THE BOOK--Playing The Percentages In Baseball

I would be so much more interested in how his list compares to other people’s list.  Just looking at his list, I have no idea whether he was good or bad.  No idea.  Granted, we can use his list to estimate the value of his prospects, but that is all we can do.  I would also want to know how his players compare to the average player of the same draft round.  For example if I can do just as well as he can by just listing players on my “prospect list” who were drafted in the first couple of rounds, then he adds nothing to the mix, right?  I also want to know how he does as compared to strictly using players’ minor league stats, either raw or as MLE’s.  Again, if I can do as well or better without ever seeing or knowing anything about these players, then what is he providing?  Even more importantly (for me), I want to see some kind of rate statistic so that I can see how good a player is when he did play.  To some extent, that gets rid of some of the noise ssociated with injuries.

So, basically, this data means nothing to me other than I can use it to approximate the value of his prospects.  Then again, the average value of his prospects might be the same as the average value of anyone else’s, a random player from the same average draft round, a list of players with good MLE’s, etc.

This is a start, but I think it is like 1% of what I want.

I never like linking to my stuff, because I generally don’t see much value in it (especially for you guys) and it always seem annoying to do on other blogs ... that said ... click on my name for my little attempt to do something similar to this. (hope nobody cares)

I started with Baseball America’s 1990 list and have only done 91 since then. If anyone thinks there’s any use for it, I’d keep going.

Anyway, it was a simple as looking up career WARP for each player and putting it in a spreadsheet. I know the flaws of WARP, but it was an easy way to get career value.

Also, is it okay to look at the correlation between the player’s rank and career WARP? I did, anyway ...

I’m not sure if career value is really what we should be after or not. I also didn’t split it up pitchers and hitters, which I could do in the future.

Let me know if you guys have any thoughts.

MB, I think it is good stuff.  I would like to see you combine all years in your database (BTW, can you send me the database of the BA tp 100 for each year) to get larger sample sizes, maybe from 1990 to 2002 or 2003.  The problem of course with using later years, is that you have players whose careers have not yet ended so you can’t combine them with players from earlier years.  For the later years, I would somehow try and “complete” their major league careers so you can combine apples with apples.

I would also like to see the same thing done with a rate state, maybe WARP per PA (or EQA or mlvr) or something like that.  You can only include players with some minimum amount of MLB playing time.  Unless someone is interested in $ value for a player’s career, which is quite useful of course fo prospects, I generally hate looking at career numbers like WS and WARP.  I usually like to know how good a player is.  For that you need a rate stat.

And see my other comments above about what I would like to see when working with these prospect lists.  As you mention in your article, unless you have something compare it to, just looking at how one list does in terms of career WARP or something like that tells us nothing about how good or bad or useful these lists are.  Just because Sickels or anyone else finds that his top 50 or top 100 players average 10 or 20 or 1000 WS in the major leaguesd is NOT putting their money where their mouth is, unless we know how some other list or methodology does as well, as I point out in my first post.

Good job, MB, and if you can send me that BA prospect database (as many years as you have), that would be appreciated!

Thanks, MGL. Unfortunately, I was just doing it and posting it as I went. So I only have 90 and 91 done. I think I’ll start working on it again this weekend or in the very near future. When I get more years I’ll be sure to get it to you.

And I think I’ll record EqA and PA’s this time, along with WARP. After I get done with BA, I guess I’ll move on to Sickels, BP (how far do they go back?), and any others ... then as you say, we can compare the different prospectors.

I think we should only care for players until age 29 or so.  That is, all we really care about is how a player performs until he’s a free agent.  After that, we really don’t care, because a team will not be able to tap into any value at all.

Philly at SOSH did the most extensive of all such studies that I’ve seen:
http://www.google.com/search?q=site:sonsofsamhorn.net+philly+"Draft+Summary"&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a

It’s not well-organized, but there’s tons in there.

Rany at BP also did a super bangup job:
http://www.baseballprospectus.com/article.php?articleid=5152

These two articles definitely require more play.

Here’s a couple of Birnbaum posts on the subject:
http://sabermetricresearch.blogspot.com/2007/11/value-of-mlb-draft-choice.html

http://sabermetricresearch.blogspot.com/2007/11/rany-jazayerlis-mlb-draft-study.html

I looked at BA’s top 100 list from 1990-1999 to come up with player value in my article here:
http://www.hardballtimes.com/main/article/the-bright-side-of-losing-santana/

While I have values for a prospect’s first full 6 years in the majors, I do not have a rate stat.  I would be happy to expand my study for another list if someone knows where to find data for other prospect lists in the 90s.  BA is really the only one I could find that was easily available.

I was going to suggest that Victor seems to have done something like this for BA.

I’ve been meaning to go back and re-do the comparison I did a few years ago.  If/when I do, I’ll definitely keep mgl’s rate stat suggestion in mind.  That’s a great idea.

It’s tough doing a good comparison simply because there aren’t a lot of epxert lists that go back very far.  BA goes back to 1990 and while they have to stand behind their institutional work, I’m sure they beleive that they’re much better now than they used to be.  I tend to think that they have improved, but haven’t looked too closely.

BP only started doing thier list - which was just a top 40 - in 1999.  From 1999-2006 the list was written by Rany Jazayerli in consultation with the other authors, but it was his baby.  I think the last couple of years there was some PECOTA input too.  But now the last 2 lists have been done by Kevin Goldstein.  And again, I’m sure BP thinks they’ve improved, but certainly however good the BP lists used to be isnt’ relevent anymore.

Sickles goes back pretty far, but some of those are in the old STATS books I don’t have and he’s also just recently switched to separate lists for hitters and pitchers.

My preference is to not do quantititive prospect studies unil all prospects hit 6 yrs of service time so that means really only the BA lists from 1990 to 2000/01 can be fully evaluated.  And there’s not a whole lot to compare them to.

I don’t have a link, but I did do a comparison of BPs Top 40 from 1999-2001 to BA’s Top 40s.  The totals are in WARP3 from 2004 so it looks like I did it the winter after 2004, which of course, is way too early, but those were the earliest BP lists.

Here are the average slot WARP3 from the eight 5-slot intervals that make up a Top 40.

BP:
#1-5: 15.1
#6-10: 15.9
#11-15: 7.5
#16-20: 11.6
#21-25: 8.0
#26-30: 13.5
#31-35: 7.2
#36-40: 6.4

BA:
#1-5: 14.8
#6-10: 10.5
#11-15: 15.6
#16-20: 7.2
#21-25: 5.7
#26-30: 9.0
#31-35: 10.3
#36-40: 9.0

BP comes out a little better, but that could easily be explained by it having a younger cohort of prospects. 

Just quickly scanning the 2001 lists there are several players who are in the 5-7 yrs of service time range.  So I really don’t think you can do much with more recent lists which means you can’t judge BA circa 2008 or BP with Goldstein and PECOTA.

At the time I was more interested in players that made one list or the other.  Roughly 2/3 of the Top 40s were the same players.  The BA only players were more successful than the BP only players.

Btw, Tango thanks a bunch for posting the service time data thru 2007.  I’ve been updating my draft stuff and that was a big help in establishing pre and post FA production.  My data is far from perfect I’m sure, but I’m much happier with it now.

Tango, can you fix that SOSH link.  It takes me to a not very productive google search.

I am much more interested in how BA, BP, Sickels, etc., do versus a ‘generic” list like by draft order or by minor league stats alone.

Any idiot can come up with a reasonable list that is going to eventually get lots of win shares or WARP just by looking at the “best” players in the minors (by stats) tempered by their draft pick status.

And anyone’s list will have the better players at the top and the worse players at the bottom if you make sure that the players at the top have better stats.

Aother way to test whether these guys add anything to the mix is by doing controlled tests, such as you take players at the top of the lists and compare them to players at the bottom who have the same statistical projections based on MLE’s.  And then you swich that around.  Players who have the same rankings but different statistical projections.

My guess is that you will find that a combination works best (stats plus scouting - who woulda thunk?).

Just looking at these isolated lists or even comparing one person to another (like BP and BA), which is problematic, does not tell me anything, other than how much value I can expect from a prospect who has a certain rank, which is no small thing of course.

MGL, it would be mighty helpful if you could generate this every year:
http://www.tangotiger.net/MLE.html

So, that would be the best way to compare if the scouts have anything more to add over a pure performance analysis.

I was just going to ask - is there a publicly available repository of historical mles?  I can’t generate them myself, but would eventually be interested in making that kind of comparison.

On a semi-related note, I should mention that one of the changes that Nate Silver made to PECOTA this year is to include draft status and/or size of amatuer bonus.  He hasn’t gone into it in depth because he hasn’t done a full PECOTA rollout, but he does seem to beleive that that kind of indirect scouting information is beneficial to separating out minor league players with similar statistical performances.

I’m curious to see how much it adds.  One effect of that change is that he know has PECOTA forecasts for June draft picks like David Price who have never played a pro game.  And I don’t think he uses college stats at all.  It’s a forecast completely based on prospects with similar draft status.  I’d be pretty surprised if there was much value in that.

I think it would have tremendous value.  He’s basically including the scouting community’s overall assessment of the player.

I’ll see if I can send Tango my MLE’s for all years that I have (what years are needed?).

I think when doing MLB projections from MLE’s, draft and prospect status would be extremely helpful.

And although it is not really necessary, an MLB projection for a newly drafted player based on draft order only would be accurate withing its limitations (IOW, a 1st round pick is going to have a much bette projection than a 5th round pick, etc.

Any computer techies on this board?  My wireless keyboard was missing letters as I typed, so I bought a new one.  That didn’t work, so I bought a new wired one.  That still doesn’t work, so it is obviously a “computer” problem.  Any suggestions?  I am using a Toshiba laptop with Windows XP.

What is it that you are requesting from your computer, and what response are you getting?

#14, ??

Do you mean from your SOSH link?

Here are the first few hits I get from Google:  Is this what you intended to link to?

Web Results 1 - 10 of 14,500 English pages from sonsofsamhorn.net for philly. (0.26 seconds)

philly sox fan - Viewing Profilephilly sox fan doesn’t have a personal statement currently. Personal Info. philly sox fan. SoSH Member. Age Unknown years old. Gender Not Set ...
sonsofsamhorn.net/index.php?showuser=35 - 53k - Cached - Similar pages

March 10 vs. Philly - Sons of Sam HornIn a delay because the Phillies are caught in a major traffic jam. They expect to start at like 3pm. This post has been edited by SoxScout: Mar 10 2006, ...
sonsofsamhorn.net/index.php?showtopic=4253 - 74k - Cached - Similar pages

Sons of Sam Horn > NCAA Lacrosse Final 4 (Philly - May 27-29)Full Version: NCAA Lacrosse Final 4 (Philly - May 27-29) · Sons of Sam Horn > Other Sports > General Sports. PT Sox Fan. May 22 2006, 03:23 PM ...
sonsofsamhorn.net/lofiversion/index.php/lofiversion/t5923.html - 12k - Cached - Similar pages

Sons of Sam Horn -> S.o.S.H. Geekage9 · philly sox fan, 6057, 17th December 2005 - 03:44 PM Last post by: Tudor Fever. No New Posts. Career path of Sox 2002 DSL/VSL players ...
sonsofsamhorn.net/index.php?showforum=28 - 87k - Cached - Similar pages

That is:
1. what is it that you are doing
2. what is it that is happening
3. what is it that you expected to happen

Is it simply a matter that you are typing the letter Q, there is no activity on the screen (no Q, no space, nothing), and obviously, you expect the letter Q to appear.

As well, does this happen with multiple applications, or just one?

Type this in Google:

site:sonsofsamhorn.net philly “Draft Summary”

For some reason, the link I have ignored the second search term.

It appears that it is only in MS IE (7.0).  Internet Explorer that is.  Which is where I do most of my typing.  I think that is the case, but I am not sure.  I am having other problems with IE as well (like I cannot move favorites around when I go to “organize favorites").

As I type a sentence, it just randomly misses letters, no matter how fast I type or how hard I press the keys.  And I have tried at least 3 different wired and wireess keyboards, so it has nothing to do with that.

I think everything is OK when I am using, for example, Word.  I am going to type a few sentences and not correct the missing letters:

OK, I am typing a sentence now.  I am typing pretty slowly and I am pressing hard on all the keys.  I am not going to correct these sentnces so you can see what I mean.  I was not looking nd it looks like it missed only two letters, one in the word “sentence” and the other in “and”.  But that is what I mean.  Smetimes it misses several letters per sentence.  See, it just missed the “o” in sometimes.  Maybe it does depend on how fast I type - I am not sure.  am going (itmissed the “I” and a space) to try and uninstall IE7.0 and reinstall 6.0 and see what happens.  This only started a week ago andI have had IE 7.0 for a few months.  It is very frustrating!

I am using Firefox.  Try downloading that, and see what it gives you.

Yup, it was definitely IE.  Friggin’ Microsoft!  One of my favorite all-time scenes in a movie (if I can be a little masochistic) is from the South Park movie when Bill Gates comes out and they just shoot him in the head.  How many computer users would like to do that (figuratively of course)?

I uninstalled IE 7.0 which automatically puts back IE 6.0.  But then I read on some forums that 6.0 becomes unstable.  Sure enough, after I uninstalled 7.0, my 6.0 was all screwed up.  Completely unworkable.

So I am now using Firefox and everything is working fine.  I imported all of my bookmarks, passwords, etc.

Anyone want to buy some keyboards (lots of them)?

Use them as some prize in some contest