Post by eric on Feb 3, 2016 11:20:33 GMT -6
There was a lot of discussion in shout yesterday about the draft players I'm making, and I don't feel I adequately addressed all the concerns so I'm going to go into it at great length.
The first point I want to stress is that I have no control over what happens in Training Camps. If there are two PGs each with 100 hidden potential and each with the same changes in potential, after three TCs one player can be -15 Handling attribute and one player can be +45. Is that exact pairing a likely outcome, no, but 95% of values will fall in the range of -6 to +36, so we are talking about very broad ranges even when potentials remain identical. Add in different potentials and different potential changes and there is enormous randomness in the system.
Second, my only input is what's in the profiles. Goran Dragic came up yesterday as an example of a player whose turnovers were too high. His profile listed him as a second round PG/SG with C+ Handling, it would make no sense for me to give him a sterling Handling attribute. I don't raise this point because I want to convince people that this is the best or only way to make draft players. I raise it because I think a very large portion of discontent comes from expectations not met, and so if you know more about my process then you'll be more likely to have accurate expectations. I'm not going to secretly make a second rounder have awesome starting attributes. Most of the time I couldn't even if I wanted to because their starting grades are so bad, that's why they're projected to go in the second round.
Third, I only make about 40 guys a year, the rest are generated randomly by the software. In a similar vein, as has been mentioned many times Soup will tweak people as he sees fit. I don't know any more than you do what he does to anyone, and nobody knows how much he is or isn't going to do in the future.
.
Now I would like to respond to certain specific claims with facts.
1. Everybody Is Terrible
Let's see how my draft classes have done so far by WS/48. It turns out that about 27 players have played >500 MP for the 3024, 3025, and 3026 draft classes, so let's see how they look and give everyone I created that didn't play a 0:
In a past article when I ran similar analysis for the 3010-3015 draft classes I found the following ranges for all 58 players:
These ranges prorated for a 35 player sample are 6, 14, 23, and 35 respectively... and they were measured for players in the primes of their careers as opposed to two or three years in. The area my draft creation is weak in so far is the late first and second rounds, and that could just as easily be because those players haven't gotten a chance to play yet. For example, DJ Augustin has a career mark of .055 WS/48, but because he's only played 369 minutes he counts as a 0. A few more years of growth, SC/RCs, etc., and I'm confident these numbers will be even better.
2. All The Point Guards Turn it Over
Here are the average TO/g leaders from 3010-3015 vs. from 3024-3027:
Here are how many point guards I have made that have been in the top 10 from 3024-3027, along with their profiles:
Kyrie Irving (three times) - "He does need to improve on turnovers", B- Handling
Derrick Rose - "pretty polished", C+ Handling
Pearl Washington - "mostly white guy", C+ Handling
George Hill - Late 1st, C+ Handling
For reference, here are some other point guards who have been in the top 10 of turnovers from 3024-3027, and their profiles:
Steve Nash - "probably the best passing PG in the class", B+ Handling
Mike Conley - "elite ball handling", B Handling
Isiah Thomas - "plays the game under control ... The classic floor general", B- Handling
Aaron Harrison (all four years) - SG, C+ Handling
My guys actually turn it over less than past guys, and the ones who do it the most ought to, and none of my guys are done growing yet.
.
I feel like these facts address the concerns I have heard. If anyone has any questions or other concerns, please voice them and I will respond. We aren't adversaries, we're all in this league together.
The first point I want to stress is that I have no control over what happens in Training Camps. If there are two PGs each with 100 hidden potential and each with the same changes in potential, after three TCs one player can be -15 Handling attribute and one player can be +45. Is that exact pairing a likely outcome, no, but 95% of values will fall in the range of -6 to +36, so we are talking about very broad ranges even when potentials remain identical. Add in different potentials and different potential changes and there is enormous randomness in the system.
Second, my only input is what's in the profiles. Goran Dragic came up yesterday as an example of a player whose turnovers were too high. His profile listed him as a second round PG/SG with C+ Handling, it would make no sense for me to give him a sterling Handling attribute. I don't raise this point because I want to convince people that this is the best or only way to make draft players. I raise it because I think a very large portion of discontent comes from expectations not met, and so if you know more about my process then you'll be more likely to have accurate expectations. I'm not going to secretly make a second rounder have awesome starting attributes. Most of the time I couldn't even if I wanted to because their starting grades are so bad, that's why they're projected to go in the second round.
Third, I only make about 40 guys a year, the rest are generated randomly by the software. In a similar vein, as has been mentioned many times Soup will tweak people as he sees fit. I don't know any more than you do what he does to anyone, and nobody knows how much he is or isn't going to do in the future.
.
Now I would like to respond to certain specific claims with facts.
1. Everybody Is Terrible
Let's see how my draft classes have done so far by WS/48. It turns out that about 27 players have played >500 MP for the 3024, 3025, and 3026 draft classes, so let's see how they look and give everyone I created that didn't play a 0:
range ws/48
1-6 0.106
7-14 0.053
15-22 0.002
23-35 -0.067In a past article when I ran similar analysis for the 3010-3015 draft classes I found the following ranges for all 58 players:
range ws/48
1-10 0.094
11-23 0.057
24-38 0.025
39-58 0.006These ranges prorated for a 35 player sample are 6, 14, 23, and 35 respectively... and they were measured for players in the primes of their careers as opposed to two or three years in. The area my draft creation is weak in so far is the late first and second rounds, and that could just as easily be because those players haven't gotten a chance to play yet. For example, DJ Augustin has a career mark of .055 WS/48, but because he's only played 369 minutes he counts as a 0. A few more years of growth, SC/RCs, etc., and I'm confident these numbers will be even better.
2. All The Point Guards Turn it Over
Here are the average TO/g leaders from 3010-3015 vs. from 3024-3027:
10-15 23-27
4.6 4.1
4.2 3.7
3.7 3.6
3.5 3.4
3.4 3.3
3.2 3.2
3.1 3.1
3.1 3.1
3.0 3.0
3.0 2.9Here are how many point guards I have made that have been in the top 10 from 3024-3027, along with their profiles:
Kyrie Irving (three times) - "He does need to improve on turnovers", B- Handling
Derrick Rose - "pretty polished", C+ Handling
Pearl Washington - "mostly white guy", C+ Handling
George Hill - Late 1st, C+ Handling
For reference, here are some other point guards who have been in the top 10 of turnovers from 3024-3027, and their profiles:
Steve Nash - "probably the best passing PG in the class", B+ Handling
Mike Conley - "elite ball handling", B Handling
Isiah Thomas - "plays the game under control ... The classic floor general", B- Handling
Aaron Harrison (all four years) - SG, C+ Handling
My guys actually turn it over less than past guys, and the ones who do it the most ought to, and none of my guys are done growing yet.
.
I feel like these facts address the concerns I have heard. If anyone has any questions or other concerns, please voice them and I will respond. We aren't adversaries, we're all in this league together.
![[Account Deleted] Avatar](http://storage.proboards.com/5589790/avatar/ZhaFVHTxVPwmyuOiOKEd.png)