Some of you may remember that I started a thread three weeks ago about how I had worst three point shooting guard in D3.

Based on some of the advice in that thread along with having to modify it due to having a truly terrible team I made some unconventional changes.

Prior to the thread I had my starters with a distribution in the range of 5-9. The guys on the bench were between 3-6. My starting guards were at the 9's and the next highest was at 8. They had the highest distribution but not by a whole lot.

Huebner is my starting point guard and Martino is my starting shooting guard.

Here was the distribution with the settings I described above after the first 14 games of the season.

R. Webb 19.64%
J. Martino 13.91%
J. Huebner 13.78%
R. Adkins 11.70%
M. Unvarsky 10.21%
G. Mercer 7.35%
E. Beals 7.22%
H. Hern 6.50%
D. Beasley 4.36%
M. Melton 3.97%
W. Rayes 1.37%


Webb had the highest distribution despite having a slightly lower number than my guards. That didn't surprise me since post players end up getting a higher percentage than you'd expect due to rebounds and interior fouls.

But then I made some changes that I expected to throw things really out of whack. I changed Huebner from a 9 to a 22. And I changed Martino from a 9 to a 28. Other than Huebner, Martino's distribution was at least 3.5 times greater than any other player and he was 4-5 higher than most others.

I played with these settings for the last 18 games. Here are the distribution results for those:

R. Webb 14.77%
J. Martino 15.64%
J. Huebner 17.63%
R. Adkins 10.61%
M. Unvarsky 9.28%
G. Mercer 7.48%
E. Beals 7.75%
H. Hern 5.31%
D. Beasley 3.28%
M. Melton 6.60%
W. Rayes 1.66%


And the difference between the first 14 and last 18:

R. Webb -4.87%
J. Martino 1.73%
J. Huebner 3.85%
R. Adkins -1.09%
M. Unvarsky -0.93%
G. Mercer 0.13%
E. Beals 0.53%
H. Hern -1.19%
D. Beasley -1.08%
M. Melton 2.63%
W. Rayes 0.29%


If I hadn't told you I pretty much tripled the distribution for my two guards after game 14, do you think you could have figured that out from the percentages?

I have zero experience doing something like this. I know from reading forum threads that when you have a very high percentage of distribution you aren't going to get the numbers you want .... but I was floored to see what a small impact my changes actually made.

I know that there are two problems: (1) these guys were often on the court together and (2) 18 games is a small sample size. But I still am struggle to figure out if what happened was a fluke or not.

Thoughts? I really don't care either way since I hope to never have to resort to such a dumb strategy. But I did find the results interesting enough to post.

distribution is difficult to master, much like playing time was before fatigue was instituted, don't worry, the computer will take care of it once the new engine is released

first off you have to isolate the numbers. Just because Martino ended up with 15.64% disto, thats not accurate for his final 18 games. It includes the first 14 in the calculation.

Taking out the first 14 games here's the percentages for your two guards over the final 18 games of the season:
Martino - 16.98%
Huebner - 20.62%
(not **exactly** correct, but very close)

looking at these numbers, i see nothing wrong with Huebner. About where I would guess his percentage would end up with a 22 distro. Possibly even a tad high. Martino is on the low end, but still within reason. Actually, not within reason (IMO), but similar to past results i've observed.

another hidden factor in distro is min per game. Martino averaged 22.2 mpg. In comparison, Webb played 26.8 (over 20% higher than Martino). So in Martino's case, less floor time, less shot opportunities.

Mrpolo09 -- I took the steps to isolate the final 18 games. You don't see the season numbers above. There are split between the first 14 and the last 18 and there isn't any overlap between the two.

Even if I goofed on the numbers (and I'm 99% sure they are right), Huebner's 22 distribution was 2.5 times that of anybody else besides Martino.

As to minutes per game, you are right. I actually figured those numbers out too but didn't post them because they aren't nearly as elegant as the simply distribution percentages. Or at least I can't figure out how to post them elegantly.

To take Webb into account who played more minutes and had a distribution of 8, here are numbers on a per minute basis. ....

The first 14 games of the season when my guards were at 9 distribution compared to Webb's 8, on a per minute basis Webb had 13.4% more distribution than both players -- yes the number for both rounds to 13.4. This makes sense in that post players get a bump from how you'd expect them to be.

The last 18 games of the season, on a per minute basis, Webb had 20.5% less distribution than Martino and 27.7% less Huebner. Given that the distribution was 8 for Webb, 22 for Huebner, and 28 for Martino that really doesn't make a whole lot of sense to me. Especially since these guys were often on the court together.


Again, I hope I never have to play like this again but it was shocking to have several games where I'd have multiple players shooting more than Martino despite him having a distribution level 4-5 times higher than the rest of his teammates.

I have never figured out distro. maybe others like OR or lostmyth can comment but I can never seem to get the magic formula for getting my players the right number of shots.

I have decided that going with a motion offense and a zero distribution is my only chance at being competitive.

i try not to overthink distro, i just use a fairly simple method using numbers 0-5 or 0-3, having the highest number as my tier one go-to guys, the next number as tier two, etc.

i think that nets good results, i then tinker with it based on who is shooting a good % and turning it over less and when people mature throughout the season.

Quote: Originally posted by mbalding on 12/06/2009I have never figured out distro. maybe others like OR or lostmyth can comment but I can never seem to get the magic formula for getting my players the right number of shots.

I have decided that going with a motion offense and a zero distribution is my only chance at being competitive.

settings distro is pretty tough. my strategy for it has changed pretty steadily since i started, and it continues to change significantly. it takes a lot of tinkering IMO, but when you do it well, i think it is a non negligible advantage over 0 distro in the large majority of cases. so, even if 0 distro works better for you today, i'd recommend setting it yourself anyway. at least some of the regular season! otherwise, you will always have that disadvantage pulling at you :(

KU - how did you figure out the per minute percentages for Webb?

To really be able to answer this, I'd have to know how Webb's distro compares to the backups for Huebner and Martino. Also, I don't think you can assume that Webb is ALWAYS on the court when Huebner and Martino are. So, in other words, if Webb has a distro of 9 and and all the backups have a distro of 1, I could easily see these numbers coming out.

Simply shots per minutes played.

That said, this isn't a big deal. And if it is a big deal -- and it is not -- Webb actually is the wrong person to focus on. He's the one person who saw a sizable decrease in distribution. I only mentioned him in that he had the 3rd highest on the team and that he theoretically is the baseline to measure how much greater the distribution should have been for the other two. What really threw me for a loop was that was everybody else on my team who I kept with a distribution of 4, 5, 6 had very small changes after I gave the two guards the super high distribution.

If I hadn't told you I made a change, the other 9 players would seem to have statistically insignificant changes. 4 of the 9 actually saw an increase and the 9 as a whole went from 52.7% of distribution to 52.0%. That is ignoring that Webb had a decent sized decline. But if you put him in the mix with the other 9 players, they were as a group still taking almost as many shots after the change as beforehand.