Post by eric on Oct 4, 2017 14:12:44 GMT -6
Listened to the Lowe Post pod with Jeff Van Gundy, who at one point suggested that Dwyane Wade could hit 40% of his threes this year. He won't, obviously, but what struck me looking at his b-r page was he is assisted on only 39% of his career three pointers, an unusually low number, so I decided to take stats.nba.com for a spin.
We know a made field goal can be assisted or unassisted.
But we don't know whether missed field goal would have been or not, because the official scorer doesn't tell us.
What an @$$hole.
So we can't really tell what a guy would shoot on assisted or unassisted looks because we don't know which attempts are which.
What we can use instead is stats.nba.com's definition of "catch and shoot" vs. "pull up" shooting.
They keep track of field goals made and attempted in each scenario.
Last year, the NBA made 17880 of 48040 catch and shoot threes and 5365 of 16605 pull up threes, or 37.2% against 32.3%, with 77% of made threes being assisted.
Okay, but we might be skewing the numbers for how valuable being assisted is by including pure catch and shoot guys: your Serge Ibakas (99%), your Marc Gasols (97%), your Klay Thompsons (87%). What if we only look at guys who took at least 100 of each kind? Well, it turns out there were 44 such players last year who combined for 3926 of 9705 catch and shoot and 2778 of 8043 pull up, or 40.5% against 34.5%, with 59% of made threes being assisted. Both %s are higher than league average, which you would hope for for guys taking this many, and the gap is also larger at about 6% instead of 5%, and since the rate of made threes assisted is so much lower the overall rate for the better shooters is only 37.9% against 36.0% league average.
.
Okay, so what does this mean for Dwyane Wade?
Well, we know his career is 431 of 1052 for 28.7%, and he's assisted on 39.4% of his made threes. Let's call these 29% and 40% for ease of following the arithmetic, and let's give him the 6% bump for catch and shoot. We define...
...that is, three point percentage is makes divided by attempts, makes is assisted makes plus unassisted makes, attempts ditto, assisted percentage is six points higher than unassisted, and since 40% of makes are assisted and 60% are unassisted that means assisted makes are 40/60 = 2/3 of unassisted makes.
And now we just plug and chug:
Which is a quadratic formula that gives us unassisted three point percentages of 26.9% or -4%, and since he's not Andrew Wiggins we know he's not shooting -4%
which gives him an assisted three point percentage of 32.9%, so even if he's assisted on 100% of his attempts we wouldn't expect his true talent level to crack 33%, although of course if he only took five all year he could easily make three.
Sorry Jeff.
We know a made field goal can be assisted or unassisted.
But we don't know whether missed field goal would have been or not, because the official scorer doesn't tell us.
What an @$$hole.
So we can't really tell what a guy would shoot on assisted or unassisted looks because we don't know which attempts are which.
What we can use instead is stats.nba.com's definition of "catch and shoot" vs. "pull up" shooting.
They keep track of field goals made and attempted in each scenario.
Last year, the NBA made 17880 of 48040 catch and shoot threes and 5365 of 16605 pull up threes, or 37.2% against 32.3%, with 77% of made threes being assisted.
Okay, but we might be skewing the numbers for how valuable being assisted is by including pure catch and shoot guys: your Serge Ibakas (99%), your Marc Gasols (97%), your Klay Thompsons (87%). What if we only look at guys who took at least 100 of each kind? Well, it turns out there were 44 such players last year who combined for 3926 of 9705 catch and shoot and 2778 of 8043 pull up, or 40.5% against 34.5%, with 59% of made threes being assisted. Both %s are higher than league average, which you would hope for for guys taking this many, and the gap is also larger at about 6% instead of 5%, and since the rate of made threes assisted is so much lower the overall rate for the better shooters is only 37.9% against 36.0% league average.
.
Okay, so what does this mean for Dwyane Wade?
Well, we know his career is 431 of 1052 for 28.7%, and he's assisted on 39.4% of his made threes. Let's call these 29% and 40% for ease of following the arithmetic, and let's give him the 6% bump for catch and shoot. We define...
3p / 3pa = .29
3p = a3p + u3p
3pa = a3pa + u3pa
a3p% = u3p% + .06
a3p = (2/3) * u3p...that is, three point percentage is makes divided by attempts, makes is assisted makes plus unassisted makes, attempts ditto, assisted percentage is six points higher than unassisted, and since 40% of makes are assisted and 60% are unassisted that means assisted makes are 40/60 = 2/3 of unassisted makes.
And now we just plug and chug:
.29 = a3p / (a3pa + u3pa) + u3p / (a3pa + u3pa)
.29 * (a3pa + u3pa) = a3p + u3p
.29 * (a3pa + u3pa) = (5/3) * u3p
a3pa / u3p + u3pa / u3p = (5/3) / .29
a3pa / u3p + 1 / u3p% = (5/3) / .29
a3p / u3p * 1 / (u3p / u3pa + .06) + 1 / u3p% = (5/3) / .29
(2/3) / (u3p% + .06) + 1 / u3p% = (5/3) / .29
(2/3) + 1 + .06 / u3p% = (5/3) / .29 * (u3p% + .06)
(5/3) + .06 / u3p% = (5/3) / .29 * u3p% + (5/3) * .06 / .29
(5/3) / .29 * u3p%^2 + ((5/3) * .06 / .29 - (5/3)) * u3p% - .06 = 0Which is a quadratic formula that gives us unassisted three point percentages of 26.9% or -4%, and since he's not Andrew Wiggins we know he's not shooting -4%
which gives him an assisted three point percentage of 32.9%, so even if he's assisted on 100% of his attempts we wouldn't expect his true talent level to crack 33%, although of course if he only took five all year he could easily make three.
Sorry Jeff.
