Using BP's "Star System" to Rank the Farm Systems

by Satchel Price on Mar 15, 2010 1:57 PM EDT in Prospect Perspective

Earlier this year, dougdirt did a nice compilation, ranking the farm system's by organization using John Sickels' letter grades and Victor Wang's prospect valuation work, translated to dollar values by the good men of Beyond the Box Score.

Considering that Kevin Goldstein of Baseball Prospectus uses his own ranking system to label prospects, known as the "star system", I thought that it would be interesting to see how the rankings play out with those grades. Goldstein uses two-, three-, four- and five-star valuations for his prospects, with the four- and five-star prospects all being included on Goldstein's Top 101 prospect list.

In order to translate all of the sub-four star rankings into dollar values, I had to correspond Sickels letter grades to each Goldstein star ranking. I decided to settle on this fairly simple, junk math solution: three-star prospects in the upper half of the Top 11 get B- rankings, three-star prospects in the bottom half of the Top 11 get C+ rankings, and all two-star prospects receive C rankings.

The top five included some surprises: 1. Tampa Bay 2. Minnesota 3. Cleveland 4. Atlanta 5. Chicago (NL)

The total results after the jippity jump.

1. Tampa Bay

2. Minnesota

3. Cleveland

4. Atlanta

5. Chicago (NL)

6. Texas

7. Baltimore

8. Oakland

9. Pittsburgh

10. Florida

11. San Diego

12. Detroit

13. Boston

14. San Francisco

15. Washington

16. Kansas City

17. Los Angeles (NL)

18. New York (AL)

19. New York (NL)

20. Los Angeles (AL)

21. Toronto

22. Seattle

23. Milwaukee

24. Philadelphia

25. Colorado

26. Cincinnati

27. Chicago (AL)

28. Houston

29. St. Louis

30. Arizona

And some notes:

-Obviously, depth doesn't have a great deal of value here, since we're really only looking at the quality of only each organization's very best prospects.

-The translation system (letter to star) was far from perfect. It doesn't necessarily make perfect sense that the No. 5 prospect in one organization should be worth more than the No. 10 prospect in another, clearly superior farm system. But alas, this was more art than science.

-Tampa Bay had far and away the best system. They accumulated $123.2M in value, nearly $25M more than second-place Minnesota. Arizona and St. Louis were each more than $7M worse in value than No. 28, Houston.

-Teams I was surprised to see so low: Cincinnati, Texas, Boston

-Teams I was surprised to see so high: Minnesota, Detroit, New York (AL)

Overall, I was relatively happy with the rankings. I'd certainly love some suggestions about how to improve them, and more specifically the translation of letter grades to star grades, if anyone has any. And the full spreadsheet is available below, if anyone wants to take a looksy.

Goldstein Star Data

Tweet Comment 12 comments | 0 recs |

Do you like this story?

Comments

Just goes to show

That especially with prospects, depending on who you ask, the results may vary dramatically.

As a Reds fan I find this interesting as KG himself said the Reds were the 9th best system in his mind, but they were pretty poor by this method. As you noted though, its only a look at the top value and KG is a little lower on the Reds top value than other placed (for example he has Yonder Alonso outside of the Top 100).

www.redsminorleagues.com

by dougdirt on Mar 15, 2010 3:42 PM EDT reply actions

I wonder how he came up with Cincy being ninth

Presumably, he weighs top-level pitching prospects more than the Wang valuation system does. Teams wouldn’t perform as well if their very best tier of prospects primarily consisted of pitchers. KG putting them ninth surprises me.

Minnesota got absolutely monster value for Hicks, Sano, Ramos and Revere, so they didn’t lose much for their lack of pitching depth.

I like baseball.
I write for Beyond the Box Score and The Hardball Times Fantasy

by Satchel Price on Mar 15, 2010 4:02 PM EDT up reply actions

Neat stuff -- was looking forward to this coming out.

Even if we knew the exact ranking of KG’s Top 330 prospects, applying Victor’s values would lack some accuracy. For example, Victor found that top 50 pitchers were all worth about the same. This could be because BA/Sickels is bad at projecting the top pitchers. If KG was better at this, a top-ranked pitcher should carry more weight.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Mar 15, 2010 4:06 PM EDT reply actions

KG seems to have more faith in his pitcher grades than Wang would presume

Because he seems to weigh the value of Chapman and Mike Leake to the Reds substantially more than the Wang system does. I wonder if his opinion of Chapman has changed recently, I wish I had a BPro subscription so I could read his organizational rankings.

I like baseball.
I write for Beyond the Box Score and The Hardball Times Fantasy

by Satchel Price on Mar 15, 2010 4:12 PM EDT up reply actions

Intuitively, the properties of hitter/pitcher values that Wang found make sense. The top hitting position player prospects are almost always older because it takes a lot longer for a hitter to start showing all of his tools during games than a pitcher, who could come out of the gate throwing 97 MPH with good sink and a plus slider at age 17.

Because of that and the relative high collapse rate of pitchers, it makes sense that the standard deviation of top-300 hitter vales to be a good bit higher than that of top-300 pitchers. To me, at least.

Doesn’t mean it’s true by any means, just makes sense to me.

Beyond the Box Score / Capitol Avenue Club / shwitter: @CapitolAvenue

by PWHjort on Mar 15, 2010 4:16 PM EDT up reply actions

Oh, I agree.

I also like that Wang’s findings match the second, underappreciated meaning of TINSTAAP: any pitcher, at any time, could figure it out and make the jump to the majors, while hitters tend to progress more continuously and give some warning they’re about to take the next step.

If you start looking at the actual rate of drop-off, however, I bet there are non-trivial differences. I think we’re just talking about two different levels of zoom, here.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Mar 15, 2010 4:55 PM EDT up reply actions

I did the exact same thing at my desk

Except didnt write up on it, but also came up with different results.

My top 10 were, Tampa Bay, Texas, Oakland, Cleveland, Cubs, Boston, Minnesota, Washington, Mets, San Diego.

And used a 5 for 5, 4 for 4 points system, and in the case of ties used the higher ranked 5 star player as the tie breaker.

by backtocali on Mar 15, 2010 4:31 PM EDT reply actions

That's not factoring

numerous discrepancies in value that Wang discovered between top-level hitting and pitching prospects. The Wang system rewards teams with lots of top-level position prospects.

I like baseball.
I write for Beyond the Box Score and The Hardball Times Fantasy

by Satchel Price on Mar 15, 2010 4:39 PM EDT up reply actions

Very True

I have to have the appearance of “working” at my desk once in a while.

I would have never thought to put Wangs values on those players though.

by backtocali on Mar 15, 2010 5:27 PM EDT up reply actions

Well, it's certainly not anything official.

I just think it’s an interesting look at how Goldstein’s system racks up the organizations.

I like baseball.
I write for Beyond the Box Score and The Hardball Times Fantasy

by Satchel Price on Mar 15, 2010 5:35 PM EDT up reply actions

Dodgers

I was surprised to see the Dodgers so high, with only 2 5 star guys and 1 4 star guy. Thanks for the list.

by Thinkblue on Mar 16, 2010 12:11 AM EDT reply actions

It helps that their best prospect was a position player

and that their depth was still in pitching though.

Top level position players are considered more valuable than top level pitchers by Wang’s method, but second-tier pitcher prospects are more valuable than second-tier position player prospects. It’s like New York’s ranking, which is held up pretty much entirely by Montero.

I like baseball.
I write for Beyond the Box Score and The Hardball Times Fantasy

by Satchel Price on Mar 16, 2010 10:30 AM EDT up reply actions