Buy The Book
Over all 5 seasons examined there is a improvement in the second attempt success rate conditioned that the first attempt was good. This data is for the mean of the individual level, eliminating any skew of the results due to aggregation.
And data is made available.
Glove-slap: mettle.
Great stuff by Ben Blatt.
All games between 1980 and 2010 in which the game was tied in the ninth inning, a total of over 2300 games, were included in the data set.
- When the home team blew the lead in the ninth, it won in extra innings 53.1% of the time.
- When the home team came back in the ninth, it won in extra innings 53.8% of the time.
Will asked me about the value of consistency.
He gave me an extreme situation, and I responded that from a pitching standpoint, you’d rather have the guy who gives up 0, 20, 0, 20, 0, 20 runs rather than 10, 10, 10, 10, 10, 10. The inconsistent pitcher will win .500, while the consistent one will be close to .000, even though both gave up 10 runs per game. The hitting is the exact opposite, so the preference would be on a consistent hitter.
So, technically, yes, consistency matters: you want it in hitters, and don’t want it in pitchers (all else being equal). Practically though, players are not like that. You’re not going to get anything close to an extreme situation that it’ll matter. That is, first you have to be able to identify a consistent/inconsistent player, and then you have to establish the degree to which he’s like that. To then put a value on that, and I’d be shocked if we’re talking about more than 0.1 wins for a player. Basically, I called it a rounding error.
It’s an interesting exercise, and well-presented. I think the most likely explanation is that the PH and non-PH games were intermingled. What do you guys think?
I’m just waiting for part 2. So far, all we’ve seen is that David Wright’s 2010 season was streaky, but we still don’t know if he just happens to be one of the extreme point (i.e., SOMEONE has to be streaky, just by luck), or if the number of streaky players is larger than expected by random.
I’ll bet we’ll only see a bit (not much, but some) of streakiness, for the following reasons:
1. you see the same pitcher in 3 or 4 PA
2. same park
3. same weather
4. same health
By re-ordering the PA as the author is doing, he is removing all this. So, just by virtue that a player is going to be the same guy in the same day facing the same pitcher on 3 or 4 occasions, we expect SOME persistence, over and above any “streaky” factor innate to him as a player. Once you factor all that out, I would say the streakiness factor will be close to zero.
Bill James study. The degree to which we observe a difference was pretty big for me:
From the years 2000 to 2009, I identified 504 “matched sets” like this in which two starting pitchers had nearly-identical records, but one was hot and the other was not. Details:
...
For example:
Randy Johnson as of September 5, 2000, had made 30 starts with a won-lost record of 17-6, 2.45 ERA, 299 strikeouts.
Randy Johnson as of September 7, 2001, had also made 30 starts with the same won-lost record (17-6), same ERA (2.45), but 320 strikeouts. But he was ten degrees hotter at that time in 2001 than he was in 2000.
...
The “hot” pitchers, in their 504 “next starts”, had a won-lost record of 199-175, an ERA of 4.28, and an average Game Score of 50.62. The “cold” pitchers, in their 504 next starts, had a won-lost record of 177-177, an ERA of 4.74, and an average Game Score of 47.94.
I like the overall idea. The results seem pretty large.
I think we need better controls. Pitchers with a 177-177 record (i.e., .500 on 354 decisions) don’t have a 4.74 ERA. If they did, they come from hitter’s parks in the higher run seasons in the 2000-2009 decade. In order to do a study like this, you have to eliminate potential sources of bias. So, runs allowed per league average, adjusted for park, would have been preferable. I don’t know that we even need to look at the W/L record. There’s a few more tweaks that can be done.
Overall, I like it, I like the idea, I like the execution. It just needs different eyes looking at it from a similar angle.
Matt takes a look, and he does it along the lines of what MGL did for the hot/cold streaks in The Book: focus on a very narrow time frame, on the idea that if something were to happen, it would be most noticeable immediately.
As you’d expect, little to no change. The article’s value is in its process, not in its findings (or lack of). People who read the article and say they learn nothing, well, you missed learning something.
***
I also want to point out, as I always do, that just because something is statistically significant, it does NOT mean that that particular observed difference is the estimated true difference. All we know is that we estimate that we have a true non-zero difference.
In short, it tells you that there is a difference, but it says nothing about how much the observed difference is the true difference (other than the higher the t-test, the more the difference is real).
Fack Youk captures it for me:
[Halladay v Yankees is] a valid storyline. He’d pitched more than a full season’s worth against the Yankees in his career and had great results. How do you not bring that up?… Essentially, what you are looking at when you try to analyze Halladay’s career stats against the Yankees (or any pitcher’s line against a certain team) are a bunch of very small samples, recorded over a very long time, and smushed together to look like one big one....But there’s two parts to this - the validity of the stats and how they are being used.… So why do we fans have the desire read about games before they happen? Why do we bother to write a preview for all 162 of them here?… More specifically, when I write a preview or read one that someone else has written, it’s because I want to have an understanding of any trends and storylines coming in and try to develop some sort of a framework that will make what’s unfolding on the field a little more coherent and interesting to me. Some of those things might be statistical, but the personalities and rivalries and the who-owns-who are compelling in their own way, even if they don’t pass the statistical smell test.
Basically, don’t conflate what has happened with what may happen, no matter how much you try to dissect each player’s career.
A perfect game is when a team’s batters gets 27 straight outs. Well, how about a single batter who gets 27 straight outs?
Andy Fox went 38 plate appearances in 2004 without reaching base safely. Joe McEwing had 34 PA in 2002. Wil Nieves had 31 over a span of THREE YEARS!
Those are the only three players with at least 27 straight PA of not reaching base, since 1993.
(Note: actually, baseball-reference counts streaks on a game-basis rather than PA-basis. I’ll leave it to someone else to see if in the surrounding games they extended their streaks.)
All players need rest. Benching a player for one day here or there because he has been cold or because you didn’t like the color of his shoes, or for whatever reason, is fine. Same with dropping a player in the batting order. It’s just not that big a deal. Just don’t pretend that somehow that is going to make your team better for that day. It doesn’t. It makes it worse. You can sometimes do the right things (rest a player occasionally) for the wrong reasons (because he has been “cold").
So yes, you need a ‘hot’ goalie to win the playoffs. It’s just not possible to figure out who’s going to be ‘hot’ until after the fact.
Jeff takes a look, and this at least has ALOT more potential than looking at OBP and SLG “trends”. Love the idea.
Walt Davis says:
Still, it’s over a 5-sigma variation which does suggest that this is a “process out of control” but all that tells us is that the “failure” rate (where each HR is a “failure") is higher than it used to be. The question then becomes what’s the new rate. It will be several hundred PA before we’ll have a good estimate of that but I’ll just point out that Adrian Gonzalez had 11 HR in May in 5 fewer PA so there’s no reason to think Mauer’s new level of talent is any higher than Gonzalez’s (who, to this point, has been a 25-35 HR guy though obviously he’s on a torrid pace right now).
And as always, I point folks towards Steve Dillard, Aug 2-17 1979—this sort of thing can happen “randomly.”
My response:
The Baseball Analysts has now become my 1A to the 1 that is Hardball Times. This time, it’s Sky that looks at how hitters perform after they’ve been on a 30-game hitting streak:
The “during streak” line is actually “performance from game 31 until his last hitless game” (as best as I can figure).
But in the 22 prior NHL seasons since the trophy was first handed out to the Edmonton Oilers in 1986, President’s Trophy winners have won the Stanley Cup seven times, including the Detroit Red Wings last year. Considering that 16 teams make the playoffs each year, that’s a remarkable run of success.
...
In Major League Baseball during that span, by contrast, just three teams — the 2007 Boston Red Sox, 1998 New York Yankees and 1986 Mets — have won the World Series after finishing with at least a share of first place for regular-season wins.
...
Since the ABA-NBA merger, 15 of 33 champs have at least tied for the best record, compared to nine titles for teams that finished in second or a tie for second. (Good news for the Cleveland Cavaliers, who won their opener against Detroit by 18 points.) And 20 of 39 Super Bowl champs since the AFL-NFL merger held the league’s best record, or finished in a tie — compared to 11 for second-place finishers.
I thought the results were important enough to warrant its own thread, rather than continuing the last one on first half, second half splits. If you didn’t follow the discussion in the thread I was talking about, here is the link:
http://www.insidethebook.com/ee/index.php/site/article/first_half_second_half_splits/
To summarize the discussion, a few people pointed out some unusually large first half, second half splits in performance for some players. The discussion is whether for players in general, those splits “mean” anything, which is the same thing as asking whether they have any predictive value, which is the same thing as asking whether they correlate to any degree from one year to another. For example, we find that platoon splits for RHB have very little predictive value. No matter what a RHB platoon splits are in any given time period, they will tend to revert to near league average for all RHB in any other time period. For LHB, there is some predictive value - the larger the sample size of data we have, the more predictive those sample results are.
For RHB (since there is some predictive value), we might need 10 years of split data to “tell us anything” about that player’s true talent platoon ratio or difference. For LHB, it might be 2 or 3 years of data.
Anyway, I was skeptical that any sample of first half and second half splits means anything, i.e., has any predictive value. Of course, even if there is a tiny amount of predictive value in any sample data, if the sample is large enough we eventually get tremendous predictive value. But, in baseball, we really only get to use one year at the least and maybe 5 or 10 years at the most, worth of data to have any practical significance, of course. If we have to wait until we get 15 or 20 years of data for it to have much predictive value, that is not particularly interesting, to me at least.
Anyway, one way to see how much predictive value there is in a certain amount of data, we can run a regression from one time period to another. If the correlation is really low, then there is little or no predictive value to that particular stat for that amount of data (the number of opportunities underlying each element in the regression). Hopefully we have enough data (both data points in the regression and a decent sample size for the underlying number of opportunities for each data point), such that the uncertainty in the resultant “r” is fairly low (a small standard error).
I did such a regression on first half, second half splits. Here is the methodology and the results:
Phil looks at it on the team level.
Phil: do you count the Expos games as “home” or “away” when in P.R.?
This is the headline and a snippet from an MLB article on the Tigers page from yesterday:
DETROIT—The Tigers’ new $153.3 million man has not lived up to expectations. But that’s not to say he won’t eventually.
Leyland apparently agrees with that sentiment.
I just looked at my current player database as of a few days ago, and Cabrera had a +33 per 150 park-neutral lwts. What the hell are they talking about?
Rany says:
Since 1956, do you know how many pitchers had made three straight starts with 8+ innings, <=3 hits, <=1 walk, and 8+ strikeouts? Here's the list, which I have put in alphabetical order for your convenience:
Cliff Lee.
That's it.
I thought maybe his definition was too strict, that maybe I can loosen it up a bit. After all, there have been many great back-to-back great games pitched. But, not 3? So, I asked B-r.com for pitchers with at least 8 innings and at most 4 baserunners. I get back just 5 pitchers. Lee, of course, Koufax and Fryman as Rany noted. Plus Jaime Navarro and Tom Cheney. Navarro managed to give up 3 runs on those 10 baserunners. Bad luck, relatively speaking. If I loosen it a bit more to at least 7 innings, Pedro in 2002 comes in with 4 straight games (no runs allowed), along with 3 in 2005 (4 runs allowed). Lee has a total of 25 IP. So, if you make it at least 7 innings per outing, and 25 IP in total, Lee is tied for 6th.
You can shake it up any way you like. But, Lee has done something quite remarkable. Good job on Rany.
Feb 11 04:03
MGL: Today on Clubhouse Confidential
Feb 11 04:02
Reader Mail of the Day: Why do we need X years of fielding data? And what about outliers?
Feb 11 02:12
Performance through the ages
Feb 11 02:10
Dwight Evans
Feb 10 23:01
For Your Soul
Feb 10 21:07
Hero of the month: Brittney Baxter
Feb 10 18:32
Moneyball at Villanova
Feb 10 17:00
Psst… wanna intern in Canada?
Feb 10 15:01
New PECOTA
Feb 10 14:28
Win expectancy charts used in football… in 1983!
THREADS
February 10, 2012
Jose Molina
February 10, 2012
Reader Mail of the Day: Why do we need X years of fielding data? And what about outliers?
February 10, 2012
Performance through the ages
February 10, 2012
Hero of the month: Brittney Baxter
February 10, 2012
Win expectancy charts used in football… in 1983!
February 10, 2012
Dwight Evans
February 09, 2012
Psst… wanna intern in Canada?
February 08, 2012
Moneyball at Villanova
February 08, 2012
MGL: Today on Clubhouse Confidential
February 08, 2012
New PECOTA
Recent comments
Older comments
Page 1 of 320 pages 1 2 3 > Last »Complete Archive – By Category
Complete Archive – By Date