Re: TO: ALEX WANG....

[Alex Wang <awang@mit.edu>]

Ryan, 

>      I know you're from MIT and you must have an incredible analytical 
>      mind, but your calculations are vague and make no sense.  I have a BA 
>      in Statistics and I currently work for a biostats company doing 
>      analysis and I have no idea where you are pulling your 2-4 points a 
>      game.

I TEACH a course in probability and stats at MIT, so don't try to impress
me with your BA. The fact that you have no idea what I'm talking about
is a reflection on your ignorance, not my "silly assumptions". My worst
students could follow that argument with ease (and I'm sure almost all of 
the other list participants did also).

>      How about the games 
>      where Walker doesn't take 20 shots?  How are those accounted for?  And 
>      what if Walker doesn't shoot 40% for a game which by the way has 
>      happened alot considering his overall FG % is 40%.

Now something that should be familiar to you (but clearly isn't) as 
someone who supposedly works in statistics is the concept of AVERAGES.
I'm embarassed for the statistics department at your alma mater that 
I have to explain this to you.  In some games, Walker takes more than 20 
shots and some games he takes less - it averages to 20. Is this starting 
to ring a bell? Similarly on some games he shoots less than 40% from the
field and some games he shoots more - again it averages to 40%. And remember, 
we were talking about AVERAGE number of points lost per game via bad shooting 
and missed passes. I'm not sure why this isn't obvious to you since you 
claim to be such a statistics expert. Really, your company should be paying
me for training you in such a fundamental statistical concept.

If you look at my last post, I explained very clearly that if Walker was
a player who say, shot like Jordan (45%), he would be hitting one more 
shot per game (2 points) on AVERAGE. Where are my silly assumptions? 
There is no disputing that if Walker hit one shot per game that he currently 
misses, he would be shooting 45%. Two shots per game (4 points) would make 
it 50%. If you know anything about pro basketball, you know that shooting 50% 
for 20 shots a game is pretty damn good. I won't repeat the rest of the 
argument because everyone else on the list probably got it the first time around.

In contrast, you said "at least" 10-15 points which you clearly just admitted 
you pulled out of your butt. Now you say 6-8 points after I showed how ludicrous
10-15 ppg was. I'm glad you're making progress. Do you have ANY data to 
substantiate this new number of 6-8 ppg, other than your emotional response to 
the games? I've justified my numbers with actual game statistics and clear-cut 
reasoning. You've done nothing to justify your claim that I am using "random numbers"
and "silly assumptions". It's interesting that you say my calculations are 
"vague" when your "calculations" are based "just from observations of several 
games". 

I challenge you to do an analysis of how you got 10-15 ppg or even 6-8 ppg. 
And analysis doesn't mean, "I watched a couple of games and man does Antoine
take bad shots" or "Charles Barkley says Antoine needs to shoot better."

Alex

> ------------------------------
> 
> Date: Fri, 27 Feb 1998 11:06:33 -0500
> From: Ryan.Lee@parexel.com (Ryan Lee)
> Subject: TO: ALEX WANG....
> 
>      I know you're from MIT and you must have an incredible analytical 
>      mind, but your calculations are vague and make no sense.  I have a BA 
>      in Statistics and I currently work for a biostats company doing 
>      analysis and I have no idea where you are pulling your 2-4 points a 
>      game.  You just throw random numbers and percentages in and make silly 
>      assumptions that make your claims look true.  How about the games 
>      where Walker doesn't take 20 shots?  How are those accounted for?  And 
>      what if Walker doesn't shoot 40% for a game which by the way has 
>      happened alot considering his overall FG % is 40%.  You can't use 
>      simple percentages and stupid assumptions to predict number of lost 
>      points in this manner.  There are alot more factors involved such as 
>      the number of times Walker touches the ball, number of shots missed by 
>      Walker in relation to open teammates, the FG% of his teammates that 
>      game and the probability they hit their shots assuming they receive a 
>      pass from Antoine, etc. etc. etc. .  My estimation was just from 
>      observation of several games ( you know random sampling right?). 10 - 
>      15 points I admit was an exaggeration, it's more like 6-8 points of 
>      lost offense a game, but 2-4 points a game is still ridiculous.  I 
>      suggest before you try imposing your scientific mind on everyone to 
>      check YOUR facts and your basic Stats classes out before you post 
>      anything.
>