View Single Post
Old 03-03-2013, 09:36 AM   #52
VanillaGorilla
All Star Starter
 
VanillaGorilla's Avatar
 
Join Date: Dec 2011
Posts: 1,371
Blog Entries: 2
Chart and Thoughts

With 4 of the 6 inductees from the latest class entering on the Veteran Standard, we see how the floor is intact just above the 2 line.

The mean GC is still 3.9. The median score is a 3.6 between Albert Belle and Smokey Joe Wood.

The standard deviation for the GC is now 1.5.

Baseball talent is not distributed in a Bell Curve format (as is often misrepresented as being the case in various forum posts). It is more like an inverse Bell Curve with a long tail on the right hand side (more players with talent greater than the norm and to a greater degree than players with talent below the norm, and those players distance from the norm being to a lesser degree).

The norm for GC of these HOF inductees (Using 1 standard deviation from the mean as the definition of "normal") is between 2.1 and 5.1. There are 2 players that are outside of this norm on the low end (RPs O'Neil and Serrano) and 12 above the norm.

The lowest plot is 0.3 points outside of the 1 StDev window. The highest plot is 5.7 points outside the 1 StDev window. This is a greater distance outside the "normal window" by a factor of 19, and this is expected to be the case because HOF scores, like the baseball talent they represent, are not distributed in a Bell Curve distribution. And the talent of players generated by OOTP when using either a historic or fictional model are correctly not Bell shaped.

The plots from left to right represent each inductee in the order he was inducted. Those that were inducted in the same year are plotted in the same order they get listed in the thread. On this chart, the plot furthest to the right represents Curt Schilling.



ADD: I was working multiple thoughts in my head, and what I typed was in error. Not to get into the weeds, too much, but when a distribution's median is not its mean, it is not a normal distribution. This is by definition. There is a calculation for median standard deviation which is different from mean standard deviation. I did not use this to illustrate my point. Instead, I shifted the range of 1 standard deviation from the mean, plus and minus (so actually 2 StDevs) to be centered upon the median in order to illustrate the frequency and degree to which the plots are heavy to the left of the mean and long tailed to the right of the mean. If I had simply typed the word "median" instead of "mean" at my edit point, to begin with, this explanation of correction would be unneeded and I would not have failed in my attempt to be both accurate and concise in my numerical musing.

I knew what I meant, and I meant what I meant, but what I meant is not what I originally said.

And you can save your "That sounds like my wife!" comments. Thank you.

Now to find my coffee wench....
Attached Images
Image 

Last edited by VanillaGorilla; 03-04-2013 at 08:11 AM. Reason: ADD, and edit*
VanillaGorilla is offline   Reply With Quote