remain fun to watch, and wholly unpredictable on a micro level, but over the course of 162 games, these anomalies experience corrections that bring the overall statistics to their normal levels. Plus, once you get to the playoffs, you can take most statistics and all but throw them out the window, since the reduced sample impedes the correctional process. THIS is what makes baseball such a beautiful sport – you can build a team so dominant that it wins its division 10 years in a row, and it’s highly unlikely that in those 10 postseasons it wins more than one or two world series.

    In summary, nobody is taking the human element out of baseball. In fact, nobody is even ignoring it. The whole reason all of us watch the games is because we DO get emotionally entrenched in the relative statistical insignificance of minor events – home runs, strikeouts, wins losses, winning streaks, losing streaks. When it comes to analysis and roster construction, advanced statistics are incredibly useful, since they relay what happens far better than our biased memories could ever do. But when it comes to watching the games and following a team, we will forever be slaves to the irrational, caught up entirely in the moment, completely indifferent to its true relevance on a macro level. Because if we were never irrational, and all we cared about was practicality, why would any of us ever care about professional baseball to begin with?

  66. Wave Your Hat January 21st, 2010 at 11:29 am

    “But we all know that there is something special about guys like Jeter, apart from the stats, that makes them good for the team and makes the team a winner.”

    This is interesting. Suppose we could build a “robot” Jeter, who could perform to all of Jeter’s stat lines but of whom we could be sure there is no “inner spark”. How sure can you be that the Yanks wouldn’t succeed just as well with the robot Jeter?

  67. Yair January 21st, 2010 at 11:36 am

    Wave Your Hat – We’re not disagreeing! Read my final paragraphs and my comments. I think we should definitely work on such measures, and that would be a good thing. But you have to admit that “clutch” exists and make plausible sense to go there. Some have argued this from a statistical perspective (Nate Silver), while I do it from a psychological one.

    Pete – Thanks for taking the time to write such a detailed response. I don’t think we’re not really disagreeing either. As I explained in the comments, Mike seems to have misread my post and its intent. It’s not an attack on sabermetrics. Other commenters have noted this – my post is just a reminder that the human element is not a myth, and that sabermetrics hasn’t yet fully quantified it. That our current statistics are a great start, but not yet finished. I was pushing back on those who think our current stats totalize the players and the game. (And the Billy Beane thing was a caricature, not a historical anecdote!)

  68. JP January 21st, 2010 at 11:36 am

    Nice job, Yair. I rarely comment here but I gotta say that was very well done.

  69. David January 21st, 2010 at 11:38 am

    Such an over-reliance on statistics obscures a genuine understanding of the game of Baseball. Well done, Yair.

  70. m January 21st, 2010 at 11:38 am

    •Yankees pitcher Javier Vazquez spoke to Puerto Rican newspaper La Perla del Sur. MLBTR’s translator Nick Collias supplies this interesting quote: “I don’t have much playing time left…I go year by year, and I don’t know if it will be one, two or three years, but I’m definitely not going to play until 40.” Vazquez is eligible for free agency after the season.
    ============================================

    How can you quantify clutch? I mean there’s a stat for late innings, RISP average, and stuff like that. But clutch is more of a reputation thing.

    And I guess Jeter is one of the best examples. We’ve often read or heard that Jeter is the last person a team wants at the plate in a late or crucial situation. Of course, A-rod or Posada or Mauer is more likely to hit a homerun. But Jeter has a reputation built on years of coming through in crucial situations. I don’t think you can quanitify clutch. And what’s the need to do so? You have a person’s body of work to tell you how they’ll most likely do in a situation.

  71. m January 21st, 2010 at 11:43 am

    How can a computer model differentiate between a crucial situation or not?

    Your team is down 6-0 in the first inning. Bases loaded, one out. At that moment and in that context this is a crucial at-bat. It could very likely set the tone for the rest of the game.

    I suppose it could be done, trying to quantify or give every situation a value. But it seems like a lot of trouble to go through.

  72. Thomo January 21st, 2010 at 11:47 am

    Yair,

    congratulations on being selected as one of the “pinch-hitters” for the blog. unfortunately while you have apparently tried to even-handed in your analysis you have overlooked or perhaps are unaware of several important elements. I won’t repeat what others have already written – I will just add that you imply that intangible (unknowns) can’t be dealt with statistics. In fact, there are entire branches of mathematics that deal with uncertainty and ambiguity. The real challenge is being able to synthesize zettabytes of data in an understandable way. since you are in Boston – try and visit the Broad Institute – they are leaders in mathematical biology (and as someone else in the blog alluded to, cells are way less complex than humans!).

  73. Ed January 21st, 2010 at 11:50 am

    I think that most successful teams blend scouting with statistical analysis. Does it have to be one or the other?

  74. Jerkface January 21st, 2010 at 11:51 am

    And I guess Jeter is one of the best examples. We’ve often read or heard that Jeter is the last person a team wants at the plate in a late or crucial situation. Of course, A-rod or Posada or Mauer is more likely to hit a homerun. But Jeter has a reputation built on years of coming through in crucial situations. I don’t think you can quanitify clutch. And what’s the need to do so? You have a person’s body of work to tell you how they’ll most likely do in a situation.

    The funny thing about this is that Jeter doesn’t always come through, or come through more often than any other player. His career average with RISP is lower than his career total average. His career late and close average is lower! He is best with 0 outs, worst with 2 outs.

    Still, all of his ‘clutch numbers’ hover around his career average, which tends to validate the supposition that as samples get larger, the player will post numbers similiar to his career averages in that situation. (regression to the mean)

  75. raymagnetic January 21st, 2010 at 11:55 am

    A person’s desire and heart can never be measured accurately. It’s not quantifiable.

    You can’t measure a guy’s will to win.

  76. Jerkface January 21st, 2010 at 11:56 am

    You can’t measure a guy’s will to win.

    But you can measure the result of said will :)
    What good is knowing a player wants to win *so bad* if he sucks?

  77. Bronx Jeers January 21st, 2010 at 11:57 am

    While the post was certainly well written, it’s not exactly overturning any new stones in here.

    Sometimes this argument reminds me of the Scopes monkey trial.

    What was it Spencer Tracy said in “Inherit The Wind”?

    “It’s a good book but it’s not the only book”

  78. Yawn January 21st, 2010 at 12:01 pm

    Great post, knocking down a non-existant strawman to make yourself seem intellectually superior. Well done. I am in awe.

  79. m January 21st, 2010 at 12:04 pm

    Jerkface,

    And yet, who would you say is more clutch than Jeter? Based on memory, Manny & Ortiz used to get big hits. Matsui has come through at times. Posada. Difficult to rank them, but I guess that’s something writers could debate.

    I’ve seen this debate in basketball. Lebron v. Kobe. Kobe v. Jordan. The thing about that is opportunity. They’ve had inordinate amounts of opportunities (compared to other players or teammates) to take the last shot as the “go-to” guy. But if I needed someone to take a 3 at the buzzer? I’d go with Jordan Farmar. If the ball’s in his hand at the end of a quarter, it’s going in. He just doesn’t get the ball at the end of the game because they’re going to live and die at the hand of Kobe.

    But baseball is different. Often the biggest hitters don’t get the opportunity because the opposing pitcher or manager can take the situation out of the hitter’s hand. Leading to fewer opportunities.

    I suppose Jeter’s reputation is bigger than it deserves to be, and that rubs a lot of people the wrong way.

    But, hey, too bad. He deserves a lot of the praises he gets.

  80. raymagnetic January 21st, 2010 at 12:04 pm

    “But you can measure the result of said will
    What good is knowing a player wants to win *so bad* if he sucks?”

    Of course talent matters first and foremost. However there are many talented people who aren’t successful. There are also people with average talent who are highly successful.

    Take players like Dwight Gooden and Daryl Strawberry. Two of the most talented individuals who ever played the game and yet none will be a HOF. What made them less successful than other players with lesser talents?

    When talented individuals have the will and desire to be great then that’s what separates them from everyone else in my opinion.

  81. Wave Your Hat January 21st, 2010 at 12:05 pm

    raymagnetic-

    Let’s say you are correct, that heart and desire can’t be measured.

    Let’s also say that your unstated premise, that is, that having more heart and desire (although how you can have “more” of something not measurable is problematic, to say the least) leads to better performance, is also correct.

    The 600 or so people who are major league baseball players are not a random sample of people – they are a very atypical collection extremely successful, highly talented, highly competitive and highly motivated people, which led to their inclusion in the group in the first place.

    How do you know, even as I said assuming your assumptions are correct, that having more heart and desire will lead to better results in competition with the other members of that very select group? You don’t, and since those characteristics aren’t measurable per your premise you will never know.

  82. rconn23 January 21st, 2010 at 12:05 pm

    Like Mike, I think this post completely misrepresents the statistical community.

    Brian Cashman and Theo Epstein, two of the best minds in the game don’t ignore the “eyeball” scouting of players.

    However, they aren’t going to drop $3 million on a player solely because a scout has a gut feeling about a guy.
    They are going to look at the numbers as any responsible, coherent GM should.

    The best orgnaizations have a healthy blend of scouting and statistical analysis.

    Scouts and stats, seperately, can’t tell the tale of a player’s value.

    There was a scout quoted in one of the D.C papers early in the season last year who said that Julian Tavarez was a “bulldog” who has the ability to be a top closer.

    Obviously, that scout is horribly misguided, not only because the stats don’t bear that out, but because he’s drifted from team to team in what have largely been mop-up roles.

    And for the record, I don’t believe in clutch and a pitcher’s ability to magically win games.

    Case in point, look at Randy Johnson’s 2004 season with the Diamondbacks. He was 16-14 with a 2.60 ERA, a .90 WHIP and a .197 batting average against.

    In 2001 and 2002, he won 21 and 24 games with similar peripherals – in some cases worse. Why? Because the team’s he pitched for were much better teams. Not because he magically forgot how to win games.

    “Clutch numbers” for hitters also vary greatly. A guy who hits .310 with RISP, is just as likely to hit .260 the next year because of a number of factors, including luck of BABIP.

  83. Guest January 21st, 2010 at 12:09 pm

    Yair: three points.

    1. Brad Lidge. He has had some of the most dissapointing playoff performances of any recent closer, but he was dominant in the Phillies 2008 post-season run. He’s had some great seasons as a closer, and some woeful season as a closer. Is he un-clutch with occasional bouts of clutchyness, or just inconsistent? I would go with the latter.

    2. The points been made earlier, but it bears repeating. I imagine the marginal difference in stress between an at-bat in the post-season and being a poor kid hitting in front of a scout knowing that your life will change if you play well on that day is at lot less than people think. Same goes for the stress a minor leaguer would feel while playing in front of scouts from the big team.

    The point is that playoff stress or red sox yankees stress is not that materially different from the stress that major league players have been playing with over a longer period time. Consequently, over a large enough sample size, a players numbers in what we fans think of as high leverage situations will fall in line with their career numbers.

    Which brings me to my third point…

    3. Alex Rodriguez. Has there ever been a more psycho analyzed baseball player? Has there ever been a player that more people were convinced reacted differently to “pressure moments” than non-pressure moments. Surely if there was ever a player who could show that some people “have it” when it all counts and some don’t, it would be A-Rod. Many have argued as much. MANY.

    And yet…now that he’s played in a significant sample of post season games, his career post season numbers are now…eerily similar to his career regular season numbers. Look them up.

    I believe certain intangibles do exist: effort, smarts, baserunning intuition, ability to deal with stress etc. Clutch, i’m not so sure about. The numbers just don’t seem to bear it out.

    Over a large enough sample, players seem to perform at their expected levels in all of the different situations they can be placed in during a baseball game.

  84. pete January 21st, 2010 at 12:15 pm

    Yair – I didn’t mean to imply that your post was an attack on sabermetrics, just that it seemed to mischaracterize its advocates as people who ignore the human element or as people who believe that a player is naught but his stat-line. The goal of sabermetrics is to better statistical analysis, not to become the sole form of analysis. In other words, saber-minded folk want to, when evaluating a player through statistics, find better ways to do so than RBIs and batting averages and wins.

    And I understand that there is a belief that there is a human element that enables players to “will” themselves beyond what their peripheral statistics indicate their performance is, or that some players can help a team win more than their numbers suggest they are. I’m not utterly dismissing that. But these are elements that sabermetricians have spent a lot of time researching, whereas it seems like traditionalists simply accept it as gospel without even attempting to evidence their claims. Fangraphs is an excellent example of this, since their win-probabilities stats track players’ performances in the clutch, and what has come of that is that for the most part, as samples grow, players either regress or progress to their career norms in those situations, and while a few players do consistently underperform in those situations, while others consistently overperform, such players are exceedingly rare, almost never the players we think of when we think of clutch or unclutch, and their rate differential is almost always quite small. In other words, there are a few players out there who are a little bit better in the clutch, and a few who are a little bit worse. The rest is simply perception bias, confirmed by selective memory. This is not an attempt to undercut the human element of the game, it is merely a suggestion that the human element is random and unpredictable, and over time corrects itself through variation. Which of course makes a great deal of sense. In fact I’ve always wondered why those disparaging statistics rely so heavily on postseason or clutch statistics based on “the human element” in their player evaluation.

    You make the claim that some players are just “winners”. This I cannot help but disagree with. For every example you give me, I can point to one of two much more likely explanations for these trends: either the player is an extraordinarily productive player, which will show up in advanced statistics, or the player consistently plays on excellent teams. Quite often, it is both, as it is with Jeter. But other times, you have cases like Joe Mauer’s, where his team was lifted from relative mediocrity by a combination of Mauer’s extraordinary offensive gifts (and of course his, along with any catcher’s, WAR is flawed because we haven’t quite figured out how to quantify or statistically evaluate catcher defense) and presumably good defense and the weakness of the AL-Central. And other times still you have the Eric Hinskes of the world, who seemed to “lift” teams to success when in reality those teams were dominant teams without him – the ’07 red sox had a dynamite rotation and a great offense anchored by Manny, Ortiz, Youkilis, Pedroia, and Ellsbury, and solid defense, the ’08 rays had no injuries to their starting rotation, the best defense in the league, and good offense, and the ’09 yankees, well, were the ’09 yankees. Of course I’ve only looked at 3 players here, but I simply can’t think of an example of a “winner” who doesn’t fall into one of those categories. If you can, then by all means point him out, but I just have a hard time biting into the theory that a player’s contribution to a team’s success can’t be quantified pretty accurately, even if we’re not quite there yet. I’m willing to be convinced, of course, but the saber side seems to consistently out-research, out-argue, and simply out-logic the anti-saber side. It’s hard for me to jump on board with the side that refuses to back up its arguments with anything empirical, and relies instead on the statistical aberrations that, while they define baseball as a sport and provide it with excitement, tend to correct themselves as samples expand. Again, nobody’s arguing that, for example, aaron boone’s 2003 HR was not special, or that we shouldn’t love him for it. They’re simply arguing that if you decided to rely on boone to continue to provide moments like that to make up for his lack of strong statistical production, you’d regret it. Sabermetrics are designed to help teams win in the regular season by accumulating enough data that statistical aberrations are corrected and not relied upon. In other words, nobody is trying to stop great moments from happening, they’re simply trying to find ways to win without relying on them.

  85. raymagnetic January 21st, 2010 at 12:15 pm

    Wave,

    Successful people all the time say that what makes them successful is not their talent, but their desire to be great.

    I remember Karl Malone had a tape out when I was in high school talking about how he had a desire to be better and would work out harder because if he didn’t someone else would. To me that quality of an individual is immeasurable.

    You can have all the talent in the world, but without a certain desire and will to be great you will likely not achieve what you could have in life.

  86. socrates January 21st, 2010 at 12:16 pm

    SJ44
    January 21st, 2010 at 10:24 am
    Doesn’t beat the 0 when you don’t have to work if you choose not to.

    You can analogize athletes and “regular” human beings in the workplace.

    Johnny Damon, if he chooses to, can say, “I’ll stay home with the kids” and call it a day if he chooses not to take a 60+% paycut to play this year.

    He’s already earned 100 million in the game. Just bought another new house in an all cash sale in Orlando.

    He has no pressing financial needs. Meaning, he will either play for a price and team he is comfortable with or he won’t.

    Its different from the real world workplace where most people have to take what’s offered to put food on the table.

    Johnny has no such pressures to do that in his present situation.
    *****************

    Although I totally agree with the above, I wouldn’t be a famous philosopher if I didn’t ask “Then wouldn’t the opposite also be true?”

    I.E. Shouldn’t Cashman be saying, if the difference between having a significant upgrade in the OF/lineup is 1 or 2 million dollars (in the form of JD starting and BG as the 4th OF and also in case Nick Johnson goes on the DL and not having to play both BG and Rule V) then why am I heeing and hawing this guy to death. Its obvious that Reed Johnson and et al just don’t tickle Cashman’s fancy or else he would have signed them to a 3 million dollar contract already.

    Just sayin

  87. murphydog January 21st, 2010 at 12:18 pm

    Twain:
    There are three kinds of lies: lies, damned lies and statistics.

    Vin Scully:
    Most people use statistics the way a drunk uses a light pole, for support, not illumination.

    As for me, statistics in this context are valuable, if at all, when used as a (not “The”) predictive tool. The visceral reaction should not be ignored. The visceral is the voice in Archie’s head telling him to marry Betty instead of Veronica because Betty has intangibles. The stat head marries Veronica because she’s better endowed financially and otherwise.

    Statistics are not the components of a crystal ball with which to see a perfect vision of the future, if we only had more statistical measures, approaching infinity. I mean, how many times can you slice the baloney and still have baloney in each slice? Human endeavor, and baseball is still that, is too complex to be captured entirely in a series of snapshots, even if spliced together. Panoramic pictures are nice, but still not the same as being there.

    It is akin to the divide between rationalism and faith. IMO, you are best served by applying both, not arguing which one should supplant the other. Personally, I’ll take Pascal’s Wager.

  88. Chip January 21st, 2010 at 12:20 pm

    On a completely unrelated note, I almost feel bad for the Mets. With the exception of Jason Bay they have had an awful winter. With that in mind I would like to take a moment to contemplate some moves for the New York Metropolitans:

    1. Sign Jarrod Washburn (2 years $16 mil)
    Joel Piniero was their first choice but the fact is that he’s probably just going to be another Jeff Suppan, Jeff Weaver type of guy who leaves Dave Duncan’s magic wand and winds up regressing. Washburn has at times been very effective in the much harder American League and pitching in Citi Field he would be a fine fit.

    2. Trade for Bronson Arroyo. He’s not nearly as good as some people make him out to be but he would give the Mets the one thing they don’t have – an effective pitcher capable of throwing over 200 innings. I’ve heard that the Mets are thinking of investing in CM Wang or Ben Sheets – and that’s fine, but given their track record with injuries lately I don’t think you can go into a season with one pitcher (Mike Pelfrey) in the rotation who was not on the disabled list for an extended period last season.

    3. Eat the Luis Castillo contract and sign Orlando Hudson. You’re not going to get anyone to take on Luis Castillo, accept that fact and just cut him and eat the loss (or move him to a bench role) bring in Orlando Hudson who wants to be a Met, would add a much needed dynamic to the clubhouse and is a vast improvement over Castillo.

    4. Sign Hank Blaylock. There was a lot of talk that the Mets were going to bring back Carlos Delgado but now he’s going to the Jays. I say let him go. Blaylock is younger and has just as much power as Delgado. Carlos may hit for a higher average but he’s also been considered a problem in the clubhouse for the last couple of seasons.

  89. m January 21st, 2010 at 12:22 pm

    5. Fire Omar Minaya.

    (I’m serious, the Mets used to be a desirable team to play for. It’s a wreck now, and it’s the people at the top. The team culture is bad.)

  90. tex's friend January 21st, 2010 at 12:23 pm

    http://bombersbeat.mlblogs.com.....light.html

    ___

    Note to Hairston. By July 31, San Diego will be salary dumping and you will be back on the team.

  91. SJ44 January 21st, 2010 at 12:25 pm

    Perhaps Cashman doesn’t see Damon as a significant enough upgrade to warrant the money it would take (at least at this time) to sign him.

  92. Wave Your Hat January 21st, 2010 at 12:27 pm

    raymagnetic-

    Let me take your previous examples of Gooden and Strawberry.

    How can you be sure that they didn’t have as much “heart” and “desire” as the next ballplayer, or even more, but because of questionable choices in their personal life embarked on a path, either because they were misinformed, were irrational or whatever, conflicted with that heart and desire and ultimately, perhaps, reduced their overall achievements from what they would otherwise have been?

    You can’t be sure. IMO, terms like “heart” and “desire”, when applied to baseball players which have achieved enough to make the major leagues, are likely to be mental constructs in the mind of the observer, which reflect the observer’s subjective judgment of the difference between what that player actually achieved and the subjective judgment of what that player should have achieved.

  93. Jerkface January 21st, 2010 at 12:27 pm

    And yet, who would you say is more clutch than Jeter? Based on memory, Manny & Ortiz used to get big hits. Matsui has come through at times. Posada. Difficult to rank them, but I guess that’s something writers could debate.

    Jeter is not clutch. He is just a good baseball player. That is my point. If Jeter were super clutch, and no one will question his WILL TO WIN, he would not have grounded into a double play as the Yankees were rallying to win Game 5. Do we believe that Ryan Madson wanted to win more than Derek Jeter?

  94. SJ44 January 21st, 2010 at 12:27 pm

    As for the “Nick Johnson will go on the DL” argument, that’s fan babble.

    The same fans making these “factual” predictions also predicted, with certainty, that AJ Burnett would spend significant time on the DL. How did that work out.

    Just about all of Nick’s injuries have been flukes, occurring when he plays the field.

    I suspect as a DH, he will have as many AB’s as Matsui did last year.

  95. raymagnetic January 21st, 2010 at 12:27 pm

    The Mets never should have fired Willie Randolph the way they did.

    The Mets are now a victim of Karma justice. :)

  96. Patrick January 21st, 2010 at 12:30 pm

    Ray won’t that quality show up in the stats though?

    For example, Derrick Coleman had world class talent. Everyone thought he’d be an all time great basketball player. He ended up being good but he never had the desire to be great. His attitude and personality showed up in his statsheet. Talent was off the charts but his lack of drive held back the results.

  97. Chad Jennings January 21st, 2010 at 12:32 pm

    New post

  98. dont-forget-where-you-came-from cheese mac January 21st, 2010 at 12:34 pm

    To hear many a sabermetrics buff tell it, Derek “Intangibles” Jeter brings nothing but a sound shortstop skill set to the table.

    I stopped reading after that. That is, like, the EXACT opposite of what any “saberhead” would say about Jeter. His fielding has consistently been overvalued by, guess who, the MSM, whereas no one can deny his offensive production (which also gets overvalued because hes Derek Jeter).

    Try to have a clue about what it is you are talking about next time.

  99. raymagnetic January 21st, 2010 at 12:34 pm

    Wave,

    If Gooden and Strawberry had a certain heart and desire they wouldn’t have let themselves fall victim to their cocaine habits in my opinion.

    Gooden and Straw had enough natural talent to make it to the majors. They had so much natural talent that if they didn’t let obstacles such of drug use get in their way they would have been great players. Instead to paraphrase Dinero in a Bronx Tale, they’re just examples of wasted talent.

  100. m January 21st, 2010 at 12:34 pm

    Jerkface,

    Not fair. That was one at-bat over the course of a long, fruitful career.

    If Jeter isn’t clutch, then who is? Are you saying there is no clutch? Sometimes players come through, sometimes they don’t but if they do it at important times or often enough the rightfully get the reputation.

    One thing, though. Once upon a time (as recently as a couple of years ago), Jeter was the rally guy. He either got it going or kept it going. Then the hand & legs things kicked in and he had a tough season hitting into a lot of double plays. But looking at his overall career, Jeter’s had the opportunities and has taken advantage of enough of them and at certain, critical moments that have contributed to his reputation.

    You can (and people have) nitpick or poke holes, but the fact remains that Jeter does have the reputation.

    Another thing is that Jeter probably had more “opportunities” as a #2 hitter that doesn’t possess an inordinate amount of power.

  101. SJ44 January 21st, 2010 at 12:35 pm

    Can’t fire Omar since they just signed him to an extension last year. The Wilpons will not eat that money.

    They will just move him to the side of the room, which they are doing right now, and not have him involved in the big decisions the team decides to make.

    I’m always leery of the “will to win” argument.

    In my lifetime, I haven’t met too many athletes who didn’t have a “will to win” at the professional level.

    Some, in fact most, who don’t win don’t win because of factors entirely out of their control.

    For example, I haven’t met as many competitive people as Dan Marino.

    He didn’t win a Super Bowl not because he lacked a “will to win”. He didn’t win one because he lacked a defense and running game for most of his career in Miami.

    ARod has never lacked a “will to win”. Yet, he has been psychoanalyized, mostly incorrectly, by just about every sportswriter and fan in America.

    Does anybody really believe this year is the first time ARod had a “will to win” and actually won?

    Or is it that he just played on a superior team this year and contributed to them winning.

  102. BK January 21st, 2010 at 12:37 pm

    Great article. I use statistics everyday in very similar way as they are used in baseball, but for sales. And what you wrote summarizes what I try to explain everyday. Stats can be a guide, but for statistics to prove forward, the future must resemble the past. Statistics should and are used as a guide, but there will always be things the numbers cannot calculate

  103. Wave Your Hat January 21st, 2010 at 12:43 pm

    “If Gooden and Strawberry had a certain heart and desire they wouldn’t have let themselves fall victim to their cocaine habits in my opinion.”

    My problem with that is, if it is true, then all heart and desire become are the residue of a career – if you achieved success, you had it, or if you achieved more than observers thought you could, you had it. If you made poor decisions and they limited your success, you didn’t have it.

    But the terms themselves, the way I see you looking at them, don’t mean anything until everything else is known. I mean, in October of 1986 you would have said Doc and Darryl had miles and miles of heart and desire.

  104. Mr. Faded Glory January 21st, 2010 at 12:57 pm

    *”The more statistically-minded baseball community has often adopted the following implicit assumption: Players are essentially machines, largely unaffected by clubhouse atmosphere, personal psychological factors or the day-to-day effects of real life. Columnists who refer to “team cohesion” or a player’s “mindset” (think Alex Rodriguez) as factors in performance are treated with indifference, if not derision, and are considered a product of a bygone era where intuition trumped hard data.”*

    Uh, where are you getting this from? Just making stuff up willy-nilly about sabermatricians because you don’t understand how numbers work is weak sauce, kid.

    Here’s the way a conversation would go, hypothetically, where you “own” a numbers guy.

    You: A-Rod must really be affected by his breakup, he’s playing like garbage lately.
    Guy: What are you talking about? Sure he’s 0-4 today, but since the announcement last week the guy’s OPS is over 1.000 and he hasn’t made an error.
    You: Uh, do you even WATCH the game?? Look at him out there right now, he seems lost!!!

    I’ve learned there’s just no arguing with people who don’t understand the value of stats.

  105. Wait till we do it all over Again January 21st, 2010 at 2:33 pm

    This article is utterly ludicrous. I’m amazed it was posted.

  106. A.D. January 21st, 2010 at 2:53 pm

    “but sometimes people exaggerate their significance.”

    Usually the opposite is true, and that the avg fan plays up the significance of what they “saw” with their eyes, and what perception they have of certain players are, which are almost completely biased by the lack of sample set someone has ever seen & the selective memory that everyone has.

    The line from Momento puts it best: “Memory can change the shape of a room; it can change the color of a car. And memories can be distorted. They’re just an interpretation, they’re not a record, and they’re irrelevant if you have the facts.”

    You use the example of “holding a lead” and “buckling under pressure”. These are things that will be played out in the stats, if a player unravels easily when things go wrong, then they will have a worse career track record & thus lesser stats. If players buckle under pressure then they will do so in various pressure situations, and thus have worse numbers.

  107. Brian January 22nd, 2010 at 8:17 am

    Any regular reader of River Ave Blues will tell you that they use advanced stats to help tell a story, or to make a story more clear. They usually do an excellent job at it as well. They also acknowledge things such as team chemistry or clubhouse atmosphere can make a difference. Psychology and Sports Psychology to play a roll its just that these variables are so hard to measure that its hard to say that they definitively play a part in a players performance on any given day. The guys at Baseball Prospectus put together a really good book called Baseball Between the Numbers a few years ago and they actually tried to quantify “clutch performance” but did so with little confidence in it. Just like we have a hard time measuring Utility or Happiness in Economics (there is no unit or measurement of these, although some economists are trying to come up with a unit called ‘utils’ to create a unit of utility) we can’t really measure Clutchness, we can only infer it using Advanced statistics such as using Leverage Index to try and determine performance in High Leverage situations (holding a 1 run lead in the ninth inning with men on base vs facing the bottom of the order in the 3rd inning with a comfortable lead). The quest in using advanced stats for baseball, as with any field that requires heavy stat usage (economics, even psychology) is to try to quantify all variables, tangible or otherwise.

  108. Brian January 22nd, 2010 at 8:31 am

    One thing I forgot to mention, using stats from certain scenarios also causes a sampling bias. If Player A gets 1000 appearances in high leverage situations and always performs above their career averages then it may be safe to say that they are clutch, but most pitchers don’t even get 1000 appearances in a career let alone in a season. And some face a disproportionally large number of High Leverage situations than others (Mariano definitely sees more than Sergio Mitre would). Thats why RBI isn’t a great stat, because if you put Albert Pujols on a team thats filled players that get on base less than 30% of the time then his RBI counting stats will be inferior to that of a player on a team full of Nick Swishers who are on base machines. Even batting order matters as a lead off batter won’t see many RBI opportunities because he has the 8 and 9 hitters (usually the worst on the team) in front of him as opposed to two of the best hitters. One of the biggest obstacles to measuring these, besides the fact that its hard to quantify clutchness or luck, is sample size


Sponsored by:
 

Search

    Advertisement

    Follow

    Mobile

    Read The LoHud Yankees Blog on the go by navigating to the blog on your smartphone or mobile device's browser. No apps or downloads are required.

Advertisement

Place an ad

Call (914) 694-3581