Monday, March 14, 2011

Inspiration and Perspiration

Got a great little lesson about stats, life and Tiger Woods last week. Well, it might not actually be a lesson about life or Tiger Woods, but it is definitely a good lesson about stats.

I recently posted a long and rambling thing about a few of the hitting stats that I find interesting. I don't like all of them, certainly do not like them all equally, but what I like about advanced baseball statistics are that they can get you thinking about HOW you might try to measure something. How would you go about trying to measure a batter's hitting ability? A pitcher's ability to prevent runs? A defender's ability to play his position? These are complicated questions with many, many layers of questions within them. It's fascinating for me to see some of the more thoughtful statistical minds attack all these questions.



Well, I mentioned in there a very interesting statistic called wOBA, invented by my e-migo Tom Tango. This stat gives a well-thought out value to everything an offensive player does. I won't go into any more details here except to point out two seeming quirks of wOBA, two quirks I touched upon in the original story.

Quirk 1: A reached on error is worth more than a single.

According to the wOBA chart I included, a single is worth .90 while reaching on error is worth .92. This seems interesting.

Quirk 2: A hit batter is worth more than a batter walking.

A non-intentional walk is worth .72. A hit-batter is worth .75. Again ... interesting.

Before I explain to you why these things are so, I should say that I came up with my own theories about why these things might be so. And before I tell you my absurd theories, I should say that baseball fans all decide how much they want to believe in things that they cannot see. That is to say that everyone will choose to believe how important leadership is for a baseball team, how significant and varying is the ability to perform in the clutch, how big a part mental qualities like self-doubt and unbreakable confidence and experience and guts and heart and all that play in the failures and successes of players.

I think there is a sliding scale -- some people think these intangibles mean EVERYTHING in baseball, some think these intangibles mean almost NOTHING in baseball, and most people fall somewhere in between. We can call this the McCarver Scale. McCarver -- and most other color commentators, to be fair -- tend to think intangibles are pretty close to 100% of the game.* And so I'd say I score 12% on the McCarver scale. Maybe 8%. I think those qualities like veteran leadership and competitive nature do play their part in the game.

*Or, anyway, that's how they talk on TV.

But I think -- and this is just my theory -- that things like that are almost always overstated because a part of us WANTS these things to matter more than they do. We WANT (many of us) to believe that players who drive in a lot of runs have some special talent for hitting with runners on base. We WANT (many of us) to believe that pitchers who win a lot of games have special talent for winning games no matter what everyone else on the field does.

And, hey, I have these same prejudices. That's how Tiger Woods gets in. Every realistic instinct in my being tells me Tiger Woods is done as the best golfer in the world. Done. I really don't think he will ever get back up to the top. I've been over my reasons a dozen times at least -- he's 35 years old (and probably even older in golf years since he has been playing, since he was 3), he's had major knee surgery, he can't find a swing that fits his current body, he has been trampled by the culture he created, and there are many very talented young golfers who grew up with Tiger Woods as their standard of excellence and are not intimidated or unfamiliar with his greatness. I am now at the point where I would be thoroughly surprised if Tiger Woods reached the top again. To tell the truth, I would be less surprised if Tiger Woods fell off the world golf map entirely.

I THINK that ... but every time Tiger Woods plays, I again hold my breath. This past week, he's playing at Doral and I held my breath. And here's why: Part of me so respects Tiger Woods' competitive nature that I cannot help think if he WANTS it bad enough, if he GETS ANGRY enough, if he FOCUSES HARD ENOUGH, then he can will himself back into the greatest golfer on earth. I may believe logically that such thoughts are silly or naive or flat misguided, but I still have those thoughts. I can't help it.

Then Tiger Woods is tied for 31st at Doral and I remember reality again.

Then Tiger Woods shoots six-under on Sunday and I again hold my breath for the next time. That's pure emotion. And it colors the view.

Similarly, when I saw those quirky baseball numbers above -- about reached on error and hit-by-pitch -- I could not help but immediately pin the reasons for the difference to something mental and vague and ambiguous. For instance, when I saw that a reached-on-error had a slightly higher value, my thought was that this might be because reaching on error has a negative impact on the the defensive team's psyche. The pitcher's ticked off that the defense let him down, the defenders let down slightly because they know they should be out of the inning, the batters have a little more confidence because they have been given a second life. I figured this was the reason a few more runs are scored.

And when I saw that hit batters are worth more than batters who walk, I had almost the exact same thought. I figured the MENTAL reaction -- anguish on one side, a little added fury on the other -- leads to scoring a few more runs.

In both cases, my theory was completely wrong and the real answers both (1) make more sense and (2) do not rely on creative accounting.

For the reached-on-error issue, Tom Tango explains that reaching on error is worth slightly more than a single only because these will include occasional errors where the hitter ends up on second or third base. Grounder to third is thrown away, that means a runner goes to second, and that's the extra value. That's it. That's the whole difference. There is absolutely no other detectable difference.

As for the hit-by-pitch, the reason it is worth more is because it happens at more random times than walks. A pitcher can have some control over a walk. He might be more likely to walk someone with first base open, for example. But a hit-by-pitch is a much more random act. Only a small, small, small percentage of HBP are purposeful. As such, they tend to lead to slightly more runs.

Of course, the numbers we are talking about are so small that nobody could possibly just notice them. That's the thing about baseball numbers. Someone who hits .296 gets hits on 29.6% of his at-bats. Someone who hit .302 gets a hit 30.2% of his at-bats. If someone gets 184 hits in 622 at-bats, he hits .296. If someone gets 188 hits in 622 at-bats, he hits .302. The difference is four hits over a WHOLE SEASON. That's fewer than one hit a month. You simply could not notice that unless you were charting it.

By charting it, you get those batting averages which tell you, decisively, which guy got more hits. But the more you chart, the more you take a little bit more myth out of baseball. Charting baseball basically proves that the tiny things that have become part of the mythology of baseball, well, they might exist, maybe, but only as tiny things. Thomas Edison said that genius is 1% inspiration and 99% perspiration. It's a good quote, but he was pretty wildly overestimating. It's probably closer to .0001% inspiration.





34 comments:

  1. Might Casey has struck out (0.0017301)

    ReplyDelete
  2. that's MightY
    error, so add 0.003 to my McCarver stone

    ReplyDelete
  3. Does the HBP vs. BB calculation account for the additional pitches thrown in a BB, which must have some value to the team of the player being walked?

    ReplyDelete
  4. This is in the same vein as your HBP explanation, but I have a theory about that: I would think that it's much more likely for a guy who walks a guy late in a game to be lifted (either due to pitch count or perceived wildness) than a guy who beans someone. An HBP is momentary wildness as you mention. But, if the pitcher is actually lifted due to a walk, it's almost a lock that the next pitcher brought in will have the correct lefty vs. lefty or righty vs. righty matchup, which favors the pitching and defense and would lead to less runs, which would make a walk worth less.

    Just a theory, though, and it's probably a combination of a bunch of things because 0.3 seems like a big difference for something that is seemingly the same on the surface.

    ReplyDelete
  5. @stephen...the difference is actually 0.03, which really isn't that big.

    ReplyDelete
  6. I dig advanced stats too, but this post seems to acknowledge something that I've felt for a few seasons, which is that when the Luddites say that sabrmetrics is sapping the fun out of the game they kind of have a point.

    On balance my enjoyment of baseball has been enhanced by the exposure I've had to advanced stats, but it's a different kind of enjoyment. Deeper, more nuanced, more objectively correct, but also less romantic, less mythical. One kind of fun is replaced by another, and you have to admit that the new kind is not as appealing as the old to a great number of people.

    ReplyDelete
  7. It's tricky to assess what part of genius is inspiration and what part perspiration because the reason guys like Tiger and Jordan worked so hard at their respective sports is because of inspiration. I mean, there are plenty of guys who can drive the ball 300 yards, there are plenty of 6'6", athletic forwards in the NBA, but very, very, very few of them approach being the best in the world.

    You can argue that it's all hard work that gives Jordan or Tiger their greatness. But what drives them to work so hard? What drives them to be so competitive? To me, that's inspiration, just as much as Little David Eckstein's scrappiness.

    ReplyDelete
  8. Interesting post, Joe. I have another idea to add to the mix. Some subset of HBP are those thrown in retribution--e.g., because a teammate was hit in the previous half-inning, or because a player was showboating a homerun trot. Those HBP often come at the WORST POSSIBLE TIME to put a man on base: the start of an inning is probably most common. If managers and pitchers were smart, they'd save their retribution HBP for the most situationally opportune time (or not do it at all). But then it wouldn't make sense as an intangible. A retributional HBP is probably as pure an intangible-driven act as you can get: you put your team in a statistically worse situation in order to demonstrate your team's moxie. It would be interesting to see whether NOT doing a retributional HBP has a positive or negative outcome, though of course I have no idea how you'd measure that.

    ReplyDelete
  9. I always heard the Edison quote as 10%-90% inspiration-perspiration. Your point remains, though.

    Since your last post was much about pain scales, and this one has a McCarver scale, I think you should combine them to make a McCarver Pain scale to evaluate announcers: "On a scale of 1 to 10, with 1 being Vin Scully and 10 being Tim McCarver, how painful is it to listen to this broadcaster?" Hey, there could be at least several columns in that.

    I would tend to agree with HBP being more random, but I would have to make an exception for Greg Maddux. Given his overall control (BB/IP, K/BB, etc.), his HBP numbers are rather high. Having watched him literally hundreds of times, I am pretty sure a fair number of those were intentional. In fact, to follow up on Jay Ess' observation, I often wondered if that was his version of an "unintentional intentional" walk, and he was saving his pitches.

    And aside from the age/injury type issues, I don't think Tiger is ever coming back because he is now making excuses for his performance, the latest being how hard it is to be a good golfer while also being a divorced father of two. The old Tiger would never, ever had excused poor performance for any reason.

    ReplyDelete
  10. "have excused" in that last sentence, obviously.

    ReplyDelete
  11. The idea I came up with for a BB vs. HBP is that people can get on base when they are behind in the count on a HBP. One common occurance (it seems) is for a pitcher to be up 0-2 or 1-2 and they are trying to set up the slider low and away (the strikeout pitch), so they come up and in... HBP. So the strikeout turns into a base. That fits into the "randomness" idea, because the pitcher can be in control of the AB, trying to get an out, but the pitch sails 3-4 inches further inside, and he's lost the hitter.

    ReplyDelete
  12. What be the standard error on those estimates? Is 0.90 really different than 0.92? Is 0.72 really different than 0.75?

    ReplyDelete
  13. @Jay Ess- It does not, just as a walk on 4 pitches is not differentiated from a walk after a battling, 13-pitch AB.

    ReplyDelete
  14. Nice analysis of how our analysis gets skewed by someone's most recent performance. Great use of two examples to show the rationality of advanced stats. Interesting comments. Thank you all.

    ReplyDelete
  15. When a batter walks unintentionally, the pitcher has thrown at least three pitches outside the strike zone that were not tantalizing enough to lure the batter into swinging. The counts will be 3-0 (a hitter's count), 3-1 (still a hitter's count) and 3-2 (a neutral count). The expectation of what will happen when ball four comes is not very much better than the expectation if ball four were instead a strike. So the fact that the last pitch is a ball (instead of the hitter teeing off on a grooved fastball trying to avoid the walk) does not much increase its value.

    On the other hand, most times there is a hit batsman, it's not on a hitter's count. Given that pitches are roughly evenly distributed between hitter's counts (where the expectation is higher than normal that a good offensive result will happen), pitcher's counts, and neutral counts, it seems that only about 1/3 of the time is there going to be a situation where the HBP is not a major improvement over the expected neutral count or pitcher's count result.

    Put another way, the average OPS on 3-0 and 3-1 counts is very high, so the HBP (OPS: 1.000) is only a slight improvement on it. The average OPS on 0-2 and 1-2 counts is the worst, so the HBP is a huge improvement over that. Yes, on some at bats the 0-2 count gets worked for a walk, but it's not the way to bet.

    And once in a rare while an umpire will toss a pitcher who unintentionally hits somebody, forcing a cold(er) pitcher into the game, getting into the bullpen (or deeper into the bullpen) sooner, all of which increases the expectations of scoring more runs.

    Yes, there is added value both to the team and to the run expectations of tiring the pitcher, making him throw four balls. Otherwise, I'd expect HBP to have much more of an advantage over a BB.

    ReplyDelete
  16. I realize McCarver can get on a listener's nerves, but he is not a bad color man. I have made this point before, but McCarver provided color man insight that, if the manager had been as smart of McCarver, would have prevented the game winning base hits in two deciding world series games. He said that infield should be playing back right before the D-Back guy (Gonzales?) hit a flare just out of Jeter's reach to win the world series - he explained that Rivera caused a lot of broken bats that produced such soft line drives. He also said that the Rangers should not be pitching to Renteria in last year's WS, just before Renteria hit the home run to win the final game. I don't believe there is not another color man anywhere who has hit on such inside color man observations in big games.

    The Tiger issue is fascinating. I just don't know if normal principles of athletics apply to golf and to Tiger. I think he has a much greater chance to getting back on top than any other 35 year old golfer.

    ReplyDelete
  17. Yeah, but he also said "A Mark Wohlberg fastball. Catch me if you can.".

    ReplyDelete
  18. Didn't McCarver just mess up the name and intend to refer to Mark Wohlers and his great fastball? I think we need to give guys on a live microphone some leeway on misstatements and even stupid statements.

    ReplyDelete
  19. The most interesting moment of Tiger Woods' golfing at Doral was when he hit a 122-yard drive in the 2nd round. That's not a typo, folks. Tiger Woods hit his drive 122 yards on the 2nd hole - the old drop-kick duck hook (it's on YouTube, at least for the time being).

    ReplyDelete
  20. I would like to weigh in regarding the peak and decline ages for golfers and how that might affect what we can expect going forward from Tiger. I work for a sports data company and have access to some pretty decent statistical analysis.

    Below is a table showing the number of tournament wins per age over the past 10 years in the European and PGA Tours combined. What it shows is that the four most common winning ages are the four years between 31-34. At age 35 (i.e. Tiger's current age) there is a sharp drop-off, and every single year from age 35-51 except one (39), players of that age win fewer tournaments than players one year younger.

    Age Wins
    17 3
    18 4
    19 7
    20 16
    21 29
    22 36
    23 61
    24 62
    25 104
    26 131
    27 148
    28 135
    29 158
    30 142
    31 171
    32 172
    33 188
    34 183
    35 145
    36 134
    37 111
    38 100
    39 122
    40 97
    41 70
    42 70
    43 50
    44 42
    45 35
    46 29
    47 19
    48 19
    49 16
    50 6
    51 1
    53 2
    59 1

    ReplyDelete
  21. An interesting analysis might be too look at the retributional HBP--that is, those times when a pitcher hits the opposing team's leadoff hitter to "send a message" after a pitcher's teammate was hit in the previous half-inning. From a stats point of view, this is purely bad: by putting a player on base with no outs your chances of winning the game decrease. An intangibles guy would argue that this is outweighed by the effects of the move on pride or morale or the down-the-road effects of stopping a team from walking all over you, etc. In other words, a retributional HBP is purely stats-vs.-intangibles: there is NO positive effect from a stats perspective, and any benefit at all, if there is one, must come from intangibles. If someone could figure out a way to measure this, it would be interesting to see the results. (And I suppose Bill James did this in 1985 or something.)

    ReplyDelete
  22. Re: the golf table two posts up, of course I meant to write: every single year from age 35-51 except one (39), players of that age win the same number or fewer tournaments than players one year younger.

    ReplyDelete
  23. Re the golfer's wins by ages -

    If you add up all the wins from ages 17 through 34 (Tiger turned 35 in December), that's 1,750 wins or 62% of the total for all ages If you add up all the wins from age 35 on, that's 1,069 wins or 38% of the total.

    ReplyDelete
  24. Nick,

    If you're taking requests... can I have the same list restricted to the 4 current majors [US Open, PGA, Masters, (British) Open] from 1934 forward?

    Thanks,
    45debc30-33c5-11e0-bfa3-000bcdca4d7a

    ReplyDelete
  25. Unfortunately, for whatever reason, our data is incomplete for golf, which is not one of our major sports. The best I can do is the four majors only from 2002, but the sample size is obviously quite small:

    Age Wins per Age
    25 1
    26 3
    27 1
    28 1
    29 4
    30 2
    31 3
    32 3
    33 2
    34 1
    35 4
    36 4
    37 2
    38 1
    39 2
    41 1

    ReplyDelete
  26. @Dan Shea --

    So, if Tiger fits the overall profile, and he now has 62% of his expected career wins, he "should" finish with a total of 115 PGA Tour wins, and 23 majors.

    I am prepared to bet everything I own (not much, admittedly, that he doesn't get within 20% of either of those career totals. In other words, he will finish with fewer than 92 wins and 18 majors, IMHO.

    ReplyDelete
  27. Thanks, Nick.

    I was just curious as to how the distributions of winners by age looked comparing majors to other tournaments.

    Based on my quick calcs it looks as if both groups have a mean age for a winner in the 32 to 33 years range, with a smaller standard deviation in the majors as very young (esp.) and very old (less so) winners are rarer in the majors.

    ReplyDelete
  28. Those numbers seem to reflect that golfers in the 35 to 39 range continue to win a significant number of tournaments, not much different than the 25 to 30 range. It seems like those number support the proposition that Tiger can make a nice comeback (which I hope he does not).

    ReplyDelete
  29. Tim McCarver also said, in regard to whether Phil Garner should have let Brad Lidge pitch in Game 6 of the 2005 NLCS to get rid of any residual effect from the massive Pujols HR in game 5, "I don't think there is one" (a residual effect) while Lidge was delivering a board-straight, down-the-middle fastball to Scott Podsednik (0 HR) that Pods turned into a walk-off HR.

    So his track record is obviously mixed.

    ReplyDelete
  30. Anything McCarver says that is correct is simply a function of his mouth running all the time. You know, stopped clock, blind pig, infinite number of monkeys, etc.

    On a per-utterance basis, he is somewhere south of the Mendoza line.

    ReplyDelete
  31. Another factor in Tiger's decline that's never mentioned:

    Perhaps Tiger's sexual escapades (blowing off tension and steam in an extreme fashion) actually made him a better golfer, and that by making that public he is now prevented from living the life that, for him at least, seemed to be incredibly compatible with high athletic achievement. So consequently we will see his golfing career decline.

    It may well be that what Tiger needs to get his game back in shape isn't a new swing or more practice, but more wild sex with hookers and porn stars. It worked before, and it didn't seem to harm his game, whereas clearly having that lifestyle exposed and broken seems to have hurt it immensely.

    Don't fix what wasn't broken, is perhaps the lesson here. The moral to this story may not be what the public wants it to be.

    ReplyDelete
  32. @Broken Yogi --

    My thoughts exactly.

    As Crash Davis put it:

    "If you believe you're playing well because you're getting laid, or because you're not getting laid, or because you wear women's underwear, then you ARE!"

    Respect the streak!

    ReplyDelete
  33. Broken Yogi,

    My guess is that the only difference between Tiger's sex life as a divorce' versus as a husband are two things that he now keeps in his bedside nightstand:

    1. pre-sex nondisclosure agreements
    2. condoms

    ReplyDelete