Post by eric on Dec 29, 2020 21:08:29 GMT
👨🏼⚕️delapandemic🚑 wee2dee Majic TimPig skrouse Logan kn88 Muzunga Mike jhb kujo killybing Druce pointyegg andrewluck troybarnes jayz93j ahebrewtoo IanBoyd
I've finished building and testing 6.0 players.
Overall the ws/48 went from .110 / .098 / .100 for bigs/wings/point in 2049 to .107 / .101 / .102 for 6.0 creation players. In the top ten were 2 points 5 wings and 3 bigs, and the top three were one of each.
Overall pts/tsa went from -1% / +2% / -2% for bigs/wings/point relative to league average to -1% / +4% / -3%. I didn't intend for it to work out this way but wings scoring better at the expense of points is a great outcome too, since point guard is the position that really goes berserk in usage in the outside offense.
I think this satisfies league desire to reach even more variety in team building without straying into the 4.0 trap.
.
Now, in addition to grades and profiles I want to reveal some stats. They won't be exact stats, but I think there are is a way people can get some information without giving too much away: take the average and standard deviation of each counting stat, then use those to bin each player, and these bins can be expressed with rounding or grading. For example, the average player got 6.8 ± 2.9 rebounds per 36, so we could say describe every player who got between 5.4 and 8.3 rebounds as 6.8, or we could describe them as "C" rebounding. Then players to get between 8.3 and 11.1 rebounds could be described as 9.7 or "B" rebounding, and so on for ABCDF in p/r/a/s/b/to. I think the grading system will be a lot cleaner but it'll be easy to output either way.
I don't think scouting of any kind really makes sense in a creation draft, although if people have opinions there of course I'll listen.
I will have the grades for creation players up with sim 8.
.
Regarding the profile writing process in general, people mostly did a decent job. These are some general notes for anyone interesting in the job for 6.0, and on that topic we do still need someone to take that job. We should have profiles for the first class after creation up by sim 8 of the first season, which I expect to be no later than February 1st and probably a little earlier.
There are a lot of things that simply do not exist in the software. This includes but is not limited to:
drawing fouls
spot-up shooting vs. shooting off the dribble or any other similar distinction
durability
streakiness or nonstreakiness
creating their own shot
good midrange shooting in a player that is a bad free throw shooter, or vice versa
It's fine to still include these in profiles, but they are chaff that has no impact on profile creation, so a profile that's mostly or entirely them is not useful. When in doubt, review the list of attributes, they are literally the only things I can adjust for player performance.
.
I think this about covers it, I have a few more things to add to the new board but otherwise 6.0 is ready to go. I feel like this process has gone pretty smoothly, good work everyone.
I've finished building and testing 6.0 players.
Overall the ws/48 went from .110 / .098 / .100 for bigs/wings/point in 2049 to .107 / .101 / .102 for 6.0 creation players. In the top ten were 2 points 5 wings and 3 bigs, and the top three were one of each.
Overall pts/tsa went from -1% / +2% / -2% for bigs/wings/point relative to league average to -1% / +4% / -3%. I didn't intend for it to work out this way but wings scoring better at the expense of points is a great outcome too, since point guard is the position that really goes berserk in usage in the outside offense.
I think this satisfies league desire to reach even more variety in team building without straying into the 4.0 trap.
.
Now, in addition to grades and profiles I want to reveal some stats. They won't be exact stats, but I think there are is a way people can get some information without giving too much away: take the average and standard deviation of each counting stat, then use those to bin each player, and these bins can be expressed with rounding or grading. For example, the average player got 6.8 ± 2.9 rebounds per 36, so we could say describe every player who got between 5.4 and 8.3 rebounds as 6.8, or we could describe them as "C" rebounding. Then players to get between 8.3 and 11.1 rebounds could be described as 9.7 or "B" rebounding, and so on for ABCDF in p/r/a/s/b/to. I think the grading system will be a lot cleaner but it'll be easy to output either way.
I don't think scouting of any kind really makes sense in a creation draft, although if people have opinions there of course I'll listen.
I will have the grades for creation players up with sim 8.
.
Regarding the profile writing process in general, people mostly did a decent job. These are some general notes for anyone interesting in the job for 6.0, and on that topic we do still need someone to take that job. We should have profiles for the first class after creation up by sim 8 of the first season, which I expect to be no later than February 1st and probably a little earlier.
There are a lot of things that simply do not exist in the software. This includes but is not limited to:
drawing fouls
spot-up shooting vs. shooting off the dribble or any other similar distinction
durability
streakiness or nonstreakiness
creating their own shot
good midrange shooting in a player that is a bad free throw shooter, or vice versa
It's fine to still include these in profiles, but they are chaff that has no impact on profile creation, so a profile that's mostly or entirely them is not useful. When in doubt, review the list of attributes, they are literally the only things I can adjust for player performance.
.
I think this about covers it, I have a few more things to add to the new board but otherwise 6.0 is ready to go. I feel like this process has gone pretty smoothly, good work everyone.
