Posted by dacj501 on 4/26/2014 9:59:00 AM (view original):
Posted by hughesjr on 4/26/2014 9:22:00 AM (view original):
The roles thing is a way to prioritize certain settings for a given position.
I do this inside a spreadsheet for all my teams.
In order to use it for recruiting, it needs to take into account potential ... to use it for actual players, it does not necessarily need to take into account potential.
For example, I have these factors that I currently use for ratings:
Position |
A |
SPD |
REB |
DE |
BLK |
LP |
PE |
BH |
P |
WE |
ST |
DU |
FT |
Off_IQ |
Def_IQ |
TOT-R |
TOT |
PG |
0.08 |
0.16 |
0.00 |
0.15 |
0.00 |
0.03 |
0.11 |
0.17 |
0.23 |
0.03 |
0.03 |
0.01 |
0.05 |
0.10 |
0.10 |
1.00 |
1.25 |
SG |
0.08 |
0.16 |
0.00 |
0.15 |
0.00 |
0.05 |
0.22 |
0.12 |
0.15 |
0.03 |
0.03 |
0.01 |
0.05 |
0.10 |
0.10 |
1.00 |
1.25 |
SF |
0.17 |
0.09 |
0.07 |
0.15 |
0.05 |
0.10 |
0.10 |
0.10 |
0.10 |
0.03 |
0.03 |
0.01 |
0.05 |
0.10 |
0.10 |
1.00 |
1.25 |
PF |
0.19 |
0.06 |
0.12 |
0.15 |
0.10 |
0.14 |
0.07 |
0.04 |
0.06 |
0.03 |
0.03 |
0.01 |
0.05 |
0.10 |
0.10 |
1.00 |
1.25 |
C |
0.21 |
0.03 |
0.17 |
0.15 |
0.12 |
0.17 |
0.02 |
0.02 |
0.04 |
0.03 |
0.03 |
0.01 |
0.05 |
0.10 |
0.10 |
1.00 |
1.25 |
I am not saying that my ratings are the best or anything like that .. I am saying I use role ratings to help me decide who is best to play, etc
riddle me this - what are your ratings for the value of each category based upon? How much suboptimalization might there be in the final PRR rating formula if each core category were under or over valued by as little as 5%? I'm not a stats geek, so I don't know, but my WAG is it is not insignificant.
by this I mean that like any other computer program, GIGO. If your values are not very close to the actual weights used by the engine your displayed PRR values would have virtually no bearing to performance or winning (well, no positive bearing) - or am I way off base on this?
If I sounded like a dick pretend I didn't - not my intent.
My factors are basically based on trial and error ... I am certainly not claiming that they are optimized. In fact, discussion could change my mind :)
The basic concept is that the factors from
A to
DU add up to 1.00 (or 100% as it were). The other factors (FT, Off_IQ, Def_IQ) add up to .25 or an additional 25%. I based the numeric score for those letter grades on this chart:
Letter Grade |
Numerical Equivalent |
A+ |
97 |
A |
95 |
A- |
92 |
B+ |
87 |
B |
85 |
B- |
82 |
C+ |
77 |
C |
75 |
C- |
72 |
D+ |
67 |
D |
65 |
D- |
62 |
F |
57 |
I use similar equations for recruiting via yatzr's recruiting tool ... which also takes into account potential. My Formulas for that are:
PG:
min((A+A_P),100)*0.08+min((SPD+SPD_P),100)*0.16+min((REB+REB_P),100)*0+min((DE+DE_P),100)*0.15+min((BLK+BLK_P),100)*0+min((LP+LP_P),100)*0.03+min((PE+PE_P),100)*0.11+min((BH+BH_P),100)*0.17+min((P+P_P),100)*0.23+WE*0.03+min((ST+ST_P),100)*0.03+min((DU+DU_P),100)*0.01
SG:
min((A+A_P),100)*0.08+min((SPD+SPD_P),100)*0.16+min((REB+REB_P),100)*0+min((DE+DE_P),100)*0.15+min((BLK+BLK_P),100)*0+min((LP+LP_P),100)*0.05+min((PE+PE_P),100)*0.22+min((BH+BH_P),100)*0.12+min((P+P_P),100)*0.15+WE*0.03+min((ST+ST_P),100)*0.03+min((DU+DU_P),100)*0.01
SF:
min((A+A_P),100)*0.17+min((SPD+SPD_P),100)*0.09+min((REB+REB_P),100)*0.07+min((DE+DE_P),100)*0.15+min((BLK+BLK_P),100)*0.05+min((LP+LP_P),100)*0.10+min((PE+PE_P),100)*0.10+min((BH+BH_P),100)*0.10+min((P+P_P),100)*0.10+WE*0.03+min((ST+ST_P),100)*0.03+min((DU+DU_P),100)*0.01
PF:
min((A+A_P),100)*0.19+min((SPD+SPD_P),100)*0.06+min((REB+REB_P),100)*0.12+min((DE+DE_P),100)*0.15+min((BLK+BLK_P),100)*0.10+min((LP+LP_P),100)*0.14+min((PE+PE_P),100)*0.07+min((BH+BH_P),100)*0.04+min((P+P_P),100)*0.06+WE*0.03+min((ST+ST_P),100)*0.03+min((DU+DU_P),100)*0.01
C:
min((A+A_P),100)*0.21+min((SPD+SPD_P),100)*0.03+min((REB+REB_P),100)*0.17+min((DE+DE_P),100)*0.15+min((BLK+BLK_P),100)*0.12+min((LP+LP_P),100)*0.17+min((PE+PE_P),100)*0.02+min((BH+BH_P),100)*0.02+min((P+P_P),100)*0.04+WE*0.03+min((ST+ST_P),100)*0.03+min((DU+DU_P),100)*0.01
(Note: I use 0, 10, 25 for low, average, high for potential for recruiting in the yatzr tool)
Again, it is not necessarily about my numbers ... its about a system that allows you to compare one player to another, based on what one feels is important to each position.
Recruiting needs to take into account potential and in my case it does not take into account FT or IQ's. One could actually look at FT percentage and factor that in .. but since its not in the yatzr recruiting tool, I don't do it.
Anyone wanting to understand the yatzr tool, here is the
link to the thread.
4/26/2014 11:51 AM (edited)