BuzzerBeater Forums

So it IS marginal. Also FT% is around 0.75, so you have: 0.223 v 0.195 + (1 - 0.195) * 0.032 * 0.75 = 0.214 when you factor in FTs.

Disregard the FTs and look at the NBA: the drop in shooting efficiency in the NBA between wide open and tightly contested shots is around 12% (but this study is a bit dated and it did NOT focus on 3s). In 2019 the difference in 3FG% for wide open vs tightly contested 3s for Curry was 52% v 36%, for Beal 41% v 21%, for Harden 46% v 32%, Durant 39% v 26%, Lillard 41% v 34%.

Now you can see how that 3% is actually a fraction of what the number is the real world. Not only that, but you have to consider that a 3 pointer in real life is WAY more likely to be wide open than in BB (with the notable exception of Hardenball).

Last edited by Lemonshine at 6/13/2020 3:00:54 PM

BTW, it is said that there was once a time when outside offenses were the primary tactics, and somebody changed something in the "3pt line" or I don't know what ... and LI is the main tactic since then. I would like to know more about that change in the GE, because I wasn't here back then.

yeah, but it lasted like 4 seasons in 2008 or something like that

it's not 0.09 it's 0.02 and it's marginal because you are assuming that you'd go from 0 assist rate to 100% which is, how should I put it, incorrect. I provided you with the information when I said 24% of 3s are assisted on average and about 47%-49% of other shots. So at a bare minimum, the benefit wouldn't impact more than 76% of the shots and that's assuming you go from 24% to 100%. If you take 88 3 pointers, 22 are already assisted on and 66 are not. If assist rate was to get 3s in line with other shots (effectively doubling the assist rate of a playoff B3 team, that is), then you'd have 22 extra 3 pointers assisted. 22*0.02=less than half a point. Even if you disregard FTs, 22*0.09 is less than 2 points.

That 3% difference is so low (and the result is similar for other shots once you factor in the FTs) that if I didn't think flow determines shot selection and shot allocation, but more importantly keeps turnovers down, I would have already tried to build a team with low passing (say around 7 throughout), to see what happens. At the very least it would cost much less.

Is there a category in between "wide open" and "tightly contested"?

For tighly contested I used the stats when the closest defender is 2-4 feet away from the shooter and for wide open I used 6+ feet. So yeah 4-6 is in between and those are considered open shots.

Last edited by Lemonshine at 6/13/2020 7:09:08 PM

Flow almost surely determines (together with other factors) who takes the shots (allocation), but it also probably impacts which shot is being taken. This is all anecdotal, but it seems clear to me that low PA players commit a lot more turnovers and settle on contested shots or non ideal ones. Turnovers are very bad, subpar shots are also very bad in BB, but contested shots not so much in comparison.

For example, in LI the best shot possible is a dunk, then putbacks, layups, inside shots and driving layups in that order.Then after a massive drop in efficiency there are jump shots and 3s. Ideally you'd want to find players who for a reason or another (one being their PA and that of their teammates) have an above average share of dunks, layups and inside shots and avoid like the plague players who tend to be forced into jump shots of any kind and at any position, as those have a 25%-35% FG% drop in efficiency compared to inside shots and layups. A small extra dip for defended shots (2-3%, if at all) pales in comparison to that.

That is easily going to make a clear difference per game, you don't need extreme assumptions. This is all because of the way BB tactics are designed to punish inside shots for outside tactics and outside shots for inside tactics, so skewing the distribution of the shots in your favour is more important than getting open shots. A guarded 'well selected' shot is always much better than an open 'badly selected' shot. People have realised this intuitively for LI, when they found out they could get guards to take more driving layups and fewer jump shots, leading to both higher efficiency and lower salaries. We know this is largely linked to other skills, but I honestly can't rule out things would look very different without enough PA.

I have no proof that PA affects shot type and distance, but I believe it does, it makes sense and if that's the case the effect is likely much bigger than guarded v open.

Last edited by Lemonshine at 6/14/2020 7:53:08 AM

I just hope they will find a good way to make those long shoots better