Posted by schuyler101 on 6/24/2010 2:18:00 PM (view original):

Posted by iain on 6/24/2010 1:32:00 PM (view original):

Posted by jimmystick on 6/24/2010 8:32:00 AM (view original):

Posted by schuyler101 on 6/24/2010 12:46:00 AM (view original):

Posted by iain on 6/23/2010 11:05:00 PM (view original):
You're aware that even an entire season isn't a statistically meaningful number of innings, right?

+1

Guardado has pitched 856 career innings and has a 4.23 ERA.  Sullivan has pitched 670 innings and has a 3.23 ERA.  So I'm looking at more than just one full season... even though this most recent season is the biggest variation of them all. 

I entertained trade offers for Guardado this season because I'm sick of dealing with his underperforming.  But in the end, I looked at Sullivan and thought there's no reason why Guardado can't do exactly what he's doing, seeing as how I view them as identical.  

I don't mean to be a dick, but here we go:

In statistics, 1000 is considered a reasonably small number.

1,000,000 is starting to have some statistical meaning.

Basically, an entire career is still a small sample set. This is why you can have an entire season that is an outlier (good OR bad ...but nobody complains about the good seasons... that's not human nature).

+1

MINUS 2

Get your head out of the math books and watch baseball!

Actually, I am a math nerd, too, but statistical analysis in this example is not akin to "math nerd" statistical analysis.  If you were to do forecasting of the stats for both players going forward, the results of that forecasting would be different than what their past results have shown.

Fair enough, randall... although you sound dangerously close to yelling at me to get off your lawn. :)

My point is that even two guys who actually do have the same ratings will NOT have the same career numbers. They just won't. They may not even be close. There is too much noise, even over their entire careers.

I agree with that.

And you can go ahead and play on my lawn if you clean up after my dogs first.

NERDS!

you're being far too liberal with your lawn sharing plan. 

Over the course of 2-3 seasons, those two should be similar enough.  Give it time.

  And that's why I hate discussing Statistics.  Back when I was in SimDynasty we had a raging debate on Clutch.  It is 'statistically unproven' because Sample Size is always 'To small or limited'.  How can you beat that argument?  If an ENTIRE CAREER isn't enough to consider a Sample Size you're just screwed!
  
  Can you imagine this-  "Nolan Ryan was a GREAT Pitcher!"  "You can't say that, his Sample Size isn't large enough to matter."

The point of my question isn't so much "Hey, these guys are exactly the same, they should have exactly the same stats!!!"  Maybe I wasn't precisely clear with my question.  But basically my question is:

"Should I expect more from Guardado, because Sullivan is proving a guy with those ratings can be a Cy Young caliber pitcher, or is Guardado doing exactly what should be expecting of him and it's just that Sullivan is overachieving?"

I really really want to believe it's the former, and I want to believe I've got 2 great pitchers here, but having Guardado for 4 seasons now and seeing him get shelled about once every 3 starts is really waning my patience.  He's coming up to his first year of arbitration.  I want to see him do something noteworthy before giving him a huge contract.

Posted by brrexkl on 6/24/2010 5:40:00 PM (view original):
  And that's why I hate discussing Statistics.  Back when I was in SimDynasty we had a raging debate on Clutch.  It is 'statistically unproven' because Sample Size is always 'To small or limited'.  How can you beat that argument?  If an ENTIRE CAREER isn't enough to consider a Sample Size you're just screwed!
  
  Can you imagine this-  "Nolan Ryan was a GREAT Pitcher!"  "You can't say that, his Sample Size isn't large enough to matter."

Exactly!  I understand that technically speaking 800+ innings isn't a true meaningful sample size, but you have to be realistic.  I'm comparing their entire careers, not just 3 weeks worth of starts.  You gotta throw me a bone here, iain. 

By the way.... why didn't my links work?  I thought one of new features in the forums was that it didn't matter if you were using Firefox, IE, Opera, etc that all the features worked in all browsers???  WTF???

  Just remember this.  1 Season is a Fluke, 2 is a Trend, 3 is a Streak and 4 is a guy eating your money and giving you crap.