The Hardball Times: Testing my career projection system
2
|
The Hardball Times found this 4/10/2008 on www.hardballtimes.com [flag] |
Tags:
MLB
Ryan Zimmerman
Comments (9)
-
tangotiger +1Good stuff. -
johnbeamer Yeah, David -- nice work.
One question though -- you've proven that your system is fair to young hitters but you could be underestimating older hitters.
A-Rod is 5 seasons from 3,000 hits (at 150 hits per year, which would comfortably exceed his recent form). Even if you factor is decline it seems like a lock. So what you are effectively saying is that there is a 65% chance he'll have a career ending injury in the next 5 years -- seems unlikely.
Perhaps you could repeat for 32 year olds with 2,300 hits and see what odds you get ...
-
Geoff Young The amount of unknowns in this line of inquiry would scare the bejeezus out of me. I'm glad you're not so easily frightened, because I find this stuff utterly fascinating. -
xeifrank Any chance you could include along with the players chances of getting 3000 hits.
1) Their career expected hit total. 2) The standard error of their career expected hit total.
vr, Xeifrank
-
David Gassko You can get all that if you purchase the THT Season Preview, which includes a spreadsheet with projected career totals and odds for every player active in 2007. The standard error can be backwards-computed from the projected totals and odds.
-
-
GuyM David: Interesting analysis. Two related questions:
1) Theoretically, your aggregate projections for the # of 3K players in a given age cohort could be right, but your projections for highest-rated players still too high. That is, Young's projection could be too high, but a bunch of other guys have a 4% chance while you have them at 1%. Do you have a way to test that?
2) Do you have a data-fitting issue here? I don't know your methodology, but did the stats of the 19 post-WWII 3,000 hitters play a significant role in creating your model? If so, then it's less of a surprise that it "predicts" many of them to get 3K hits. Right?
-
David Gassko Guy,
Good comments as always. The first theory I tested at the end of the article, when I looked at the 10 guys with the best odds of reaching 3000 hits at each age. As I concluded, "In total, there was only a very slight over-prediction, and that was probably due just to random chance."
The second comment is spot-on, though I don't think this is a huge issue. Those players make up a very small part of the whole sample, so they shouldn't have too huge an effect on the final equations.
-
GuyM Good points on both counts, David. I guess the projection I wonder about is Delmon Young. Given his Ks, and very high BABIP, can we really forecast him to produce hits at the necessary rate? Did you factor in BABIP?-
David Gassko No. As I wrote in the Season Preview, this is meant to be a simple system, so all I look into is the statistic in question. You can certainly quibble on individual players -- my system could be much improved if it took into account a lot of other information, but it's never going to be accurate enough for me to be motivated to do that.
-
-
-
Links (2)
THT: Gassko: Testing my career projection system
Published 4/10/2008 at BBTF's Baseball Primer Newsblog
THT: Gassko: Testing my career projection system Gassko’s latest toy… Those numbers are still significantly higher than what Bill James’ Favorite Toy projects. So maybe we should run one more test: What if my system strongly over-projects young hitters with high odds of making it to 3,000 hits, like Young and Zimmerman? Well then, let’s run the same test as above but restrict ourselves to only the players at each age who were in the top 10 in projected odds of gathering 3,000 hits. Among 20-year-olds, my system thought that 4.7 of the top 10 players would reach 3,000 hits; ...
Links of the day 4/10/08
Published 4/10/2008 by Ryan at The Victoria Times
When you're done reading Beth's post, go ahead and take a look at some of these. The Hardball Times seems to think Delmon Young has a pretty good shot at 3000 hits. Boy I hope so. ...
