Author Topic: ST stats a predictor of future success?  (Read 293 times)

0 Members and 1 Guest are viewing this topic.

Offline Lintyfresh85

  • Posts: 33933
  • Next year, maybe?
ST stats a predictor of future success?
« Topic Start: May 25, 2010, 12:05:34 PM »
Was surfing the net this morning when I cam upon an article from John Dewan about how players that post monster ST stats have a 60% chance of having a better than career average year.... His qualification for the stat was that the player has to of had at least 200 at bats in his career, and 40 in the current ST. He also has to post a SLG% .200 higher than his career average.

Here's the list for 2010. A lot of hits for such an obscure observation.

http://actasports.com/sow.php?id=243

Possible Breakout Players
Slugging Percentage 200+ points better in Spring Training

Jose Bautista, Blue Jays - Currently co-leads MLB with 14 HR's and is slugging a whopping .566
Mitch Maier, Royals  - Not doing much, but his OPS is the highest of his career at .691
Ryan Zimmerman, Nationals  - Zimm is sporting a .951 OPS and the highest SLG% of his career at .579
Colby Rasmus, Cardinals  - His .915 OPS is .201 points higher than last year. SLG% of .531
Jerry Hairston, Padres - Down year. Miss.
Delwyn Young, Pirates  - Hasn't played enough to make an impact.
Conor Jackson, Diamondbacks - Injured most of the year. (What else is new)
Troy Tulowitzki, Rockies  - Currently boasting his second best OPS of his career at .845
Hunter Pence, Astros  - Started off slow, but is coming along. I'll count it as a miss.
Gregg Zaun, Brewers - Miss. Although, his OPS is higher than his career average.
Aaron Rowand, Giants - Hurt most of the year. Miss.
Nelson Cruz, Rangers - Big time hit. SLG .735... an OPS some guys would love to have. OPS of 1.154
Justin Upton, Diamondbacks - Terrible start to the year... but coming on. Miss.
Will Venable, Padres - Miss.
Alberto Callaspo, Royals - Highest SLG and OPS of his career so far. Hit.
John Bowker, Giants  - Miss. Bowker got his chances but didn't make anything of them.
Mike Aviles, Royals  - Career highs in SLG (.486) and OPS (.842) so far.
Mark Kotsay, White Sox - Miss.


Pretty interesting list... and I'm surprised that it had as many "hits" as it did. Great calls on Bautista, Cruz, Rasmus and Zimm.

Offline PANatsFan

  • Posts: 37068
  • dogs in uncensored, nudes in gameday
Re: ST stats a predictor of future success?
« Reply #1: May 25, 2010, 12:11:11 PM »
4 of 17. Eh.

EDIT:

6. Your bolding was inconsistent :lol: Scouting probably would do better.

Offline Lintyfresh85

  • Posts: 33933
  • Next year, maybe?
Re: ST stats a predictor of future success?
« Reply #2: May 25, 2010, 12:15:20 PM »
4 of 17. Eh.

EDIT:

6. Your bolding was inconsistent :lol: Scouting probably would do better.

Fixed!

I do find it interesting that he hit on Bautista though as his number 1 breakout player.

Offline PANatsFan

  • Posts: 37068
  • dogs in uncensored, nudes in gameday
Re: ST stats a predictor of future success?
« Reply #3: May 25, 2010, 12:16:51 PM »
Last time this sort of thing happened, Brian Roberts lead the league in home runs. Just saying. But if it's true, then the Pirates let yet another one go.

Offline soxfan59

  • Posts: 1201
  • Gough, Gough White Sox!!!
    • John R. Russell, Ltd.
Re: ST stats a predictor of future success?
« Reply #4: May 26, 2010, 12:10:58 PM »
I think this is another example of statistical analysis that really is no analysis at all.  Think about it.  A player who starts the spring hot has a 60% chance of having a better than average year.  Wow.  This is like saying, the odds of a player being better than average when they start well is a little better than 50%.  Average is 50%.  So if you're having a better than average spring, odds are you'll have a better than average year.  And the odds for that are about average.  Somewhere in this is the same argument about the odds of the next child born will be a boy or a girl.  I'm really not impressed by this.

Its possible that this is so because I am a fan of the White Sox.  Note the White Sox player on the list is a bust.  This is common for White Sox players, especially prospects or new players.  White Sox history is littered with the careers of players who had dynamite spring training stats who did not amount to anything.  Daryl Boston, John Cangelosi, Brian Anderson etc etc.

Offline PANatsFan

  • Posts: 37068
  • dogs in uncensored, nudes in gameday
Re: ST stats a predictor of future success?
« Reply #5: May 26, 2010, 01:01:08 PM »
I think this is another example of statistical analysis that really is no analysis at all.  Think about it.  A player who starts the spring hot has a 60% chance of having a better than average year.  Wow.  This is like saying, the odds of a player being better than average when they start well is a little better than 50%.  Average is 50%.  So if you're having a better than average spring, odds are you'll have a better than average year.  And the odds for that are about average.  Somewhere in this is the same argument about the odds of the next child born will be a boy or a girl.  I'm really not impressed by this.

Its possible that this is so because I am a fan of the White Sox.  Note the White Sox player on the list is a bust.  This is common for White Sox players, especially prospects or new players.  White Sox history is littered with the careers of players who had dynamite spring training stats who did not amount to anything.  Daryl Boston, John Cangelosi, Brian Anderson etc etc.

If you can quantify "breakout year", then you can use regression to see what fraction of the variation in the data is described by "spring training performance", which he does quantify. We all learned about the correlation coefficient in school, you just square it.

But spring training performance could be masking a true variable that explains the increase in performance, but correlates with spring training performance. Like better off-season conditioning.

Also, how he obtained the 60% is not made clear, and there's no error on that prediction. I hate this kind of thing. Maybe that's why it's "Acta" sports.

Offline JCA-CrystalCity

  • Global Moderator
  • ****
  • Posts: 18527
  • Platoon - not just a movie, a baseball obsession
Re: ST stats a predictor of future success?
« Reply #6: May 26, 2010, 01:45:56 PM »
I wonder if the study has any sort of correction for aging.  I would think younger players would be the ones more prone to year-to-year improvement and would have the better seasons following springs with big jumps. Zimmerman, Cruz, Rasmus, and Aviles would fit that mold.