Warriors select Jonathan Kuminga with the 7th pick of the 2021 NBA Draft

[FNQ]

He definitely has plenty to learn, but I don't think I'd agree that he's demonstrated no tangible growth.

I should have been clearer - statistical tangible growth. Like are those decisions resulting in anything different, is his rating going up, things like that. Overall his rating has cratered, but that's entirely on defense. And tracking, for all the good things it can bring up, will still miss things that it calls intangible (thus how I worded it) like making a better decision with the basketball. Even tracking can't piece together things like 'an open 3 is still a bad shot if there's X time on the SC' and similar situations, and he has done better there. He's had some flashes of improved man D too - only problem is early in the season, he was excellent at man D, it cratered also, and now he's below average with some impressive possessions in between

Overall though, the sheer amount of points/blown assignments hes had definitely overshadow those things. And maybe its Kerr's fault for throwing him into a scenario where he really had no chance. But at the end of the day, if there's one *player* who you can point at and say "they really hurt us", its him. People here are pointing at Wiggins, Klay, Dray.. but Kuminga's adventures on defense are the definite culprit according to tracking, and I whole-heartedly agree

[FNQ]

I should note too, while the rebounding numbers may have jumped, its not sustainable. He's boxing out at a lower percentage, fouling on rebounds at the same percentage, and just overall doesn't have anything to point to that shows improvement. His current style doesnt mean he wont ever rebound, it means that the variance of his impact will be very high because its more about luck than skill at this point. He could get away with that when he was one of the strongest/jumpiest players on the court. But not in the NBA, now he has to put in the work. And until that boxout % goes up.. I dont think there will be any consistency there. Its why he's one of the few Warriors who shows a deficit in team rebounding when on/off (1.5%).

To give context to that number, Nemo often subs in for Looney, and Nemo's deficit is -1.5% as well.

[FNQ]

rats jumping ship when the 19yo has a bumpy stretch in his rookie season.

Bumpy stretch? We're talking about his last 700 minutes. That's the majority of his season.

Its simple: you draft a player for the value they return in their first 5 years. Teams are not guaranteed to anything beyond those 5 years of control. Franz is so far ahead of Kuminga, and is trending upward while Kuminga is trending downward, both consistently. If you land on the option that Kuminga will be better than Franz in 5 years, that's a very homerish take. If you suggest that Kuminga will return more value than Franz in those 5 years, that's a hail mary.

Saying that Franz would have been a better pick, even without considering that we are trying to win now, isn't jumping ship. It's stating facts. He's much better right now and Kuminga has a lot of work to do to get to where Franz is now, even if Franz plateaus

[ILOVEIT]

Kuminga MIGHT be one of the kids that never picks up the ability to focus defensively. If that's the case then a Franz Wagner was the better pick.

But come on...."hail mary"? That's silly. He's 19...and as a 19 year old he's show the ability to take over and dominate offensively.

I'll take a gentleman's bet that Kuminga will be equal or better than Franz in 3 years.

[EvanZ]

At least according to DARKO they have fairly similar trajectories. I don't think we can say with much confidence who will be better in 5 years tbh. And I say this as someone who clearly favored Franz coming into the season.

[FNQ]

At least according to DARKO they have fairly similar trajectories. I don't think we can say with much confidence who will be better in 5 years tbh. And I say this as someone who clearly favored Franz coming into the season.

All boxscore/DRTG based +/- systems cannot account for the details of defense. The fact that Kuminga is trending up, in any metric, is a giant red flag to its accuracy. I mean just by the eye test we can see him be visibly confused and cause chaos on our defense.

To emphasize that, in terms of tracking, Kuminga is rated closer to Wiseman than Oubre last year in terms of defense. Again, not necessarily his fault, I wouldn't ever advise throwing a 19 year old rookie into the SF/PF position in our defense as those are easily the the two most demanding positions, but that's what he's going to play here, if he's successful. Meanwhile Franz's needle on defense has shot up since the ASB and is looking borderline elite.. he's essentially playing to the same impact that Wiggins was for us in the 1st half of the season, with less efficient offense.

Tracking obviously cant account for potential, but if people have been beating the drum about Wiseman being a bad fit and having an uphill battle to do well in our system, then Kuminga should be getting the same treatment. Expectations need to be lowered for the guy, not this future all-star/Giannis/Kawhi stuff.

And then on top of that, the things a good offensive option would need in today's game - a pullup, vision on drives, the ability to read defenses - is not there as well. Is it possible he's better than Franz in 4 more years? Sure.. but would it be playing the odds to think so? Because the MLM has that at a very low percentage, based on their current scores (and thats not even factoring in current trajectories, which in tracking, are going opposite directions)

[EvanZ]

At least according to DARKO they have fairly similar trajectories. I don't think we can say with much confidence who will be better in 5 years tbh. And I say this as someone who clearly favored Franz coming into the season.

All boxscore/DRTG based +/- systems cannot account for the details of defense. The fact that Kuminga is trending up, in any metric, is a giant red flag to its accuracy. I mean just by the eye test we can see him be visibly confused and cause chaos on our defense.

It's pretty clear that chart shows the last 4 games are trending down exactly what our eyes told us. He was playing well for a portion of the season until Kerr put him in the starting lineup.

[EvanZ]

It's a pretty clear mistake to have 100% certainty at this point about those two. Foolish imo.

[FNQ]

Kuminga MIGHT be one of the kids that never picks up the ability to focus defensively. If that's the case then a Franz Wagner was the better pick.

But come on...."hail mary"? That's silly. He's 19...and as a 19 year old he's show the ability to take over and dominate offensively.

I'll take a gentleman's bet that Kuminga will be equal or better than Franz in 3 years.

Yes, hail mary. Playing the odds, its extremely unlikely. And thats not speaking to Kuminga, thats speaking to Franz. Tracking sees him as a top 70 player right now (top 50 since ASB) and is trending up defensively. That means he's one of the more impactful defenders on his team while still being an average efficiency scorer at 15-18 ppg. That's really, really good.

So why am I saying its playing the odds? Part of the query machine we have access to has a MLM (machine learning model) and it basically looks at all the rookies since tracking's inception (only 2014 fwiw) and looks at their trajectory based on a ton of factors - age, size, position, and then all the relevant tracking stats as well, and it tries to predict out what the likelihood of a player rated lke Kuminga to a player rated like Franz. And basically, what its saying right now is that based on Kuminga's current numbers and impact, he will have a hard time matching Franz's impact currently, based on historical data.

Everyone is unique so obviously its not cut and dry, and the data only has 7 years of tracking history to go on. But the implication is that if Franz is already a key contributor defensively and has an offensive efficiency that should raise with experience and with the quality of teammates rising, he has the potential to be a great #2 option on a winning team.

Another factor is age - and people think this favors Kuminga, but it doesn't with regards to the 5 year draftee window. Franz will be 21-26 for his contract, Kuminga will be 19-24. Even if we say Kuminga will be as good as Franz when he hits 21, Franz will be 23 then, and where will he be?

So yes, if I'm looking at the data and it says that there's a ~10% chance that Kuminga is more impactful than Franz is right now, I'd say its a hail mary if Kuminga turns out better than him in the draftee window. It still could happen, but my guess is that if I polled non-Warriors/Magic fans about it, and then just Warriors fans about it, there'd be a steep variance in who thinks who will be better in that time. And thats all I try to do - take emotion out of it, present the data as is, even if it sucks to hear

[FNQ]

It's a pretty clear mistake to have 100% certainty at this point about those two. Foolish imo.

Kinda like reading what I wrote and thinking 100% is what I was saying?

[FNQ]

At least according to DARKO they have fairly similar trajectories. I don't think we can say with much confidence who will be better in 5 years tbh. And I say this as someone who clearly favored Franz coming into the season.

All boxscore/DRTG based +/- systems cannot account for the details of defense. The fact that Kuminga is trending up, in any metric, is a giant red flag to its accuracy. I mean just by the eye test we can see him be visibly confused and cause chaos on our defense.

It's pretty clear that chart shows the last 4 games are trending down exactly what our eyes told us. He was playing well for a portion of the season until Kerr put him in the starting lineup.

No, he had a couple games where he played well. It was sandwiched between a lot of games where he didn't. The second Draymond went out, the bottom fell out. Defensive metrics have never been good at identifying this, which is why everything is trending towards tracking. Its not like his defense was good all season and then only recently took a turn.. is that the implication?

Its a shame we cant play Portland and LAC all season, because for whatever reason he actually missed less than 3 rotations in those 2 games.. everything else has a large number by it. Its not a coincidence that Wiggins/Dray take their spots back, Kuminga gets almost no time, and all of a sudden our defense looks much better.

Put it this way: Kuminga has 26 games of over 20 minutes. In how many of those games has Kuminga been a positive impact player? Does DARKO have the ability to go game by game? Because tracking has 5 of those games as him being a positive impact player. Expected with a raw rookie.

[EvanZ]

Put it this way: Kuminga has 26 games of over 20 minutes. In how many of those games has Kuminga been a positive impact player? Does DARKO have the ability to go game by game? Because tracking has 5 of those games as him being a positive impact player. Expected with a raw rookie.

That's literally what DARKO was created to do. So yes. But it has "momentum" so it takes some time to convince the model that a player is extremely positive or negative. As it should be. I'll bet on DARKO long term over your eyes. And it's telling me it's way too soon to declare Franz the winner. Personally, I'd say it's like 70/30 Franz is better long term. But I have enough doubt to be reasonable about it. Especially having witnessed the Jordan Poole phenomenon recently.

You can read about DARKO yourself:

https://apanalytics.shinyapps.io/DARKO//

[EvanZ]

And thats all I try to do - take emotion out of it, present the data as is, even if it sucks to hear

You have also been presented with objective data (eg DARKO). FWIW.

[EvanZ]

Its simple: you draft a player for the value they return in their first 5 years. Teams are not guaranteed to anything beyond those 5 years of control.

With RFA it's really 8-9 years. You can count on one hand (much less actually) the number of players that took the QO and became unrestricted free agents after their rookie contracts.

[FNQ]

Its simple: you draft a player for the value they return in their first 5 years. Teams are not guaranteed to anything beyond those 5 years of control.

With RFA it's really 8-9 years. You can count on one hand (much less actually) the number of players that took the QO and became unrestricted free agents after their rookie contracts.

It definitely trends that way, but we're talking guarantees vs. likelihoods. You can only guarantee 5 years of control, but it is far more likely that you get more. I dont have any say in that measurement, but I'd imagine its 5 out of convenience for the people doing tracking, since tracking data only has 8 years behind it. It would make sense that they eventually kick it out further.

But still beyond 5 years has a great deal of variance that the first years likely wouldnt. Matching a player who doesnt want to stay, matching to recoup pick value, getting overbid, those kinds of things that are a bit more intangible. But without knowing for sure, I'd still wager they eventually do kick the window out as the dataset grows

[FNQ]

Put it this way: Kuminga has 26 games of over 20 minutes. In how many of those games has Kuminga been a positive impact player? Does DARKO have the ability to go game by game? Because tracking has 5 of those games as him being a positive impact player. Expected with a raw rookie.

That's literally what DARKO was created to do. So yes. But it has "momentum" so it takes some time to convince the model that a player is extremely positive or negative. As it should be. I'll bet on DARKO long term over your eyes. And it's telling me it's way too soon to declare Franz the winner. Personally, I'd say it's like 70/30 Franz is better long term. But I have enough doubt to be reasonable about it. Especially having witnessed the Jordan Poole phenomenon recently.

You can read about DARKO yourself:

https://apanalytics.shinyapps.io/DARKO//

You think I'm using my eyes here? Where do you think all this info is coming from? Its coming from data, and frankly, better data. Because DARKO is automated, numeric, it doesn't have any subjectivity to it. How can DARKO account for missed rotations?

As I said before, which is a trend with you that is really really troubling - Franz has been the better selection thus far, by a lot. If someone is saying that Kuminga would be better in ORL, or that you think he'll be be better than Franz based on what we know now, that's a homer take, and is on the wrong side of 90% in terms of the odds. I really think its weird to have a high horse of you eyeballing 70% as an ok number, yet a machine learning model based on the last 8 years of rookies has it at 90.3%, and having one of those numbers be foolish and inexcusable

[HiRez]

At least according to DARKO they have fairly similar trajectories. I don't think we can say with much confidence who will be better in 5 years tbh. And I say this as someone who clearly favored Franz coming into the season.

All boxscore/DRTG based +/- systems cannot account for the details of defense. The fact that Kuminga is trending up, in any metric, is a giant red flag to its accuracy. I mean just by the eye test we can see him be visibly confused and cause chaos on our defense.

The overall trend doesn't seem wrong. Because Kuminga did clearly trend up in the second half of the season, just as he is now trending sharply downwards, and the graph shows that. The graph shows Kuminga reached a peak and started declining about 10-15 games ago (the line appears to be a smoothed moving average that will be delayed from the data points). Seems about right. And honestly the entire team has been playing terrible defense lately.

Kuminga ranks #1 in PTS/36 and #7 in REB/36 among qualified rookies. Considering he's only 19 (Franz is almost 21) he's had a pretty spectacular season. Some bumps on both ends of the floor aren't unexpected and I think it's too soon to draw conclusions that he won't be good defensively in the future. If you had told be after the first couple months of the season that Kuminga would finish the season shooting over 30% from 3 and would be one of the top rookies in TS% and EFG% I'd have said you were nuts but here we are. And despite some pretty bad defensive performances, I feels like he was better than expected on that end as well. That said, Franz is definitely better right now and has been all year. But Kuminga's future looks bright if he can hold onto the ball.

[and1GS]

I think Kuminga is a pretty solid on-ball defender most of the time. Where he breaks down is in passing off that assignment around a screen, or general off-ball confusion. Have seen him trying to process the 'do I help? do I stay home? wait is this my switch?' many times.

I think he will pick it up. Moody was always expected to be a better/more cerebral defender and only now is he (mostly) understanding our system. Think it's important to put this all in perspective that we expected nothing from Kuminga this year and that he would be a massive project. Seeing progress throughout the season was promising and he's flashed an incredible skillset. One that needs to be rounded out, for sure, but he has the tools.

[EvanZ]

Put it this way: Kuminga has 26 games of over 20 minutes. In how many of those games has Kuminga been a positive impact player? Does DARKO have the ability to go game by game? Because tracking has 5 of those games as him being a positive impact player. Expected with a raw rookie.

That's literally what DARKO was created to do. So yes. But it has "momentum" so it takes some time to convince the model that a player is extremely positive or negative. As it should be. I'll bet on DARKO long term over your eyes. And it's telling me it's way too soon to declare Franz the winner. Personally, I'd say it's like 70/30 Franz is better long term. But I have enough doubt to be reasonable about it. Especially having witnessed the Jordan Poole phenomenon recently.

You can read about DARKO yourself:

https://apanalytics.shinyapps.io/DARKO//

You think I'm using my eyes here? Where do you think all this info is coming from? Its coming from data, and frankly, better data. Because DARKO is automated, numeric, it doesn't have any subjectivity to it. How can DARKO account for missed rotations?

I haven't seen your data. Where is it? Why don't you share some of it with us so we can make our own
decisions on it? Or are we just "supposed to trust you"?

As for "subjectivity", why would I want that in a ML model? You want ML models to be objective. That's the whole point.

[Big J]

I don't think you guys realize how dominant he'd look on Orlando, Detroit, or Houston with big minutes. He'd easily be in the ROY convo.