Knicks Morning News (2022.12.16)

on yesterday’s topic of teammate pairings and how they do or do not affect individual statistical production, here are some random (not random) on/off pairings i have handy:

1. myles turner scored 17.4/100 on 57.5% TS with sabonis and 24.8 on 57.9% without him.

2. Klay’s career TS with steph (13K mins) is 58.26% (28.17/100). without him it is 54.7% (30.67 / 100).

2b. Wiggins on the dubs is 59.45% TS with steph (25.63/100) and 53.51% without (28.54/100).

3. CJ with lillard was 57.1% (28.6/100) vs 52.45% without (34.2/100).

4. Brunson with luka was 60.65% (20.6/100) vs 57.79% (28.28/100). and the assists obviously skyrocket from 4.7/100 to 9.9/100.

5. RJ has only played 1429 mins without randle in his career, but in those he is 47.4% TS (29.67/100) vs 52.19% (25.2/100) with him.

6. Kawhi and PG dramatically affect each others’ volume but not TS at all.

7. Murray’s TS is 55.8% with jokic and 53.5% without, on identical volume.

7b. Porter’s TS is 65.9% with jokic and 61.05% without on identical volume.

8. Chris Paul has not had a positive delta with the individual TS on pretty much any of his primary scoring teammates. Blake Griffin: none. Ayton: modestly negative delta. Booker, modestly negative delta. even Deandre Jordan. Yet he has a massively positive ORAPM on team performance.

9. AD’s volume with lebron is lower but his TS is identical.

10. Mitchell’s stats were essentially the same with or without gobert in utah.

What do you think about interaction effects exactly?

i think they often matter quite a bit. people sometimes overstate them bc nba player production is on balance surprisingly resilient and people love to tell stories. but merely resilient.

To take one name from that list, Brunson. We have samples with Luka and without. We have now a sample in New York where his usage has jumped 25% and he has completely different teammates.

Across some very varied environments the last three season, he has posted these WS/48 numbers. .144 .144 .166

this doesn’t really cut your way bc his ws/48 in dallas with luka on was more like .10 vs .17 with luka off.

Should we have expected Brunson to play better without Luka? Should we have expected him to play better or worse at a higher usage? Should we have expected him to play better or worse next to Randle and Barrett specifically? Should we have expected them not to impact him at all?

this is just it. most “interaction effects” are difficult to parse. vague, deceptive signals drowning in dynamic confounders. but you don’t have to feign omniscience to buy shorts for summer. i posted here more than once about jalen’s sans luka stats last year in thinking about what he might look on an orange team utterly lacking luka. ditto for fvv sans lowry when he was a free agent. this doesn’t require having takes about whether austin reaves benefits from troy brown’s back cuts.

I can’t say exactly what predictions the interaction effects theory would have supplied about Brunson before he came to NY, but it doesn’t seem all that coherent with, for instance, the impact Randle’s high usage is purported to have on RJ, Grimes, and IQ, or the expectations people have expressed about how Obi or IQ would do with a larger role.

i don’t know the putative randle effects you’re referencing. i kinda doubt i would buy them. there is no grand unification of interactions effects theory i’m aware of. just combinatorial chaos that sometimes, more than sometimes, i’d say, burps loud enough to get your attention.

To me, in the aggregate, these effects seem kind of random and for the most part pretty marginal. I would have said Brunson was going to be more or less the player he was his final two seasons in Dallas, despite the fact his role and teammates have changed dramatically. That’s how it has turned out so far.

same and different are just two dots on the numbers line but i’ll pick the one on the right. he is scoring 31/100 on 10 ast/100 vs 25 and 7 his prior two years. the vastly different bpm and hardly different ws48 are distractions meant to fog the window. we’d rather have this guy, which is who he already was when luka sat, than the averages from the last two years. 31 and 10 per 100 on solid efficiency and stingy turnovers is a scarce and excellent thing. better and scarcer than 25 and 7 on slightly better than solid efficiency and stingy turnovers.

As I have mentioned, and this is just directed at the air, IQ is one puzzle for Knicks fans at the moment who believe usage and efficiency should move in opposite directions. He is taking less shots than he was as a rookie and his ts% is down from then. Given the growth you would naturally expect for a guy entering his third year, the whole thing is strange. And yet, he is still more or less the same guy, basically within 20% of his median.

in a nod to your well known revulsion toward the gnostic tradition i will drop all pretense of analysis and say i am an iq believer. i agree with jk47 that his shot is aesthetically displeasing but think he is totally wrong that this means it won’t go in. it needs to go in.

Someday I am going to release my quantum of talent theory on the world and blow some minds. Just have to sand off some rough edges on this emerging idea 😉

i will happily buy a prelease signed copy of your opus applying lorentz invariance to jr and ish smith but i might discreetly drop a coaster or two on it when entertaining.