in fact if you look at just the 2.5 years with OG on they are +8 whether they have zero of your bad defenders or at least one bad defender. it is only with a bad defender and no OG that they drop to +1.0 (2303 mins).

To me that supports the visual and statistical evidence that OG is a great defender. He probably disappears on offense some of the time because he’s so exhausted covering for Brunson and KAT. 😉

So maybe if we get KAT out of the equation and add a defender it will help OG’s offense too. 😉

KAT for RWIII-Jrue would be really selling low on KAT imho.

It’s a preposterous idea.

It’s been more than 2 years since he has.

lol…that’s actually fair. One can argue that he may be due…

Theoretically, the conclusion you’d be wanting to sis that we should dump KAT because our defense is much better with 4 defenders while our offense is just as effective.

i’m a big fan of getting creative to make bigger sample sizes (how else could tnfh have pulled?). but it’s fraught. for example the 3.5 years you posted only included 28 “good defender only” minutes from 22-23 and 2351 “bad defender minutes,” during which they were only +2.2. but they were actually +3.9 with jalen off that year. this is a compositional bias because instead of asking “are jalen offenses pareto better with four good defenders irrespective of their offense” we could instead be unintentionally asking “were the bad nyk defenders who played with jalen in 22-23 on average worse players than the bad defenders in other years?” this is an especially bad mistake if we’re going to try to use it to draw conclusions about a player who had been here only during the latter sample.

i’m not saying that this kills the hypothesis that it’s a good idea to stick jalen with defenders, which surely is right in most cases. i know the delta mostly holds up even if you eliminate 2022-23 from your run. but then the sample size gets smaller and the data suddenly looks no more compelling than competing hypotheses. for example, this one: zero bad defenders is irrelevant; the key is to play jalen with a great defender who doesn’t kill you on offense like anunoby (+8.0 3215 mins witj jalen vs +2.8 without jalen 3190 mins). in fact if you look at just the 2.5 years with OG on they are +8 whether they have zero of your bad defenders or at least one bad defender. it is only with a bad defender and no OG that they drop to +1.0 (2303 mins).

an obvious counter to applying this to kat is that since kat’s arrival they have been +1.8 with jalen on and kat off (1447 min), +4.2 (2232 mins) with both on and +8.1 with kat on and jalen off (1614mins). and no, i don’t believe this means that kat is better or they ought to trade jalen. more that the lineup data will inherently push you and pull you and beg you to dance for your dinner. i actually think it’s pretty obvious that both jalen/randle and jalen/kat were far from ideal pairings vs a more defensive-oriented player and the lineup data would have to scream in my fact to change my mind much.

I’m convinced Leon is not trading KAT for 2-3 pieces that may be a better fit with Brunson. Giannis or we ridin as is.

I, too, am convinced that Leon won’t do anything smart.

It’s been more than 2 years since he has.

Problem is, their playoff offense stinks relatively worse without him on the floor.

A “Jalen Brunson team” is one that is poor on defense and tries to outscore their opponents, often successfully. Karl-Anthony Towns has been thrown into that mix, making the offense even better — but the defensive issues that trail Jalen Brunson everywhere, regardless of context, persist.

I frankly doubt that this is really a winning formula — which is why they’ve continuously come up short. I was pretty consistently on the “Yes, he’s a tentpole” side of the equation, but I’m not sure I’m there anymore.