Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Baseball Newsstand > Baseball Primer Newsblog > Discussion
Baseball Primer Newsblog
— The Best News Links from the Baseball Newsstand

Thursday, March 20, 2014

Baseball Prospectus | Overthinking It: The Big Questions from the 2014 SABR Analytics Conference

A lot of good stuff in the article.

Quite a bit, apparently. Dave Allen and Kevin Tenenbaum compared three projections for minor-league players: one based on a “naïve” statistical model, one based on the more complex ZiPS projection system, and one informed by Kevin Goldstein’s five-star prospect tiers at Baseball Prospectus from 2007-2012. The scouting-centric model won head-to-head showdowns with both of the others.
[data snipped]
That shouldn’t come as a shock: Scouts, after all, have access to stats, but the stats don’t draw on the scouting grades. Not surprisingly, the method that makes use of more information wins. Still, it’s nice to see that outcome confirmed. Teams, of course, have much more granular scouting grades than KG’s old star system at their disposal, so they can probably build much more accurate projections.

Jim Furtado Posted: March 20, 2014 at 08:08 AM | 9 comment(s) Login to Bookmark
  Tags: sabermetrics

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. villageidiom Posted: March 20, 2014 at 09:28 AM (#4674374)
•On the panel on front-office decision-making, Mariners GM Jack Zduriencik and Bill Geivett and Bobby Evans of the Rockies and Giants, respectively, agreed that moves can come together much more quickly now that communication and access to information is almost instantaneous. Geivett mentioned being able to do a deal that was proposed two minutes before the trade deadline because both teams involved were so well informed about the players involved.


OK, the race begins now. Find that transaction.
   2. Lassus Posted: March 20, 2014 at 10:06 AM (#4674390)
"Based" on?
   3. Best Regards, President of Comfort, Esq. Posted: March 20, 2014 at 10:26 AM (#4674398)
OK, the race begins now. Find that transaction.
It was probably a trade of minor leaguers.
   4. Walt Davis Posted: March 20, 2014 at 05:11 PM (#4674642)
The table on the scouting results don't make sense. The first table supposedly shows scouting vs. the naive stat model. To say "scouting wins" means you don't understand statistics. An RMSE diff of .001 is not meaningful, probably even less meaningful than the meaningless .007 edge in R-square for the naive stat model.

But is that table correct? Cuz the next table compares scout to ZiPS and ZiPS sucks, performing far worse than either of the other two. It's hard for me to imagine ZiPS being that far off. Did the analysts remember to re-adjust the playing time assumptions of ZiPS?

And finally ... y'know, I've been saying this for what must be at least 12 years now ... COMBINE THE TWO YOU MORONS! Sorry, it's been a long time. There is absolutely no reason not to include scout rankings in a statistical model. This will help you see whether one adds useful info to the other.
   5. bjhanke Posted: March 21, 2014 at 03:57 AM (#4674728)
Walt - Combining the two worked real well for me the one time I tried it. In 1990, I did a book called something like "The Top 100 Prospects in Baseball" with a sportswriter named Rob Raines. Rob collated a bunch of scouting reports and produced a list of the 100. I did sabermetric projections, basically by Bill James' old MLE method, plugging the MLEs into his old BROCK2 system. It was really amazing how close those two lists were in terms of who ranked where. And it turned out to be a VERY good predictor of who would and would not be able to play in MLB. Better than that if you discount the injury cases (Geronimo Pena, Mickey Pina). - Brock Hanke
   6. bookbook Posted: March 21, 2014 at 09:13 AM (#4674757)
Interesting discussion on FO pay in the comments section which serves to point out how overpaid fortune 500 CEOs are rather than to suggest GMs are massively underpaid. How do you know these 30 are the 30 best possible GMs? Indeed. And some Fortune 500 CEOs aren't among the top 5,000...

The pure meritocracy model breaks up against the reality of timing, chance, and happenstance. ( which is why "survival of the fittest" models only really apply to large populations over multiple generations),
   7. AROM, Instagram Gangsta Posted: March 21, 2014 at 09:30 AM (#4674769)
You won't get any argument from me on Fortune 500 CEOs being overpaid, but that is not a valid comparison to the GM of an MLB team.

Typical team probably has something like 200-400 million in revenue. Let's compare this to Exxon (420 billion), Apple (171 billion) or Walmart (470 billion).

If GMs are paid in the 1-5 million range, is out of the ordinary for similar revenue companies?
   8. Mike Emeigh Posted: March 21, 2014 at 12:25 PM (#4674870)
If GMs are paid in the 1-5 million range, is out of the ordinary for similar revenue companies?


It doesn't appear to be. The going rate, from a quick check of public companies on EDGAR, seems to be that CEOs of companies with similar revenue to MLB teams are paid in the 500-750K range and that COOs are paid about half that. Many of those executives also get stock options as part of their compensation, though - and sometimes those can be considerable.

-- MWE
   9. valuearbitrageur Posted: March 21, 2014 at 01:50 PM (#4674917)
Yea, GMs can't be paid their marginal contribution level, because sometimes they do marginal damage. Their value is difficult to measure (for most owners), so paying some fraction of their presumed contribution seems reasonable.

Also because if you have an employee who you think adds $300k a year in net profits to your company, clearly you don't pay them $300k, wjphats the point in that, you aren't running a business and taking risk to pass all profits to employees. You pay them as little as the market allows so you can capture as much of those profits as possible. If their value is obvious, easily transferable and non-volatile, they end up getting paid well over $200k, if not probably quite a bit less.

In the case of GMs, the market is small, the results are volatile, and the skills aren't easily transferable. And a GM isn't CEO of a $300m business, team president is. The GM is usually a VP level manager responsible for a $100m cost center. And baseball GMs make far more than their typical counterparts in non sports businesses. Guys running factories with $100m in labor costs annually don't get $3m paychecks.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
A triple short of the cycle
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogOT: Monthly NBA Thread - October 2014
(672 - 6:57pm, Oct 31)
Last: Squash

NewsblogJoe Maddon is to become Cubs manager, sources say
(140 - 6:55pm, Oct 31)
Last: Andere Richtingen

NewsblogMLB.com - In pursuit of Maddon, Cubs dismiss Renteria
(16 - 6:53pm, Oct 31)
Last: Zach

NewsblogAP: Sandoval, Shields Among 121 Free Agents
(21 - 6:52pm, Oct 31)
Last: cardsfanboy

NewsblogSending Gordon | Joe Blogs
(31 - 6:37pm, Oct 31)
Last: Don Malcolm

NewsblogRelix: Watch the Exact Moment the Phish Crowd Found Out the Giants Won the World Series
(6 - 6:37pm, Oct 31)
Last: Bhaakon

NewsblogOT: Politics, October 2014: Sunshine, Baseball, and Etch A Sketch: How Politicians Use Analogies
(4904 - 6:35pm, Oct 31)
Last: Merton Muffley

NewsblogFull Count » Red Sox sign Koji Uehara to 2-year contract
(37 - 6:34pm, Oct 31)
Last: Commissioner Bud Black Beltre Hillman

NewsblogNo, Alex Gordon wouldn't have scored an inside the park home run
(158 - 6:31pm, Oct 31)
Last: Lance Reddick! Lance him!

NewsblogBoston.com: Youk Retires
(14 - 5:49pm, Oct 31)
Last: Bourbon Samurai

NewsblogStatcast: Gordon stops 90 feet from tying Game 7
(18 - 5:46pm, Oct 31)
Last: Barry`s_Lazy_Boy

NewsblogOT:  October 2014 - College Football thread
(545 - 5:43pm, Oct 31)
Last: Infinite Joost (Voxter)

NewsblogAFL Saguaros at Rafters
(4 - 5:15pm, Oct 31)
Last: Mike Emeigh

NewsblogFree Agency Contract Calculator - Beyond the Box Score
(1 - 4:38pm, Oct 31)
Last: fra paolo

NewsblogDeadline: World Series Ratings: Game 7 Scores Home Run For Fox
(33 - 4:11pm, Oct 31)
Last: BDC

Page rendered in 0.2551 seconds
52 querie(s) executed