Quote:
Originally Posted by olivertheorem
I'd think the easiest way to generate combine results for the drills would be to randomly generate results based on the underlying (hidden) ratings within a normal distribution. For example, say a player's speed rating is 50 on a 1-100 scale (obviously at this point, I'm making up numbers). Now, let's say that a 50 translates to a true 40-time of 4.80, but one standard deviation out makes the typical range 4.75-4.85, and farther out for higher SD. It would have to be made clear what the results-to-ratings translations were, but you would end up with most players that you were fairly certain about their measurables, but you'd still have players have flukishly high or low performances that would help or hinder their draft stock accordingly.
But then, from what I've read so far I'm fairly certain Francis has considered the problem already. :P
|
That would work but now be very realistic. I am pretty sure that the are a lot of factors (Fatigue, Motivation, Pressure to preform well, etc.) that go into how a player runs the 40, not just speed and random chance.