THE BOOK--Playing The Percentages In Baseball

Friday, February 25, 2011

Scorecasting review, part II

By , 05:47 AM

I wrote this to the authors on their web site:

Dear Sirs:

I am a professional (having worked for several MLB teams, notably the Cardinals in 2004 and 2005) sabermetrician and I have been working extensively in this field for over 20 years. I am the “inventor” of one of the most widely used advanced defensive metrics, UZR, I am one of the co-authors of the sabermetric book, “The Book,” and I host, with one of my colleagues (Tom Tango), a popular and highly respected sabermetric blog, http://www.insidethebook.com (click on the “blog” link).

I recently read your new book, Scorecasting, and I liked it very much. It was well-written, clearly presented, and well-researched, as far as I can tell. I believe it has broken some new and important ground as well. I have recommended it to many of my colleagues and to our (blog) readers.

I was particularly interested in your baseball research of course, especially that pertaining to home field advantage (HFA), a topic that I am not unfamiliar with. While it has been researched some over the years, it is admittedly one area that we (sabermetricians) know comparatively little about.

While you provided some well-researched, eye-opening insight into the role (both in quality and quantity) that umpires may play in the HFA in baseball (and other sports), I must say that some of your research in that area conflicts with similar research I conducted recently, after reading your book.

I would appreciate it if you could take the time to read my comments below (they are edited and reprinted from my blog) and address each issue as you see fit to do.

Kindest regards,
Mitchel G. Lichtman

To recap part of what I wrote in Part I, the authors in Scorecasting said this:

To test our theory, we first compared all pitches, about 5.5 million of them, from 2002-2008 made in stadiums using QuesTec versus those without it. For example, we looked at all called pitches when the Astros visited the Cardinals (at their non-Questec stadium) and when the Cardinals visited the Astros (at their Questec equipped stadium).

What did we find? Called strikes and balls went the home team’s way, but only in stadiums without Questec, that is, ballparks where umpires were not being monitored. This is consistent with an umpire bias toward the home team causing the strike-ball discrepancy. We also found something surprising. Not only did umpires not favor the home team on strike and ball calls when Questec was watching them, they actually gave more strikes and fewer balls to the home team. In short, when umpires knew they were being monitored, home field advantage on balls and strikes didn’t simply vanish; the advantage swung all the way to the visiting team.

Basically, they say that the home team bias exists in non-QT parks in 02-08, but that in QT parks, it was reversed – there was a road team bias.

My data do not support that claim, using the same 11 QT parks that they did (they listed them in the footnotes) in the same years (02-08). I found a .58% bias in favor of the home team in QT parks and a .46% home bias in non-QT parks. So not only was the bias was not reversed in QT parks, as the authors assert, but it is higher (in favor of the home team, as always).

Now, as it turns out, comparing the home/road differentials in QT and non-QT parks is not fair anyway, in terms of isolating the effects of the park. The reason should be obvious and the authors, at least one of them, should have addressed this issue.

When you compare the H/R bias in QT parks compared to non-QT parks, you are comparing the effects of the park and the effects of the QT teams.

Imagine that all the QT teams had league average batters but pitchers who threw many more strikes (per pitch) than league-average pitchers. And let’s say that the park had no influence on the umpires or anyone else. Well, in QT parks, the H/R differential would be enormous, because it would be driven by the home team pitchers who threw lots of strikes. In non-QT parks, in games where the QT teams were not involved, the H/R differential would be normal (the teams would not affect the numbers) and in games where the road team was a QT team, the H/R differential would be greatly reduced or reversed, because the road pitchers would be strike-throwing pitchers.
So the bottom line is that if we want to see if the park (QT or non-QT) has any effect on the differential, we can’t just look at one subset of parks (like QT parks) and compare them to another subset (like non-QT). We have to control for the teams. They can’t do it and I can’t do it. So their finding (that the differential is reversed in QT parks) doesn’t mean anything and neither does mine (that the differential is larger in QT parks) until we control for the teams.

One way to do that, at the risk of losing some sample size, is to just look at QT teams in QT parks and compare that to non-QT teams in non-QT parks. That will tell us the true effect of the park, more or less, since the home and road teams in both sets of data are essentially the same.

Here is that data:

QT v. QT teams

H: .323
R: .319

.4% home bias.

Non-QT v. non-QT teams

H: .318
R: .312

.6% home bias.

So this time, the QT parks do indeed have a smaller home bias, but by no means is it reversed, as the authors claim.

(BTW, as it turns out, the QT teams are not league-average when it comes to the called strike percentage of their pitchers and their batters. Their pitchers have a .6% advantage compared to non-QT pitchers, and their batters have a .3% disadvantage, again, as compared to non-QT batters. So the QT teams have a net advantage of .3% in called strike percentage, whether they are the home or road team and regardless of what park the play in.)

To recap the leverage issue that I discussed earlier and then expand on it, the authors assert that the home bias, and I’ll sometimes call that umpire bias, even though I am not certain how much of it, if any, is due to the umpire, is much larger in high leverage situations (when the game is on the line) than in low leverage situations (when the outcome is not too much in doubt).

I also found that to be the case using my original, simple definition of low and high leverage. Here is my original data for 02-08:

HL
.74%

LL
.24% (I originally posted .18%, which was wrong)

So I found more than 3 times the called strike differential in HL situations than LL ones.

I reran the data using a much more rigorous definition of high and low leverage. This time I used Tango’s complete LI chart (rpg, inning, relative score (home v. away), bases, and outs) to separate all PA into low and high leverage ones.

The Scorecasting authors did not say what criteria they used and unless they licensed Tango’s charts (which I don’t think they did) I’m not sure how they would do it, unless they duplicated his work. They did reference Tango when talking about LI though.

Anyway, I defined low leverage (LL) as any PA where the LI was less than .1. That seems low, but around 8% of all PA were LL.

High leverage (HL) PA were defined as LI > 2.0 and constituted about 9.5% of all PA).

I think that is a pretty fair, albeit arbitrary (and round), division. Again, I have no idea how that scheme compares to the authors’.

Remember that one of the authors’ central theses in the chapter on baseball HFA is that most of it is due to umpire bias, especially in ball and strike calls. Furthermore, they say that because of the psychology of this bias, they are much more biased (in favor of the home team) in HL situations when the game is on the line, and much less biased in low leverage situations, where it doesn’t really matter (so they might as well just call them as they see them).

Of course you could create the reverse argument from a psychological standpoint - that the umpires favor the home team and the crowd, but only when it doesn’t influence the outcome of the game. But that is neither here nor there, and I am certainly no expert in prospect theory or cognitive or behavioral psychology.

The authors claim in their book that in low (and medium) leverage situations, there is no home/road bias in terms of the percentage of called strikes (of all pitches not swung at). In fact, if you look at their “Difference in Percentage of Called Pitches that Are Called Strikes on Home v. Away Batters,” you will see that in low and medium leverage situations (however they define them), the bias is actually reversed in favor of the road team.

Here is what they said about HL and LL situations and umpire bias (my comments are in parentheses):

…in low leverage situations, when the game is not much in doubt, the home team advantage in receiving fewer called strikes and more balls (they mean to the batters, BTW) goes away (it actually reverses). But as the following chart shows, the called strike advantage for home teams grows considerably as the game situation gets more and more important.

Again, in my initial look, where I used a gross definition of high and low leverage, I found that from 02-08 there was a .74 home bias in HL situations and a .24% positive difference in LL situations. While this is a large difference, these numbers contradict the authors’ claim that there is a reverse bias in low and even medium leverage situations.

The difference between my old results and theirs could easily be due to our definitions of low leverage. As I said before, because my initial criteria were so coarse, my low leverage PA included lots of actual higher leverage situations.

With my new definition, that is not the case. Low leverage was any PA with a LI of less than .1. So what was the home/road called strike differential in this pass (in all parks from 02-08)? It was .3%, a little larger than when I used the much simpler definition of high and low leverage. So again, I have to dispute the authors’ claim that the bias “disappears in low leverage situations.” I say that with some trepidation because I don’t know how they bifurcated the PA (LL and HL) and I am not quite sure of the rest of their methodology as well as the integrity of their and my pitch database. (Plus, either one of us could have made one or more computation errors.)

What about the new high leverage differential? The old one was .74%. The new one was 1.1%, 50% higher. So their claim that the home/road bias is much higher in HL situations is once again confirmed. Interestingly, in their chart, in “very crucial situations” they have the bias at only around .55%.

As I said, I am not 100% convinced that most or all of the difference is due to umpire bias.

What about their Questec claims with regard to leverage? They are pretty bold. They say that in non-QT parks, where they know they are not being monitored, the home bias is large in HL situations, and that in QT parks, where they are being evaluated, the bias in HL is actually reversed, that is, the road team gets favored. (In the last thread, I mis-reported what they said. I stated that they asserted that in QT parks, in HL situations, the bias was smaller than in non-QT parks, and that in LL situations, the bias was reversed. In fact, they say nothing about the difference between QT and non-QT parks in LL situations.)

Here is exactly what they say in their book:

That is, when the game is on the line, home teams in non-QT stadiums get a big strike-ball call advantage and those in QT stadiums get a huge strike-ball disadvantage.

As I said before, I would be skeptical of those claims off the top of my head, especially the last one (reverse bias), but the data will speak for themselves. We are not dealing with anything subjective here on the part of the researchers or me, other than perhaps their choice of words, like “large”, etc.

Using my rigorous criteria for high and low leverage, albeit likely different than the authors’ (unknown criteria), in HL situations in QT parks, I get 1.3%, which is higher than the overall 1.1% (all parks).

In non-QT parks, in HL situations, the home bias is .9%, as you would expect (after subtracting the 1.3% weighted by 11 parks from the overall 1.1% weighted by 30 parks) and after rounding.

So there is no reverse bias in the QT parks as I suspected. Not even close. In fact, the home bias is larger (likely due to the fact that QT teams have a .3% called strike bias to begin with, as compared to non-QT teams, and they are always the home teams in QT parks of course). I cannot imagine how the authors came up with this result from the data. There does not seem to be any justification for it, no matter what their definition of high and low leverage situations is.

I can only assume that they accidentally reversed the home and road called strike percentages. In fact, I ran the numbers twice just to make sure that I didn’t make the same mistake.

BTW, in LL situations, QT parks had a home bias of .5% and it was .3% in non-QT parks, but again, the difference is likely due to the teams and not the parks. The “real” difference (between QT and non-QT parks in LL situations), after adjusting for the teams, is probably around the same.

As I mentioned before (in case you didn’t read the last post or forgot it), I believe that the authors make another error in the same section on leverage and Questec. They say:

In practical terms, when an umpire is not being monitored by QT, a home batter in crucial game situations will get a called strike only 32% of the time if he doesn’t swing. In the same situation, a batter from a visiting team gets a called strike 39% of the time.

Those numbers appear to be wrong. In their own charts, the authors show the H/R differences to be on the order of less than 1% and not 7%. In fact, I think they mean something like 30.2% and 30.9% and somehow it got printed as 32% and 39% - or some such thing.

Then they go on to say:

Now consider the same two situations when the umpire is being monitored by QT. Here the home batter gets a called strike 43% of the time, and the away batter only 35% of the time.

Again, those numbers appear to be wrong, as all of the called strike percentages are on the order of 30% to around 32%, and the differences between home and road are only as large as around 1.3% (not 7 or 8%), according to my data and analysis. Plus, they have the home and road effect reversed, as far as I can tell from my data output, as I mention above (the home team always has the advantage whether it is a QT park or not – small in LL situations, and large in HL situations – there is no “huge reverse bias”).

• Sabermetrics • Books • Statistical_Theory

#1 David (see all posts) 2011/02/25 (Fri) @ 13:46

It used to seem a little bit pretentious when sabermetrics was described as a “science”. Technically, I guess that it fit the definition (which is kinda a flexible definition, anyway), but it seemed a little bit too haughty and even pretentious a word.

This, though, is exactly what I like to think of when I hear the term “peer review” - other scientists, hobbyists, and enthusiasts having their B.S. antenna raised and then, driven by a dedication to honesty, investigating the issue with rigorous dedication to empirical facts.

In fact, I think that this sort of article is BETTER than much modern science because I know that many issues are so politicized and/or corporatized that the ultra hyped “peer review” process has been shown to be totally ineffective and is, in many scientific studies, just a rubber stamp.

So, anyway, I think that this little investigation is a great testimony for sabermetrics and I hope that (a) you continue your scrutiny of Scorecasters and (b) you continue to publicize your work.

If the authors don’t respond honestly and completely, then I think that you should write strong letters to their publisher. I was loosely involved with the retraction of a fraudulent book by the scumbag “scientist” Charles Pellegrino and we had to send faxes and make calls to the publisher and lots of mainstream media outlets.

(There’s a small formatting error in the “quote box” near the top of your article when you’re quoting from the book. One of the quoted paragraphs isn’t in the box. No biggie.)

#2 (see all posts) 2011/02/25 (Fri) @ 14:17

We know there is HFA. We seem to know that there is some differential favoring Home Team in called strike %. Let us take the null hypothesis that there is no umpire bias. Because there is HFA the home team is more likely to be ahead, and moere likely to be significantly ahead. Wouldn’t this affect the tactical decisions of pitchers and possibly hitters? With a 4 run lead wouldn’t a pichter be told to throw strikes, lowering the opposing OBP, while possibly raising BA and SLG?
Couldn’t this explain some (all?) of the differential in called strike %? Also, batters follow different tactics depending upon being home or away. Down by 1 late in the game the home team is more likely to play for the tie, and use one run strategies such as the sac bunt or the stolen base. Putting these strategies into effect would affect the batters, (and probably the pitchers), decision making in whether or not to swing, which would affect the called strike % in a systematic way without umpire bias.

I think the data exists to test the above hypothesese. The Scorecasting authors seem to me, (I haven’t read the book yet), to have seen a result in the data, and stopped at the first explanation that came to mind. (And perhaps aren’t enough subject matter experts to have seen these issues and non-umpire bias explanations of their results.

#3 MGL (see all posts) 2011/02/25 (Fri) @ 14:39

kds, sure, I can easily control for the score and see if your hypothesis is true.

The authors present pitch f/x data that supports their thesis that umpires have a markedly different strike zone for the home and road teams, although at this point, I cannot accept any of their findings at face value.

Other pitch f/x researchers (Dan Turkenkopf, Mike Fast, Phil Birnbaum, and J-Doug) have shown a small difference in strike zones between the home and road teams (accounting for around 16% of a typical HFA of 4% - 54/46), but not the large difference that the Scorecasting authors claim.

David, wouldn’t the publishers of a book be pretty hostile toward someone questioning the integrity or competence of one of their authors?

#4 (see all posts) 2011/02/25 (Fri) @ 14:49

(There’s a small formatting error in the “quote box” near the top of your article when you’re quoting from the book. One of the quoted paragraphs isn’t in the box. No biggie.)

Fixed. Thanks.

#5 David (see all posts) 2011/02/25 (Fri) @ 15:08

MGL wrote:

“....wouldn’t the publishers of a book be pretty hostile toward someone questioning the integrity or competence of one of their authors?”

The question seems to imply that the opinions of some businesspeople trumps or should be allowed to contaminate empirical data, so that’s kind of an odd question.

I have no clue what the emotions or thoughts of the people at the publishing house would be. Really, it doesn’t matter anymore than it should matter what any employers feel when presented with abject wrongdoing on the part of their employees. They have a responsibility - professional, civic, and human - to do the right thing. (I’m not saying that companies won’t or shouldn’t factor in loyalty when they’re deciding how to respond to the wrongdoing - even if the authors betrayed any loyalty through their fraud. But the loyalty to the employee should factor in at the punishment stage, not at the acknowledgment of wrongdoing stage. In other words....I think that publishers have the same obligation to be honest as I believe everybody else has. Just ‘cause I might have a bond with an employee doesn’t give me the right to deliberately tell lies to the entire world.)

For me, when I sent contacted the publishing house of Charles Pellegrino, I never got a personal response. All I know is that, within days of the letter that I sent out (which was paired with some other work being done by others), the publisher said they’re going to revise the book. When this then triggered more evidence of fraud, they decided to pull it in perpetuity.

If you’re interested, you can read about my little odyssey at my original blog article:
http://jamescameron.blogspot.com/2010_02_01_archive.html

Or at this Amazon.com post I wrote:
http://www.amazon.com/review/RDJTBP7EWVMI2/ref=cm_cr_rev_detmd_pl?ie=UTF8&cdMsgNo=8&cdPage=1&asin=0805087966&store=books&cdSort=oldest&cdMsgID=Mx36PI5OVGE7BP4#Mx36PI5OVGE7BP4

(Quick aside on the issue: I personally felt kinda spent and lousy after the whole Charles Pellegrino thing. I believe this was because it’s always better and more challenging to create then to destroy, even if the destruction - of lies, in this case - is necessary and righteous. So I enjoy debunking and catching fraudsters as much as anybody, but I personally found that, for me, it’s not good for the soul to do it for too long.)

Hyperlinks aren’t allowed here, right?

#6 Tangotiger (see all posts) 2011/02/25 (Fri) @ 15:15

If you put http followed by :// , then the blogging software MIGHT flag it as spam, and I have to go unlock it.

If you put www. something.som, then the blogging software let’s it go through, and hyperlinks it for you.

In your case, I just edited it to put in the necessary text to hyperlink you.

#7 David (see all posts) 2011/02/25 (Fri) @ 15:16

That explains it. Thanks.

#8 MGL (see all posts) 2011/02/25 (Fri) @ 18:26

David, wow, I just skimmed your long post (and links) to the Pellegrino issue. I applaud your effort to debunk a fraudster. I absolutely hate people that misrepresent facts for profit. It is an epidemic, as anyone who listens to satellite radio knows (every single commercial - and I mean every single one - is selling a fraudulent or misrepresented product or service).

That being said, how is this guy (Pellegrino) any different from any of the thousands of people who publish books and web sites (and sell products) spouting nonsensical ideas and misrepresenting facts?

BTW, what is the purpose of your web site and what is your association or interest with James Cameron? For the record, I think he has done some good (not great) movies and some awful ones. And I think Avatar, despite the wonderful animation technology, was one of the worst made movies I have ever seen (a slight exaggeration), and I lost all respect for Cameron after walking out on it in the theater. In short, the script (which is 90% of the quality of a movie, IMO) was awful, which is inexcusable for a high budget movie that took years to make. It seemed to me that he was just about ready for production after working on the movie for years, and someone said, “Hey James, we forgot a script.” And he responded, “Oh yeah, crap! I’ll whip one up in a few days. I’ll have a couple of highs school students work on it.”

#9 (see all posts) 2011/02/25 (Fri) @ 18:49

@mgl #8 - i had very similar feelings after seeing Avatar. Considering Cameron made T2 and Aliens, i’ll probably keep seeing any movie he makes (it’s not like they come out all that often) and I’m glad i got to see Avatar in 3-D and in theaters, but i’m never going to rewatch it, unless its re-released in theaters and in 3-D 15+ years from now and I’m feeling particularly nostalgic.

#10 MGL (see all posts) 2011/02/25 (Fri) @ 18:57

In addition, I think it was a travesty that it was one of the favorites for the best picture Oscar last year (thank God it didn’t win). I guess that was politics and power though…

#11 David (see all posts) 2011/02/25 (Fri) @ 19:26

The reference to the fraud with Charles Pellegrino was just to illustrate that it’s very possible for bloggers to identify fraud and to take action against it - sometimes even with the extreme effect of having the book withdrawn.

If the authors of Scorecasting don’t show their work and there’s a solid explanation for why their research was the opposite of yours, then I definitely believe that a retraction should be forced. That’s just my opinion.

As far as James Cameron goes, I just run a fan blog and podcast about his works - that should be pretty clear from the links there. I’m writing a (self-published) biography about him that’s focusing on the technical side of his movies, too. But I’ve never met him and have no professional affiliation with him.

Obviously, everybody has their own opinions about movies. Some of my favorite movies of recent years - ‘Speed Racer’ and ‘Kingdom of Heaven’ come to mind - have been disappoints or outright flops. I don’t expect everybody to like them, and I don’t like everybody to like Cameron’s movies, either - even if his last two are the biggest blockbusters of all time, by a Barry Bonds ‘01 thru ‘04-caliber margin.

Here’s the hilarious margin by which Cameron’s movies obliterate every other movie:
boxofficemojo.com/alltime/world/

Here’s a painting that Cameron did before making ‘The Terminator’:
img.photobucket.com/albums/v703/DBrennan3333/TerminatorConceptPoster.jpg?t=1298675800

Here’s one of many patents that Cameron has:
http://www.google.com/patents/about?id=91weAAAAEBAJ&dq=5189512

Here’s a speech Cameron gave to NASA when he was on their board of advisers, despite refusing to become an American citizen:
nasa.gov/pdf/185156main_1st_exp_conf_cameron_transcript.pdf

Not bad for a Scottish Canadian who dropped out of community college and is absolutely reviled by the American media.

#12 MGL (see all posts) 2011/02/25 (Fri) @ 21:40

Part of what I am trying to say is that if every book that was shown to have inaccurate claims, facts, and data were retracted or republished, how many books would that leave? Not too many, as the industry is flooded with books about garbage and nonsense. Many of them are obvious - so if publishers were loathe to publish inaccurate books, none of these would get published. They do, and most publishers are only interested in one thing, and we know what that is. Now, sometimes a publisher gets forced to retract certain things in certain serious books, due to overwhelming publicity, but that is another story and that is an economic decision as well - not an ethical one.

So what I am saying is that no matter what I or anyone else would send to the publishers about the claims or data in this book, the chances that they would print a partial retraction or retract the book completely are next to zero.

Wow, you are pretty defensive about Cameron. All I said was that I hated Avatar, which is pretty much the consensus among any serious movie-goer than I know of (other than that it had some great technical effects). The fact that it was a blockbuster means that it was a blockbuster. Some movies are popular, in part, because they are good quality movies, and some movies are popular for other reasons.

Now, what constitutes “good or bad” quality in a movie, independent of its popularity among the average movie goer is another story and somewhat of a philosophical issue (and unanswerable, I think, at that).

I wasn’t aware that the media “hates” Cameron. Why is that? And how is he so smart, which you seem to imply, yet he teams up with this character Pellegrino?

#13 MGL (see all posts) 2011/02/25 (Fri) @ 21:54

To address kds comments in #2 above:

I ran the numbers for 02-08 when the game was tied only (lots of high leverage situations of course).

H: .325
R: .316

H/R diff=.9%

(In all games at all times, the H/R diff is .5%.)

BTW, a game is tied in 27% of all PA.

When the home team is ahead:

H: .327
R: .304

H/R diff=2.3%

PA=468,550

So there is a 2.3% “advanatge” for the home team when they are ahead.

The home team is ahead in 37% of all PA.

When the visiting team is ahead:

H: .313
R: .326

H/R diff= -1.3%

PA= 536,508

The road team is ahead in 42% of all PA.

So there is a 1.3% advantage for the road team when they are ahead.

So it appears that which team is ahead has a significant effect on the H/R called strike percentage differential.

Now, we don’t know which way the causality goes. When the home team is ahead, is it because they are throwing more strikes or do they throw more strikes when they are ahead? Probably both, but probably mostly the former. So I am not sure we are getting any insight into the issue of umpire bias versus pitcher approach from this investigation, other than the fact that in a tie game, the home team still has a large (.9%) advantage in called strike percentage, suggesting that it is not the pitcher’s (or batter’s) approach that is driving it.

BTW, does anyone know why the road team is ahead more often than the home team?

When the road team is ahead the average margin is 2.941 runs. When the home team is ahead, it is 2.996 runs (obviously if the road team is ahead more often, the margin has to be higher for the home team, since they score more runs overall). That includes all PA, other than the last one when the home team scores the winning run in a walk off victory.

In any case, if the road team is ahead more often than the home team (despite losing more games and scoring fewer runs of course), and it is true that pitchers will throw more strikes when ahead (which they will), then that would mean that the umpire bias may be MORE than we might infer just by looking at the called strike to ball H/R ratio.

Now, just looking at called strikes and ball percentages tells is nothing about the umps, per se. If home pitchers pitch better and home batters recognize pitchers better, we would expect to have H/R differentials like this, even if umpires had no bias in how they call pitches.

That is where pitch f/x come in. The pitch f/x researches can look at a pitch and the umpire call, and independent (mostly, but not completely) of the batter and pitcher, determine if there are pitch calling biases by the umpire in favor of one team or another (or any one thing and another).

J-Doug, Mike Fast, Phil Birnbaum, and Dan turkenhopf have done excellent work in this area. Dan and J-Doug came up with an umpire home team pitch calling bias on the order of .12 runs per game (difference between home and road teams), which is around 30% of the total HFA.

(They both say around 15%, but that is not correct. The home team scores around .2 runs more than average and the road teams scores around .2 less than average, and .12 is 30% of .4 runs. Maybe someone can clear that up for me.)

Here is what the authors say about home and away pitchers and pitch f/x:

First they say that pitchers are just as accurate home and away, which, if true, is a pretty important observation. We have always assumed that one of the components of HFA is that pitchers pitch “better” at home, presumably independent of the umpires. If it is true that the accuracy of pitches is the same, home and road, which admittedly is only one aspect of pitching skill, that puts a serious dent in that assumption, I think, plus it strongly supports the notion that umpires ARE indeed responsible for most if not all of the H/R called strike differential (batters may be responsible for some part of it).

Anyway, the authors write this:

..on average MLB pitchers are equally accurate at home as on the road, throwing a ball within the strike zone 44.3% of the time at home and 44.5%...on the road.

Can anyone (Mike, Dan, J-Doug) confirm that this is true?

They also say that according to pitch f/x, they throw with the same velocity and movement (not sure how they measure that) at home and on the road, even when controlling for the type of pitch thrown.

Like the aforementioned sabermetricians the Scorecasting authors also looked at the pitch f/x strike zones to determine home/road pitch calling bias by umpires. Using the average value of missed calls, they found that around (they say it is a “rough estimate") “an extra 7.3 runs per season were given to the home team by the plate umpire alone.” They don’t exactly say whether that “extra” is compared to the road team or compared to an average team (do they not have people who proofread these things for logic?), but let’s assume that it is “as compared to an average team.” So, that is 7.3 / 81 games (again, I am assuming that by “whole season” they mean 81 home games, but they could mean “per 162 games” I suppose.

So 7.3 / 81 is .09 runs, which is only slightly more than the sabermetric researchers found (.06 for the home team and another .06 for the road team).

Now, here is really where the researchers get it wrong, which Phil pointed out several weeks ago. I should have put this into my letter to them.

They say:

...this adds up to an extra 7.3 runs per season given to each home team by the plate umpire alone. That might not sound significant but cumulatively, home teams outscore their visitors by only 10.5 runs in a season. Thus more than 2/3 of the HFA in MLB comes by virtue of the home plate umpire’s bad calls.

I don’t know why they used the word “cumulatively” there, but that is neither here nor there. I don’t think you need to be a subject matter expert to know that there is something wrong with that logic. 10.5 runs a season, or .06 runs a game for the difference between the home team and visiting team run scoring seems awfully low, doesn’t it (and from 02-08, I get .117 rpg or 18.9 runs per season)?

That would be equivalent to about a 50.6/49.4 home road winning percentage, just a trifle low.

Obviously the 10.5 runs per gameis because the home team doesn’t bat in the bottom of the last inning almost half the time. You have to use runs per 9 innings or “adjusted” runs per game to figure the home team run advantage. And that is around an extra 40 runs or so, of course. So rather than 10.5, it is around 50 runs (actually in 02-08, it was 65 runs, or .4 runs per 9 innings). And 14.6 (7.3 for the home team and 7.3 for the road team) divided by 50 is 29% (the umpire’s share of the HFA), around the same effect as J-Doug, Dan, Mike, and Phil noted (although they said 15%, or half of that).

BTW, since they authors are dividing the 7.3 (and not 14.6) by the supposed 10.5 run difference between the home and road teams, they are implying that the 7.3 “extra” runs is above and beyond the road team, and not a neutral team, as I assumed. In that case, you would divide 7.3 by 50 runs, to get 15%, the influence of the umpire’s strike zone as compared to the total HFA, as reflected by a difference of 50 runs per 9 innings, between the home and road teams.

It is quite a bit different to say that umpire ball and strike bias accounts for 2/3 of the total HFA and to say that is accounts for less than 1/3! That little “mistake” changes the tone of their whole thesis, even based upon their own numbers. All of sudden the umpires, at least vis-a-vis their ball strike calls, are not responsible for a majority of the overall HFA.

#14 David (see all posts) 2011/02/25 (Fri) @ 22:01

I think that publishers - and the media - are pretty arbitrary and selective about what they determine to be fraud and/or worthy of retractions. So there’s definitely no guarantee.

But I think that wanton fraud is a no-brainer cause for a retraction. This isn’t a trivial issue like writing the wrong year or saying that a flood happened in the wrong part of China (not that those things are okay, but errors are gonna happen, we agree). No, what seems to have happened (possibly, not certainly) is that the authors literally made key data up out of thin air. If they don’t show their work and/or it can’t be duplicated, then it’s a huge error or outright fraud. Either way, a retraction should be made. Maybe they can just post something at their website - even newspapers usually do that much, and they produce dozens of articles every single day!

It’s also worth noting that this stuff is always interesting, too. Whether it’s World War 2 popular author Stephen Ambrose or Doris Kearns Goodwin, there are many writers who are never penalized for their fraud - even abject plagiarism (when other authors are keelhauled over far less). But it’s still fun to go to websites and read about this stuff.

In short, you can’t be certain that a retraction can be forced. But you can at least try. All it takes is five minutes to type up a letter with your findings and then two minutes to mail or fax it to the publisher. That’s what I did.

I don’t know why the American media hates James Cameron, but if I showed you the abject lies they’ve written about him - repeatedly - and the absolutely insane mob mentality against ‘Titanic’ before it came out, you’d probably find it really fascinating.

I guess I could take a few guess about why they hate him, but that’s all they’d be is guesses. (By the way, I’m speaking in really general terms here - it’s kind of like saying that the media hates A-Rod. Does everybody in the media hate him? No. Are many of them probably just writing what they think will be popular with readers? Probably. But it still happens.)

Cameron’s affiliation with Pellegrino was pretty loose, and he has a very, very large circle of colleagues - as everybody of similar stature in Hollywood has. I’m sure that some of his other colleagues are sleazeballs, too. When you work with hundreds and hundreds of people on dozens and dozens of projects, some of them are gonna be great people, some are gonna be scumbags.

Bill James spent years writing about how Pete Rose was probably innocent. So brilliant dudes aren’t exempt from lapses in judgment. I think that this should be common sense.

I don’t know what constitutes a “serious movie-goer”, and I don’t know that your private conversations with a few people about ‘Avatar’ should be the arbiter of whether it’s good or bad.

But what I do know is that common man everywhere has voiced his opinion in the most resounding of terms and made ‘Avatar’ the greatest box office champ ever (by a landslide). I also know that it received overwhelmingly positive reviews from critics (83% rating at RottenTomatoes.com). I also know that it’s Blu-Ray disc shattered records within days of its release.

(And, incidentally, it was also a sustained hit, being the #1 movie for seven consecutive weeks, which is almost unheard of nowadays. It wasn’t like ‘Transformers’ or ‘GI Joe’ where it makes all of its money its opening weekend and then limps out of theaters two weeks later because everybody told their friends to avoid it.)

Anyway, if you don’t like ‘Avatar’, that’s fine. It’s all a matter of opinion, obviously. But to say that it was unpopular is bogus to the extreme.

But I hope that you continue your great research on Scorecasters, and I hope that you pursue some sort of justice for this. If the authors’ core argument (the HFA chapters are the crux of the book) is premised on lies or falsehoods, it shouldn’t go unpunished, in my opinion.

If nothing else, you can edit its Wikipedia page to let people know through there.

#15 Tangotiger (see all posts) 2011/02/25 (Fri) @ 22:04

"BTW, does anyone know why the road team is ahead more often than the home team? “

Presuming this is true, it’s because the road team is always 0 to 3 outs ahead. After the first 3 outs, the road team got to bat, and the home team didn’t. The road team will have scored a run 30% of the time, so right away, they will be ahead of the home team 30% of the time after the third out in the game.

After the sixth out in the game (end of the 1st), the home team will be ahead just barely over the road team.

Basically, from the home team perspective, it’s a case of “one step back, 0.05 steps forward, 0.95 steps back, 0.10 steps forward”. They are always playing catchup.

#16 David (see all posts) 2011/02/25 (Fri) @ 22:06

Good call, TangoTiger.

As long as you’re here, can you confirm that the formula for converting wOBA into RC is:

wOBA/1.15 * PA

Thanks.

#17 Tangotiger (see all posts) 2011/02/25 (Fri) @ 22:11

Not exactly.

runs above average
= (wOBA - lgWOBA) / 1.20 * PA

That 1.20 is what it is these days.

To convert to actual runs, you add around .12 runs per PA.

That .12 is whatever the league average is.

#18 David (see all posts) 2011/02/25 (Fri) @ 22:25

Hmmm.

At the risk of being solicitous, I’ll say that I’ve got most of the forecasting systems for this season. (In the past, I just used PECOTA, but I read here and elsewhere that PECOTA was thrashed even by Marcel the past few years, so now I’m expanding to others.) I’m just trying to create RC and then RC/G as a core, baseline projection stat. But I’m having a terrible, terrible time trying this. Every time I try to calculate it, for any system, the results are always way off.

So, if there’s anybody who could assist me with generating RC for the forecast systems, I’d appreciate it. (David—at—TactilePortraits.com.)

#19 Tangotiger (see all posts) 2011/02/25 (Fri) @ 23:20

Why not simply use wOBA as the baseline stat? Its akin to OBP.

#20 David (see all posts) 2011/02/25 (Fri) @ 23:25

I like RC because of its context to the game itself (the same reason that some prefer wOBA because how it parallels OBP or EqA as it corresponds to AVG I guess) and also because I can then create individual W%. These are oldies from the Bill James Abstracts, and I like them a lot.

#21 Tangotiger (see all posts) 2011/02/25 (Fri) @ 23:35

Say someone is a .400 wOBA in a league of .335, and he has 600 PA. His runs above average is:

(.400-.335)/1.2*600 = +32.5

If the league scored 22,000 runs and the league had 190,000 PA, then we have .116 runs per PA.

The baseline therefore is .116 x 600 = 69.5 runs

This player therefore is 69.5+32.5 = 102 runs

His RC+ is 102/69.5 = 1.47 or 147

#22 MGL (see all posts) 2011/02/25 (Fri) @ 23:41

David, you sure you are not Cameron’s long lost son? You might want to get a DNA test. If you are, you might be in for one heck of an inheritance!

Seriously though…

“So brilliant dudes aren’t exempt from lapses in judgment. I think that this should be common sense.”

You are right. Mine was a throwaway comment.

“I don’t know what constitutes a “serious movie-goer”, and I don’t know that your private conversations with a few people about ‘Avatar’ should be the arbiter of whether it’s good or bad.”

Right. My experience is anecdotal and should not be generalized in any way, shape or form. Plus I am biased, as you are.

“But to say that it was unpopular is bogus to the extreme.”

Never said (or implied) that. Be careful about writing falsehoods!

#23 MGL (see all posts) 2011/02/25 (Fri) @ 23:58

Given that it does appear that the umpire home bias is greatest in high leverage situations (we think), it would be interesting to see if that is enough to create a significantly lesser HFA for games in which teams are mismatched. IWO, is there enough of a difference in the amount of time that high leverage exists in games where the teams are mismatched and those where the games are close to affect the HFA? Tango, do you think? I can test that on the games in my database - for each year, create a file of team strength based on their record for that year (it is not really necessary to regress it), and then look at the percentage of high leverage situations (say, > 2.0) for each matchup based on whether it is between two mismatched teams or not.

David, right now, I am not ready to accuse these authors of malfeasance. Other than a couple of glaring errors, much of the stuff I mention could be open to interpretation, as far as the methods and data are concerned. And there is the possibility that I made one or more computational or programming errors. It is not like I double checked all of my work. And I’ve only put in maybe 6 hours compared to their thousands (I assume). Plus, the entirety of the book seems well-researched and written in good faith, and I am pretty sure that a significant portion of it is correct.

We’ll wait and see if the authors respond to my email, and if yes, what they say.

I have a feeling, and it is only a feeling, that there is more sloppy research in mainstream books (with some data made up) due to a lack of time and effort on the part of the authors and researchers, than many people assume. In this case, and in others, it could be the fault of researchers and not the authors themselves, although of course they are ultimately responsible.

Unfortunately, I did not send the letter to either of their email addresses (I don’t know if they are publicly available). I put it in a “contact” form on their web site. It could go an assistant and that person could simply discard it or it could get lost in hundreds or thousands of emails…

#24 (see all posts) 2011/02/26 (Sat) @ 02:48

MGL #13. If the game is tied in 27% of PA, the road team is ahead in 37% and the home team is ahead in 42%; then we have 106% and a problem. Maybe in the %’s and maybe in the PA’s.

#25 Peter Jensen (see all posts) 2011/02/26 (Sat) @ 03:21

kds - MGL forgot to specify BATTER_EVENT_FLAG = true when he ran his queries for home team ahead, visitor team ahead and tied, including non batter events in those numbers. But apparently he excluded non batter events when he ran the total events and it looks like some other events as well, possibly bunts. That’s how he got more than 100%. The correct numbers are home ahead 454979: 34.6%, visitor ahead, 521048: 39.7%, tied 337193: 25.7%.

#26 MGL (see all posts) 2011/02/26 (Sat) @ 03:37

kds, there were just so many PA’s, I lost count!

Seriously, I’ll re-check.

02-08

Home team ahead

PA=442,664 (35%)

Visiting team ahead

PA=506,833 (40%)

Tied

PA= 326,691(25%)

BTW, I just sent off my letter to one of the author’s (Moskowitz) University email address.

#27 James Holzhauer (see all posts) 2011/02/26 (Sat) @ 04:01

mgl, is this your only issue with the ‘facts’ as reported in the book? i found it to be full of fuzzy mathematics and half-truths.

#28 MGL (see all posts) 2011/02/26 (Sat) @ 05:01

James, I read the whole thing and nothing else jumped out at me. That is not to say that there are not other subtle or even glaring errors. I am not proficient enough (nor do I have the data or the interest) in the other sports to test some of their assertions.

Overall, I thought it was a good book. It probably has no more errors in it than the typical mainstream book, but I am not sure about that. And that is not to justify shoddy work. Anything you put in a book should be accurate. That is simple, basic ethics and integrity. A few honest mistakes are allowed of course. In The Book, I made a mistake about the run value of a pitcher bunting in a certain situation (I think it was with runners on first and third). As far as I know, that is the only mistake in The Book and we did a lot of research. Chances are, there are other mistakes, although probably not that critical, but you never know…

#29 MGL (see all posts) 2011/02/26 (Sat) @ 06:08

I looked at matchups among teams that had similar season w/l records and those that did not, to see if the percentage of high and low leverage situations would be different, and thus if the H/R called strike percentages would be different. Here is what I got:

For games where the two teams were within 1.5% of each other in season w/l record (e.g. one team is .570 and the other is from .555 to .585)

11% of all games were like this. HL occurred 9.9% of the time and LL, 7.6%.

LL was < .1 LI and HL was > 2.0.

For games where one team was at least 18% better (in their season w/l record) than the other

8.5% of all games were like this. HL occurred 9.4% of the time and LL, 8.5%.

So, as you might expect, with two teams of similar quality, HL occurred slightly more often, .4% and LL slightly less often, .9%. I doubt that would have much effect on the overall HFA though. In fact, the total H/R called strike differential for the evenly matched teams was .6% and for the lopsided matchups, it was also .6%.

#30 David (see all posts) 2011/02/26 (Sat) @ 13:28

#27 and MGL,

I know you guys probably read this already, but the review at BP had a number of pointed criticisms, and the totality of the review was that the book was somewhere between disingenuous and just plain bogus.

http://www.baseballprospectus.com/article.php?articleid=13003

Here are some selected quotes:

”I checked team home/road stats at the NHL website, and what I found was different than what the authors suggest. It turns out that home teams outscore the visitors in even-strength situations almost as much as on power plays.”

“If the Scorecasting argument is correct, you’d expect more HFA in the late innings (where clutch situations tend to cluster) than in early innings. But the actual figures show the opposite trend.”

“There are more than a few other cases where I think the authors got it wrong. Sometimes it’s because the study they’re quoting is flawed. For instance, they’ve got a chapter on a study that says that batters hitting below .300 late in the season wind up hitting over .400 in their last at-bat because they’re motivated by their personal goal. It turns out that that isn’t true: after you adjust for selective sampling (players who reach .300 tend to sit out thereafter, which means the causation goes the other way: a hit causes the at-bat to be the last one, rather than the other way around).”

(It’s worth noting that the author immediately conceded that last point when Jonah Keri told it to him in a podcast interview.)

So it looks like the authors repeatedly didn’t run any sort of controls - as in the NHL or .300 AVG examples - and/or cherry-picked stats to support whatever argument they were advocating. (Then again, maybe the authors were the responsible one and it’s MGL and Phil Birnbaum who were sketchy in their work!)

#31 MGL (see all posts) 2011/02/26 (Sat) @ 18:16

Yes I have read the reviews on the book (Birnbaum and Jaffe). There does appear to be lots of problems with the book. It is true that we could be wrong and they right. In general, if you have to take sides, take sides with someone who has little or nothing to gain, in terms of money, prestige, etc. Unfortunately, that is not the case with this book and its authors. And of course the biggest difference between the authors and people like Phil and me is that they are not SME and we are. It is easy for us to spot likely or potential problems (for example, when we see a result like the .299 hitter thing, the first thing that comes to mind is “is this a selective sample?,” or, “that can’t possibly be true; there has to be another explanation.") and it is easy for us to properly research the issues.

#32 MGL (see all posts) 2011/02/26 (Sat) @ 18:33

It looks like Phil has uncovered another problem with the authors’ research:

http://sabermetricresearch.blogspot.com/2011/02/why-is-there-no-home-court-advantage-in.html

#33 David (see all posts) 2011/02/26 (Sat) @ 18:54

I think that I’m playing Devil’s advocate a little bit - I have an innate bias against the pop science books (and the pop science community, I guess) of the past few years. I still get aggravated when is see Freakonomics or The Tipping Point cited as some sacrosanct record of facts and insights, and I personally think that lots of the authors and the media suck-ups (Wired mag) in this community are pompous douches constantly obsessed with the myth of their own rebeliousness. More importantly, I haven’t even read the book (and I don’t plan to). So, I feel a little bit awkward championing others to critically scrutinize this.

Having qualified my comments that heavily, I’ll say that not being an expert in “SME” (is that a synonym for sabermetrics?) is no excuse for their incompetence. The authors clearly had access to lots of great minds and, since the book apparently presents itself as science, they could’ve just offered it up for some sort of peer review. (I’m a huge critic of the peer review process, but still!)

Then there are other issues. For instance, with the “more .300 hitters than .299 hitters” thing, Bill James did (what I think was) a great job of this....three years before Scorecasters or the stupid NY Times article. I remembered being blown away by the article a few years back - and I think it was far better researched and written than what I’ve read of Scorecasters. The authors had to have known about this, definitely the dude who’s “scientific” paper they were premising the chapter on had to have. (I’m kinda guessing that he shamelessly ripped James off, to be honest.) Yet the Baseball Prospectus review writes that Bill James isn’t even mentioned in the book, let alone credited with spotting this phenomenon before them! (And James did it so much better - he noted that RBI figures appear to be targeted the most of all.)

I was also going to then link you Birnbaum’s article about their insane off-base reasoning with free throw shooting, but I see that somebody just did.

Also, I think that you’re claim that it’s easier for us to research these issues is totally off. In the first place, we’re not being paid. (If somebody wants to pay me, I’ll be happy to do so!) In the second place, even if that were true, then why didn’t the authors, their editors, their fact-checkers, or their publisher contact this site or THT or BP or Fangraphs or a hundred other people before they published the book? We’re just talking about sports, but what if it was a book about, for instance, nutrition for cancer survivors? Or about safety for industrial workers? Or about how to save money on your mortgage? Or anything else that involves more substantial issues? Is it okay to say, “Ah, who cares about lies and errors? Just sell the damn thing.”

We all have a responsibility to be honest and diligent. These authors were clearly dishonest and/or incompetent. That’s a bad thing. It’s a serious thing because accepting it and brushing it off sets the precedent that there are no repercussions for lies and incompetence. Like I said, if there are books about more substantial things written with this much self-assuredness and it had this many “mistakes”, it could do serious harm in the real world.

(This is kind of a strange comparison, but I remember reading a story a few years back about how the singer Madonna always propagated this “origin myth” about how she moved to Manhattan, poor, young, and alone. But through only her hard work and special force of will, she said, she was able to survive in Manhattan and become a famous millionaire. Many, many, many young girls who idolized her were wide-eyed and entranced by this story, and they tried to copy what she did....and their lives were usually pulverized because, of course, the odds of a teenage girl being able to support herself in Manhattan are tiny. The story was then brought full circle when it was revealed that Madonna just made the story up. In fact, she came from an upper-middle-class family, she had an existing social circle in New York City, and her dad subsidized her on her move. At a glance, Madonna’s little fable would probably be dismissed as a harmless exaggeration - maybe she even deluded herself into believing it. But you know what? That seemingly innocuous white lie was literally the cause of thousands of wide-eyed teenage girls having their lives totally devastated. The point is....most of this pop culture stuff might seem trivial, and I’m sure that it usually is. But whether it’s lies about baseball or the start of a pop singer’s career, lies can contaminate the world in really insidious ways. That’s what I believe.)

#34 MGL (see all posts) 2011/02/26 (Sat) @ 20:22

David, I don’t disagree. I am typically as incensed as you are when someone (author, pundit, commentator, journalist, whatever) spins or distorts the truth. I am just trying to be a bit more solicitous that I normally am. One reason is that I am trying to start a discourse with these guys, and ripping them is probably not the best avenue toward success.

#35 David (see all posts) 2011/02/26 (Sat) @ 21:40

MGL,

That’s totally cool, and it makes more sense, too.

But if the authors or editors don’t provide a legitimate response (besides the patronizing, then I don’t think there’s anything wrong with ditching the honey and pulling out the vinegar. (Or, actually, the water, because I believe that the goal probably isn’t to be mean, but to advocate the truth.)

And, to be clear, I have a kind of negative agenda here: I take a schadenfreude delight in seeing some of these pop media figures get called out on their routine malfeasance. Take this saga about Doris Kearns Goodwin:

slate.com/id/2061056/

(And don’t feel pity for these media darlings. Kinda like corrupt prosecutors or murdering cops never suffer any serious repercussions besides maybe, maybe being suspended or losing their jobs, no matter how evil their offense, these people hooked in with the main media just get a little bit red in the face for a few days, issue the now-stock, “I don’t apologize for anything. Critics don’t matter, I’m the one in the arena” phony tough guy crap, and then resume life as usual and all gets forgotten. Same thing would happen with Scorecasters. Guaranteed.)

#36 MGL (see all posts) 2011/02/26 (Sat) @ 22:34

David, trust me, you and I are cut from the same cloth with regard to this issue!

There is WAY too much deception in this world for my taste!

If I don’t get a satisfactory response from these guys, I may hit you up for some advice on how to proceed.

I expect 1 of 3 responses:

1) Thank you for your input, however, we stand by what we wrote.

2) Yes, we are aware of some inaccuracies in the book, and we are working on correcting them. Etc.

3) No response.

#37 David (see all posts) 2011/02/27 (Sun) @ 17:53

MGL,

Keep updating, please. In fact, it might be fun to read reviews from other sabermetric-style sports sites (beyond baseball) and see whether there are any issues there.

Just to be clear about myself, I don’t work for any sort of publishing group and I have absolutely zero mainstream media connections. I’m definitely not a professional watchdog or anything. The issue with fraudster Charles Pellegrino was just one of those random things that just emerged. I investigated it myself over the course of two or three days (although those were very busy days!) and, after writing my blog article, I sent e-mails and faxes to any media parties who I thought might be interested: his publisher, critics who’d praised his books, etc. I think I sent six letters in all. (Plus many of the people who were assisting me and also investigating him.)

So I think that the criticisms of Scorecasters are interesting and, assuming there’s no legit explanation, they should warrant retractions. But I’m, ya know, not somebody with any special power to make that happen.

#38 MGL (see all posts) 2011/02/27 (Sun) @ 20:31

David, sounds good. I’ll keep everyone informed via this blog and if I need any help, I’ll drop you an email.

If more of the general public took an interest in exposing fraudsters (again, I am not in any way, shape, or form lumping the Scorecasting authors into this group), we would reduce the amount of deception in the world considerably. It is a noble deed.

#39 David (see all posts) 2011/02/27 (Sun) @ 20:50

MGL,

It is a lot of fun, and I kinda secretly hope that there is indeed fraud and lies involved with this case, because it is kinda fun to see.

For whatever it’s worth, though, (probably not much), I’ll repeat what I wrote earlier - and I just say this for philosophical reasons - that I personally found myself kind of spent and depressed after the Pellegrino thing had run its course. I knew that it was good and right and it was a lot of fun. But it is definitely better and more challenging to create than to destroy, and so I think that it’s best if people just make “debunking” or fraud investigation a periodic endeavor, not the automatic nature of their writing and research. Again, that’s just based off of my little experience.

#40 MGL (see all posts) 2011/02/27 (Sun) @ 22:45

There are two ways to look at it, David. Exposing someone else’s lies is a manner of creating, no less noble a thing to do then to independently create something. In my opinion at least. You do what your heart, soul, and mind tell you to do, and as long as the overall cause is for the good, there is nothing to feel bad about. I assume that your primary motivation was not to destroy this guy Pellegrino, even if you may have felt anger and betrayal towards him, which is a natural reaction, BTW.

With these Scorecasting guys I have to resist the urge to feel anger and resentment. After all, I work very hard at trying to get at and present the truth, and when someone else does not, I sometimes feel betrayed even though it has nothing directly to do with me.

In any case, this is not a mission of mine. I had an interest in some of their findings and I wanted to find out if they were accurate. It looks like some are and some are not and I think that they or someone else should eventually correct the record.

#41 MGL (see all posts) 2011/02/28 (Mon) @ 04:10

Here is one of many examples of BS that the media (MSNBC, etc.) perpetuates:

http://www.msnbc.msn.com/id/41783544/ns/technology_and_science-science/

At least the author of this article clears things up (as if we needed someone to tell us that the boy is not “magnetic"), other than the fact that he doesn’t know what the word “adolescent” means:

“Sometimes (as in the case of 7-year-old Bogdan) it’s because the person is an adolescent and has not reached puberty.”

#42 (see all posts) 2011/03/02 (Wed) @ 20:05

If anyone’s still reading, I ran these splits at baseball-reference.com for 2010:

High leverage, home team: OPS .749
High leverage, road team: OPS .701

Low leverage, home team: OPS .749
Low leverage, road team: OPS .704

This does not appear to be consistent with the Scorecasting (and MGL) finding that there is a bigger HFA (due to pitch-calling) in high-leverage situations. Any idea what’s going on here?

#43 MGL (see all posts) 2011/03/02 (Wed) @ 23:04

Phil, well I definitely confirmed that the H/R “called strike and ball” bias is much higher in HL leverage situations. That doesn’t mean that the other aspects of HFA are, especially those that are non-umpire related.

Plus, without controlling for the pool of pitchers and batters in each group (they could easily be different for the H and R teams), I would not infer too much from those numbers.

So I don’t see the problem necessarily…

#44 (see all posts) 2011/03/03 (Thu) @ 00:02

Right, but something must be going on ... different pitchers seems like the most likely. Or, maybe IBB are making the high-leverage look worse than it should.

Interesting.

Page 1 of 1 pages

<< Back to main

THE BOOK--Playing The Percentages In Baseball

Friday, February 25, 2011

Scorecasting review, part II

Latest...