That's right folks. In 12 minutes without Luka the team gets better! I wonder what they do in 48-minutes without---

Huh. Well then they must really suck when Luka--

Well, it's probably just a fluke. Luka's team was bad the next--


Wow, Luka actually looks really good in the regular-season! And he must be really really good in the playoffs! But let's just check the net-rating, I'm sure it doesn't suggest he's a nega--


Ooof. Well maybe the haters were right. Luka doesn't impact--


Or maybe he does?
Okay, let's combine our samples:


Phew. Luka is good!
For a little less uncertainty, let's extend this a bit:
[img]https://media.discordapp.net/attachments/807803459331555363/1169265589144010793/image.png?ex=6554c670&is=65425170&hm=5ee1e7fb7b14d0a0f0abc4e0b015c0e566ead3324786994ad3deb1401a474b43&=&width=1434&height=435
[/img]

I think the answer to "does luka impact winning" is probably yes. And yes, that was something someone directly questioned because of that on/off graph(i wonder if there's anything to learn from this...). But maybe he isn't --as-- impactful as you would expect. Of course if anything this confirms that player-evaluation is an exercise in uncertainty, and he does do a fare bit better by record(helio-led teams do tend to do better in the clutch).
He also shares alot in common with, at least emperically, most valuable offensive players ever with a high-volume of ball-handling to go with high-volume and effecient scoring as well as creating a bunch for teammates. Maybe he just got unlucky!
Regardless, I think whatever you decide Luka's ability to affect winning is, it's important to look at the whole picture, and not just try and stack-metrics indiscriminately.