RotoGuru Baseball Forum

View the Forum Registry

XML Get RSS Feed for this thread


Self-edit this thread


0 Subject: Comments: A lengthy review of Covering the Base

Posted by: Sludge
- [16109168] Tue, Dec 19, 2006, 14:40

Please place comments, criticisms, arguments, etc. here. Thank you.
1Steve Biz
      ID: 1511371623
      Tue, Dec 19, 2006, 22:00
Not very lengthy... I'm guessing there's more coming? Of course, there doesn't really need to be. These guys sound like a couple of babies/idiots/pric**.
2Sludge
      ID: 45541422
      Tue, Dec 19, 2006, 22:05
Yes, more coming. Forgot the "To be continued..." to close the post.
3Steve Biz
      ID: 1511371623
      Tue, Dec 19, 2006, 22:09
"since the two stats don’t share a common denominator, they can not be expected to yield a telling number when added together"

So easy to disprove it's absurd. Idiots writing for idiots.
4Sludge
      ID: 45541422
      Tue, Dec 19, 2006, 22:36
I was going to include an example that went something like this:

The Poisson distribution has the property that the mean and variance of the population are both equal. Thus, for a sample gathered from a Poisson distribution, the sample mean and sample variance are both unbiased estimates of the population mean (commonly refered to using the Greek lowercase lambda). The formulas for these two look nothing alike yet taking the average of them produces an unbiased estimator for the mean and variance of the Poisson. By the logic of the authors, I'm betting they would argue that this would be a ludicrous thing to do under most circumstances. Under the Poisson, however, it is perfectly valid.

I didn't include it because it was a bit on the technical side, and doesn't quite fit the bill as well as I would like since there'd be someone who would look at the formulas for the two and notice that the denominators are n (for the sample mean) and n-1 (for the sample variance) and argue that they're close enough; nevermind that one numerator is the sum of the observations (sample mean) and the other numerator is the sum of the squared deviations (sample variance).
5Razor
      ID: 211341921
      Tue, Dec 19, 2006, 23:22
Not sure if I would have made it through an entire book with such a pathetic critique of OPS as you posted in the other thread. Kudos for that.
6blue hen
      Leader
      ID: 710321114
      Wed, Dec 20, 2006, 10:47
Isn't the denominator the same? Both OBP and SLG are "out of" 1, right?
7Razor
      ID: 2107611
      Wed, Dec 20, 2006, 10:52
No, SLG is out of 4.000. OBP is out of 1.000.

It's still a silly argument. Anyone that has even the loosest grasp of what OPS is about knows that it's just a quick and dirty tool and is not to be used as an exact measure of anything.
8Sludge
      ID: 45541422
      Wed, Dec 20, 2006, 12:04
The "denominator" is the bottom number in a division. The "numerator" is the top number. What bh and Razor are referring to are the "range" of the OBP and SLG (0-1 for OBP, 0-4 for SLG).
9Seattle Zen
      ID: 49112418
      Wed, Dec 20, 2006, 12:18
How about actually talking about the merits of the results OPS is trying to express? Would they have been happier with:

(Total Bases + BBs + HPBs/Plate Appearances)

If so, how different is that from OPS? Very little, do the math. If they can't grasp the value of measuring total bases with on-base percentage, then they are worthless.

Razor is right, OPS is a shorthand that's easy to grasp and calculate.
10Nerfherders
      ID: 64532914
      Wed, Dec 20, 2006, 12:48
So far, so bad, it seems to me. OPS is a quick and dirty tool for seeing how good or bad players are relative to eachother. But I certainly wouldnt say a player is 'X good' based purely on OPS. I would have to break down his numbers to do this. If I am doing draft prep and I see two players with the same OPS, ie within 20 points of eachother, I look at all of their stats, weigh OBP heavier, see who else is in their lineup, etc, etc. But nobody would be able to convince me given similar plate appearances that a player with say 50+ less OPS than another player is the better of the two.
11blue hen
      Leader
      ID: 710321114
      Wed, Dec 20, 2006, 13:09
I agree with Zen agreeing with Razor. OPS is generally simple to calculate (I even measure my own success in beer-league softball with OPS). Calling it a complex formula is not really correct.

Nerfherders, what are your other criteria? I mean, if Albert Pujols is up, down one run in the bottom of the ninth, I'd much rather have Scott Podsednik as the man on second than Jim Thome. Clearly, there shouldn't be as much weight on that (Thome will reach base and go farther per hit than Podsednik, both by a mile), but there are certainly other things to consider. What are they? And isn't it possible that Bill James has covered some (or all) of them?
12Sludge
      ID: 16109168
      Wed, Dec 20, 2006, 13:11
So far, so bad? I never argued that OPS is a good diagnostic tool. Only that it does, in fact, estimate what it is purported to estimate (how well or poorly it estimates it is not something I'm interested in debating). All estimates have variability, but the fact that you indicate in a roundabout way that there is some kind of threshold beyond which you would look at the OPS of two players and conclude that Player A is, in fact, better than Player B shows that you agree.

In case it's not clear, the point of this exercise is not to validate sabermetric techniques. There will be some of that when I feel it is warranted, but typically only from an overhead view (e.g. the goal is perfectly valid, but nothing to be said about the methods used to achieve that goal except that some of the criticisms they have of the methods are absurd). bh asked me to critique their critique, and believe me, there's plenty to critique. Or don't believe me and wait until I am able to post more and make up your own damn mind.
13Sludge
      ID: 16109168
      Wed, Dec 20, 2006, 13:13
Calling it a complex formula is not really correct.

I'm not aware that anyone has, not even the authors of Covering the Bases.
14Perm Dude
      ID: 31113208
      Wed, Dec 20, 2006, 13:17
#12: Hmmm. I guess comments which are critiques themseves aren't welcome? I don't think you need to jump down Nerf's throat, even when you disagree. I'm assuming it was not the point of this thread to only welcome comments which merely complement your own points.
15blue hen
      Leader
      ID: 710321114
      Wed, Dec 20, 2006, 13:49
Actually, I don't think Nerf was aiming at Sludge. Did he mean so far, so bad about the book authors? That's what I got out of it.
16Sludge
      ID: 45541422
      Wed, Dec 20, 2006, 22:37
#12: Hmmm. I guess comments which are critiques themseves aren't welcome? I don't think you need to jump down Nerf's throat, even when you disagree. I'm assuming it was not the point of this thread to only welcome comments which merely complement your own points.

I don't see how I'm jumped down anyone's throat, PD. I do not see the tone of my post being anywhere near that level. Perhaps the start of Nerf's comment was not aimed at me, but my point remains the same, and I would have likely made it regardless given the direction the comments were headed. While everyone else can feel free to debate the efficacy of sabermetric techniques here, I do not want anyone to misunderstand my goal here, which is only a thorough review of Covering the Bases. If they make some valid points, I will be sure to highlight them. But honestly, I have yet to see many worth mentioning.

Anyone who's been around the politics forum for a while will know that my particular forte' (call it my schtick if you prefer... it's apt, and I make no bones about it) is often not taking a side in an argument but in analyzing the arguments themselves. Despite what you may think of me for it, I take great pleasure in it, and it's why I made the offer to bh in the first place. What so often happens, however, is that because I'm picking apart an argument I get pigeonholed as falling squarly on the other side of that argument, which as often as not is not the case. I am trying to head that off at the pass here by stating quite clearly and quite often that I am not here to defend sabermetric techniques.
17Perm Dude
      ID: 31113208
      Wed, Dec 20, 2006, 22:41
I don't think you have any idea of what I think of you, Sludge. Certainly not enough to use it as a point of reference to defend yourself.
18Sludge
      ID: 45541422
      Wed, Dec 20, 2006, 22:41
And for the record, critiques of my critiques are quite welcome. I will respond to them, however, explaining why I think they're wrong or misplaced if I think they're wrong or misplaced. I will also tip my cap (wearing a cap might help) to those which have merit.

I promised bh I was going to keep the discourse academic and even-tempered on my part, and I am. Nobody should take it personally if I tell them I think they're wrong; I'm not here to walk on eggshells.
19Sludge
      ID: 45541422
      Wed, Dec 20, 2006, 22:43
Sorry, PD, that was meant to read "Despite what anyone may think of me for it...". I probably wrote "you" instead because you most of all recognize my schtick most readily.
20Motley Crue
      Dude
      ID: 439372011
      Wed, Dec 20, 2006, 22:54
I brought this up a long time ago (maybe years), but I don't feel like I received a satisfactory explanation. It seems to me that the most efficient way to calculate a player's value on offense in terms of how often he gets on base and how well he can slug would be to multiply OBP and SLG%.

Why do we add them instead?
21Perm Dude
      ID: 31113208
      Wed, Dec 20, 2006, 22:56
Fair enough, Sludge. Critique away.

[BTW, I think it is unfortunate that you are really getting your first "professional" look at sabermetrics through such schlocky writing. The fact that these authors are real idiots doesn't validate what sabermetrics tries to do. I would love to have a critical eye on Bill James directly.]
22The Beezer
      Leader
      ID: 191202817
      Thu, Dec 21, 2006, 06:55
Nice read so far, Sludge. I'm a big fan of James' work and find his work inspiring when I'm trying to find ways to get useful information from the data generated by a technical support desk. Despite that bias, I think an overcorrection is due in terms of recognizing the limits of statistical analysis in its present form and how it applies to the real world. I had heard about this book, but I assumed that detractors were at least somewhat overstating the case.

I assumed incorrectly. The authors of this book should have the number system taken away from them before they perpetrate further crimes against humanity.

MC, I think the primary reason people add the figures instead of multiplying them is just that it's easier to do in your head when your looking at a stat sheet with OBP and SLG on it. I think regression studies have shown multiplying the terms has more predictive power than adding them (and there are certain factors for each component that do better yet). I don't have a link handy to back this statement up, so take it for what that's worth.
23blue hen
      Leader
      ID: 710321114
      Thu, Dec 21, 2006, 11:06
Excellent work so far, Sludge. The second installment is fantastic.

And I agree with Perm Dude in #21 - this is a terrible introduction to sabermetrics. But the benefit for the rest of us is that we are getting analysis from someone who has no idea who Bill James is and is about as unbiased as it gets.
24Seattle Zen
      ID: 49112418
      Thu, Dec 21, 2006, 13:45
From the book review:
However, Pujols would have to go 1 for 2 with a home run and two walks, without grounding into a double play, in order to “create” two runs (actually 2.1085 according to the runs created formula). Does it matter if the home run was a grand slam or a solo shot? Of course not. Runs batted in are a mere triviality in Bill James’ runs created formula. Pujols could have driven in anywhere from one to seven men (had the walks come with the bases loaded and the out been a sac fly) yet, through the runs created formula, he is credited with two runs, regardless of the hard factual numbers. Finally, we have to agree with Bill James. This is illogical. (pg. 17)

These guys seem to pride themselves on their baseball knowledge, yet they don't even know that a sac fly is not an at bat. So sad.
25Perm Dude
      ID: 171110219
      Thu, Dec 21, 2006, 13:55
Is there more to the quote, Zen? The quote doesn't say that a sac fly is an AB, only that it is an out.
26blue hen
      Leader
      ID: 710321114
      Thu, Dec 21, 2006, 15:09
Zen is right - if a guy goes 1 for 2, the out is NOT a sac fly. Unless he was 1 for 2 with a sac fly in addition.

However, their argument is unaffected - an RBI groundout (or 2 RBI groundout!) would have the same affect.
27Perm Dude
      Dude
      ID: 030792616
      Thu, Dec 21, 2006, 15:19
Ah, I see what you mean. Thanks.
28Nerfherders
      ID: 64532914
      Fri, Dec 22, 2006, 03:46
I think you misunderstood me. My reference was to the book's authors and not Sludge's review. I am in agreement here.
Rate this thread:
5 (top notch)
4 (even better)
3 (good stuff)
2 (lightweight)
1 (no value)
If you wish, you may rate this thread on scale of 1-5. Ratings should indicate how valuable or interesting you believe this thread would be to other users of this forum. A '5' means that this thread is a 'must read'. A '1' means that this is a complete waste of time.

If you have previously rated this thread, rating it again will delete your previous rating.

If you do not want to rate this thread, but want to see how others have rated it, then click the button without entering a rating, or else click here.

RotoGuru Baseball Forum

View the Forum Registry

XML Get RSS Feed for this thread


Self-edit this thread




Post a reply to this message: (But first, how about checking out this sponsor?)

Name:
Email:
Message:
Click here to create and insert a link
Click here to insert a random spelling of Mientkiewicz
Ignore line feeds? no (typical)   yes (for HTML table input)


Viewing statistics for this thread
Period# Views# Users
Last hour11
Last 24 hours11
Last 7 days55
Last 30 days1111
Since Mar 1, 20072380809