I understand that guard types will get more shots in this engine, given equal distribution. It seems to me to be an engine flaw, but it is what it is. This result seems particularly odd

Heres where my distribution was set:
8- Huang - 15 shots in 32 minutes
8- Salinas - 3 shots in 22 minutes
7- Gordon - 6 shots in 31 minutes
5- Woodring - 7 shots in 18 minutes
3- Ortiz - 5 shots in 29 minutes

Also of note, for when boxscore and pbp are unavailable - the opponent played a press set at +1; only one of Huang’s 15 shots came in the final 3 minutes. Colorado definitely has strong frontcourt defense (which is why I didn’t pump Salinas’s distribution up to 10 or 12, as I’ve done throughout the year).

I have some thoughts, but I’m curious to see other people’s take on how the variance got so wide. I’m especially interested to know if anyone has data or insight on increasing the distribution closer to 100, and if that tends to reduce instances of wide variances like this.

It's not that uncommon with my teams. Distribution refers to number of touches, but doesn't necessarily mean he will shoot based off of those numbers. And as you mentioned, guards will generally take more shots than low post players. You had close to even distro with one guard(Huang) and 1 forward and 1 center, so naturally when that happens, the guard will get a lot more shots. And if you're wondering why Woodring had so many shots in so few minutes is because he was on that second line, where more than likely your second line subs didn't see that much distro, so it was probably a 5-1-1-1-1 ratio.

Have your distribution add up to 100 and you wont run into nearly as many of these variances.

Posted by mullycj on 12/29/2018 1:31:00 PM (view original):
Have your distribution add up to 100 and you wont run into nearly as many of these variances.

But even if he upped it to x3 (24,24, 21, 15 and 9) that should still come out to the same percentage? 24/93, 24/93, 21/93, 15/93, 9/93 compared to
8/31, 8/31, 7/31, 5/31, 3/31.

You're probably right, I'm just wondering why there would be such a variance?

It probably has something to do with the press defense and double teams. Low pass/bh players tend to be more susceptible tobdoubles

“You had close to even distro with one guard(Huang) and 1 forward and 1 center, so naturally when that happens, the guard will get a lot more shots.”

Interesting theory. So if it was reversed, and I had two guards with relatively even distribution, and one front court player with that same distribution number, do you think that front court player would have gotten more shots? That doesn’t really make any intuitive sense. If I set distribution relatively even between three players, wouldn’t it make more sense each of those players be expected to get roughly the same amount of shots?

Between those three players, they took 24 shots from the field. Their total distribution number was 23. And the guard took more than 50% more shots than the other two combined, despite having distribution set at roughly 1/3. Does this really not seem out of whack to you?

I dont know, but that's been my experience.

Posted by mullycj on 12/29/2018 1:31:00 PM (view original):
Have your distribution add up to 100 and you wont run into nearly as many of these variances.

That’s getting at the last question, I’m curious if you have any data for that. I’ve used higher distribution closer to 100, and haven’t really noticed any difference, but then again, this is the largest variance I’ve seen (or noticed, anyway).

If you REALLY want to be anal track who else was on the floor when the shots were taken. Funky sub patterns can lead to one 8 shooter on the floor with 0s and the other 8 shooter on the floor with the 7 and 5.

Posted by Trentonjoe on 12/29/2018 2:04:00 PM (view original):
It probably has something to do with the press defense and double teams. Low pass/bh players tend to be more susceptible tobdoubles

This is one of my thoughts.

Posted by mullycj on 12/29/2018 1:31:00 PM (view original):
Have your distribution add up to 100 and you wont run into nearly as many of these variances.

I don't think distro adding to to 100 would solve anything. And i think this is the least used method of distro among coaches. I could be wrong, but I don't know any coaches that do this honestly. And i attempt to talk to every coach that I cross paths with!

However, I do think higher distro numbers DO produce better results for some reason. As far as getting what you want out of shooters. I recently discussed this with a successful coach. I used to work a 1-10 distro scale. Giving 10 to only my best scorer, and decreasing from there. So my starting 5 would be 10-7-4-1-0 or something. And my shot distro was coming out to be too similar for my liking. So I doubled things.... 20-14-8-2-0 and my shot attempts resemble more of what I was aiming for.

it's odd because it's all the same ratio. But it definitely seems to be different with bigger numbers

Posted by shoe3 on 12/29/2018 2:08:00 PM (view original):

Posted by Trentonjoe on 12/29/2018 2:04:00 PM (view original):
It probably has something to do with the press defense and double teams. Low pass/bh players tend to be more susceptible tobdoubles

This is one of my thoughts.

That’s because you are a smart guy and know who to listen to.

Posted by topdogggbm on 12/29/2018 7:20:00 PM (view original):

Posted by mullycj on 12/29/2018 1:31:00 PM (view original):
Have your distribution add up to 100 and you wont run into nearly as many of these variances.

I don't think distro adding to to 100 would solve anything. And i think this is the least used method of distro among coaches. I could be wrong, but I don't know any coaches that do this honestly. And i attempt to talk to every coach that I cross paths with!

However, I do think higher distro numbers DO produce better results for some reason. As far as getting what you want out of shooters. I recently discussed this with a successful coach. I used to work a 1-10 distro scale. Giving 10 to only my best scorer, and decreasing from there. So my starting 5 would be 10-7-4-1-0 or something. And my shot distro was coming out to be too similar for my liking. So I doubled things.... 20-14-8-2-0 and my shot attempts resemble more of what I was aiming for.

it's odd because it's all the same ratio. But it definitely seems to be different with bigger numbers

Mine always adds to 100, so there is one.

Higher district numbers don't affect anything. It's the percentage. Probbaly anecdotal data that's effectively pseudoscience.

Posted by cubcub113 on 12/30/2018 9:16:00 AM (view original):
Higher district numbers don't affect anything. It's the percentage. Probbaly anecdotal data that's effectively pseudoscience.

I always thought this as well. It made sense to me for years. So much that I argued the point with a couple coaches for a long time (coaches that I think you and I both associate with, cubcub). But by trying it, it changed the results I get. Consistently.

Try 5-4-3-2-1 for a week
Try 15-12-9-6-3 for a week.

you'll see differences. The differences in shot attempts between a 15 and a 6 will be a wider margin than the difference in shot attempts between a 5 and a 2.

For mully who uses 100, i definitely don't think that's wrong. I did that for a long time as well. As I felt that was the way it was supposed to be done. There is no "wrong" distro. It blew my mind at the time to think that it wasn't necessary to use 100. But now I feel like using 100 may force the ball into certain players hands at the end of the game. I don't have any proof of this, and i could EASILY be wrong about this part. But at the end of games, if a player or players have not hit their distro mark, I would think since it's a more definite number that the game would force the ball into their hands.

But with leftover room by not adding up to 100, I feel like it may be more forgiving, and it wouldn't force the ball into a player's hands in the same manner. Could my thoughts be completely wrong? Absolutely. But that's how i think about distro. By leaving some flexibility by not using the full 100, i've always felt like the engine would not be as forceful trying to hit its specific numbers. I have absolutely nothing to back that up. That's just how i look at the situation.