The logic of Steve Kerr's coaching decisions pt. 1: Eric Paschall vs Juan Toscano Anderson, and POOLE WATCH BABY!!!
Eric Paschall's brutal lineup combos, JTA the net rating god, and irrational Jordan Poole optimism.
AUTHOR’S NOTE: I wrote this on April 26, before the Warriors’ disaster of a game against the Dallas Mavericks the next night.
For all of my frustrations with Steve Kerr this year, I’ll happily concede that he’s done a mostly great job of coaching the Warriors since James Wiseman went down with a torn meniscus. Kerr’s coaching decisions in the wake of Wiseman’s injury will be discussed at further length, but they basically come down to playing Juan Toscano Anderson consistent minutes, Eric Paschall being out of the rotation, banishing Nico Mannion to the shadow realm in favoring of letting Jordan Poole run the second unit, and playing Kelly Oubre off the bench and/or in lineups where he’s the 4 or is next to one of JTA or Draymond Green.
So let’s go through these decisions one by one.
Paschall vs JTA
Juan Toscano Anderson is a legitimate NBA role player, if not a starter on the right team — say Miami or Portland — but he didn’t get consistent minutes in the rotation until Steve Kerr put the kibosh on Eric Paschall. I’m going to talk at length here about Paschall because his bad performance is what opened the door for JTA to solidify himself in the Warriors’ rotation.
I can understand why Eric Paschall received minutes to start the season. He proved himself an efficient paint scorer on last year’s dumpster fire of a roster and he had a really nice streak at the beginning of this season where he was punishing full sized 5’s off of the dribble. But Paschall is a 6’6 two foot jumping ISO specialist with a bad jumpshot. A player with those limitations has pretty much no margin for error and the moment that Paschall started showing up on injury reports with minor tweaks, his performance suffered. This makes sense — if you’re a 6’6 foot two foot jumper whose shot doesn’t command respect, you probably have to be at your physical peak to manufacture buckets in the paint at an efficient rate.
Over the summer, there were various reports published that mentioned Paschall re-working his jumpshot. That lasted, what, all of two games? It’s possible that Paschall is better off shooting that dolphin kick jumpshot — he increased from 28.7% to 32.1% 3P on identical volume per 36 minutes — but the bigger issue here is that Paschall’s impressive restricted area efficiency went down and at this point, he’s such a bad rebounder and defender that he’s difficult to find a spot for. In his rookie season, Paschall shot an impressive 68% in the restricted area, 42.6% in non-restricted area paint, and 43% from the mid range. This year, those numbers were 64.2%, 37.3%, and 49.3%. The mid range number is obviously encouraging, but frankly, mid range shooting is not something that the Warriors desperately need from their bench units.
It’s possible that these percentage drops would be more tolerable if Paschall had shown literally any improvement as a rebounder or in his ability to rack up counting stats on defense, but that just didn’t happen. Paschall averaged 6 boards per 36 minutes as a rookie, 1.8 of them offensive rebounds (I think this is demonstrates the fundamental problem with Paschall; his impressive stature and explosiveness really only show up on the box score when he has a chance to score) and only averaged 6.6 boards per 36 minutes this year, a fairly below average rate at the power forward position and an even more alarmingly low rate for someone spending significant minutes at the 5, which accounted for 67% of his minutes this year per BBRef. Paschall averaged exactly 1 stock a game both this year and last, indicating no progress on the defensive front.
It’s certainly possible that Paschall, like Wiseman, was the victim of playing lots of minutes in lineups with poor spacing (he’s only a part of three positive two man combos to receive >100 minutes: Steph/Paschall +1.4, Bazemore/Paschall +6.6, Mulder Paschall +13.8), but simply put Paschall isn’t an organization priority as he was a late 2nd round pick AND JTA fits this team much better. The two man combos with negative net ratings are littered with Paschall’s name. Behold, these particularly gruesome numbers:
Paschall/Oubre -20.8 in 281 minutes
Paschall/Wiseman -20.2 in 102 minutes
Paschall/Poole -17.9 in 156 minutes
Paschall/Looney -17.3 in 103 minutes.
Now look at JTA’s notable two man combos to receive >100 minutes (I’ve bolded the positive ones):
Draymond/JTA +10.5 in 362 minutes
Bazemore/JTA +6.8 in 199 minutes
Steph/JTA +6.6 in 463 minutes
Poole/JTA +6.2 in 224 minutes
Wanamaker/JTA +6.2 in 224 minutes
Wiggins/JTA +6.1 in 514 minutes — JTA has played 783 minutes on the year so this is seriously impressive as this accounts for 65% of his minutes on the year!!
Lee/JTA +5.6 in 214 minutes
Mulder/JTA +2.6 in 137 minutes
Looney/JTA -0.1 in 106 minutes
Oubre/JTA -3.7 in 442 minutes (!! 54% of his minutes on the year)
JTA/Wiseman -4.3 in 154 minutes
Mannion/JTA -5.4 in 108 minutes
Pretty much the only way to get bad minutes out of Juan is to pair with Wiseman, Oubre, or Mannion, whose unnecessary and unearned minutes piss me off so much that I’m using all my restraint to not spend a whole ass page laying out in excruciating detail how terrible he was this year. Simply put, Juan is a winner. He reminds me of a mix of the best traits of Draymond Green and Matt Barnes. I don’t think Juan has a much higher ceiling than what he’s already doing because I doubt he’ll ever turn into a volume three point shooter and he religiously avoids mid range shots — he’s taken literally 5 mid range shots in his entire career — and that probably indicates a limited ability to create his own efficient looks.
But what Juan does, he does extremely well. He’s a super smart passer on the short roll, he has a sixth sense for where Steph is at all times and is Draymond-esque in his ability to hunt DHO’s for Steph 3’s, and he knows when to slip DHO’s for dunks. Per Cleaning the Glass (as of April 26th), Juan is in the 96th percentile at his position (“big” per CTG) in points per shot attempt at 1.41 PSA. Considering Juan pretty much only shoots open threes (41.9% with 4.1 attempts per 36 minutes) and on mildly contested drives, this makes total sense. If Juan has one glaring flaw, it’s that he turns the ball over A LOT on PNR situations (28.6% TOV frequency when he’s the roll man, Draymond by comparison turns it over at “merely” 15.6% TOV frequency as the roll man). But much like Draymond, these turnovers come from trying to make home run passes. I can live with recklessness from smart players, which is also something that I think applies to Jordan Poole, who we’ll talk about next.
POOLE WATCH BABY!!! — AKA, Why was Jordan Poole not handling the ball to begin the season?
I’ll admit, I’m irrationally excited about Jordan Poole. This is probably in part because his success allows me to take a victory lap on a prediction I made in the midst of Poole’s awful rookie season when I said he’d have a better NBA career than Eric Paschall. I think that looks conservative in retrospect and frankly, I wouldn’t be shocked if Poole is worth north of $20 million a year halfway through his second contract.
What excited me more about Poole’s game more than anything last year was his obvious feel for the game. Despite being physically overmatched, Poole managed to shake loose against NBA defenders (even if he couldn’t finish his shots or get off clean, uncontested looks) with an impressive array of ball fakes, jab steps, and misdirections. Sometimes it would be as simple as Poole leaning one way before exploding the other direction. In a way, it reminded me of Klay Thompson, who always made use of the supreme threat of his shot to make the most of those leans and misdirections. But obviously, Poole was historically bad shooting the ball last year (33/28/79 splits), so the Klay comparison pretty much ends there.
As Poole’s rookie season progressed, he started showing passing flashes. Seriously, go watch this video. You’ll see an impressive ability to manipulate defenders by looking off the actual passing target, Poole holding onto the ball just long enough on a drive to find his cutter, Poole hitting shooters with crosscourt laser passes, Poole lobbing it up over and over for Marquesse Chriss — we miss you! — and so on. I figured that if he could become even a league average shooter and get the ball in his hands every now and then, he could be an interesting combo guard off of the bench, so long as he got his defense up to Kerr’s standards.
I had my doubts about his shot, but there were encouraging signs. After D’Angelo Russell got traded and Poole no longer had to let Jacob Evans handle the ball, he averaged 14.2 points, 3.9 assists, 2.6 rebounds, and 0.7 stocks on 48/33/73 splits. At that point Ky Bowman was the only other player with any credible claim to the point guard position, so Poole started getting more reps on-ball and his passing looked good. But most importantly, as his shot improved, defenses started treating his shot with genuine respect. It was as if teams understood that Poole was not someone you wanted to get hot.
At some point in the summer after last year’s dumpster fire of a season, @antonin_org posted on Twitter that Poole had shot nearly 60% in the restricted area after being recalled from the G-League on January 4, 2020 AND he’d finished the year shooting 50% from floaters. Couple that with Poole’s solid FT% (79.8) and I had reason to believe that he had the type of natural touch that would eventually manifest itself. A few months before the 2020-21 season, I came across an interesting interview with someone who I thought said a lot of smart things about Jordan Poole. That person was... Kirk Lacob. Seriously, read this response to a question about Poole’s shooting ability:
His natural ability with the ball and shooting the ball is extremely high. And it's something we identified when we drafted him. The reason he maybe didn't go higher than where we were drafting is because he was still figuring out how to harness it. You've got all this natural ability, but he had to learn how to actually utilise it.
I think his base was a little weak. And like I said, our focus was on adding strength, he has put on a lot of core strength. He has worked a lot on his base. He tends to shoot difficult shots and difficult shots are even more difficult when you really don't have that base and get square.
One thing that is, not to compare the two, but one thing that's always been so impressive about Steph Curry is he takes all these wild, seemingly wild, crazy shots, but he's so strong in his core and his base, that he's actually able to get square and get settled on all those attempts.
We really worked with Jordan, something he identified in film with the coaches was, okay, you have the natural ability to shoot all these highly difficult shots for most players, but you have the touch. It's all about just getting that foundation, that base down so that even though you do something crazy off the dribble, crazy on the screen, you're actually getting square every time and allowing your natural ability to take over. And that's something that he's worked a ton on, and that we think is going to be a strength moving forward.
The game just needs to slow down for him. Frankly, he has all the tools in the world. And I think he's gotten better and better about understanding that you don't have to play, you know, in like sixth gear all the time, you can actually shift up and down and sometimes playing in second and going to fourth and then back to third can actually be valuable. So he's learning a little bit about change of pace, not just in the way he's moving, but in the way he's thinking and the moves that he's making. And I think that's going to be really valuable.
I’ll admit, this quote appealed to me because it lines up with my belief that it’s easier to maximize the body of a guy with feel than it is to teach an athlete how to play. Most of what Lacob says here is pretty sensible. Poole looked weak last year. He looks less weak this year, but even then, his shooting splits at the time of writing (April 26: 43/35/88) are pretty much in line with what he did after the DLo trade.
So, going into the season I was pretty excited about the idea of Jordan Poole handling the ball in the second unit with Brad Wanamaker next to him to defend and hit open shots (The Athletic’s Anthony Slater noted when the Warriors signed Wanamaker that he’d shot it at 48% in catch and shoot situations the previous year with the Celtics). This was not what happened. Poole played in 6 of the 7 first games of the season, mostly as an off-ball guy while Wanamaker operated with the ball in his hands. Poole racked up a -28 +/- in those games before Kerr excised him from the rotation. After Brad Wanamaker slowly played himself back into the fringes of the NBA to the tune of 35/21/89 splits (this is actually worse than Poole shot in his rookie season), Steve Kerr’s hand was forced and Jordan Poole finally got a shot in the rotation again... but next to Nico Mannion. I said I wasn’t going to spend a lot of time on Mannion, but here we are. Talking about Mannion is necessary to discuss Poole’s role on this Warriors team, what it means for the future, and how Poole does and does not fit with Kerr’s conception of basketball.
I had to filter for two man combos to play >50 minutes to account for Mannion’s mercifully brief stint as a rotation player to look at his net ratings, which are... quite bad.. He’s only a part of one positive combo with Mulder (which was almost certainly accumulated in garbage time) and he had five two man combos with a net rating worse than -10. What exactly did he provide? I’m still not sure, but Steve Kerr claimed to like that Mannion was a “true point guard,” which as I understand it, means that Mannion is someone he believed would keep the ball moving, not take risks, and wouldn’t go out of his way to hunt his own shots. It’s funny though, my last memory I have of Nico Mannion playing meaningful minutes is of my Twitter timeline, Marcus Thompson included, melting down at the audacity of Mannion’s shot selection.
If you’ll recall, one of Steve Kerr’s purported peeves about Poole is that he’s a high risk, turnover prone combo guard, which Marcus Thompson has hinted at in a few Warriors Plus Minus podcasts. About those turnovers... Jordan Poole now ranks 82nd percentile in TOV% at his combo guard position, per Clearing the Glass. Nico Mannion ranks... 2nd percentile at the point guard. Whoops!
This is really quite funny if you think about it! We know that Steve Kerr really values defense, to the point of having sacrificed offensive firepower for defensive prowess on countless occasions — most famously in Game 7 of the 2016 NBA Finals. Nico Mannion is a rookie and most rookies are bad on defense, particularly rookie point guards. Jordan Poole, while an intriguing player who has made some defensive improvements, is still mostly bad on defense, so the two of them together would probably be a punching bag on defense. And yet, Steve Kerr was willing to tank his bench defense in order to take the ball out of Jordan Poole’s hands and give it someone he thought would be more trustworthy, only it turned out Nico was INSANELY turnover prone!
So bringing it back to Poole and Kerr — there was a reasonable body of evidence that suggested Brad Wanamaker would be best utilized off-ball with someone else, say, maybe a combo guard with passing skills and microwave scoring ability, taking the lead ball-handling duties. Steve Kerr decided to play Brad Wanamaker on-ball and tossed Jordan Poole into the dustbin after 7 games and this decision to stick with Wanamaker in that role did not help either A. chase wins or B. develop the young players. It was only after Wanamaker failed miserably and Poole got some minutes in the G-League that we finally got to see Poole get real minutes and even then, they were at first tied to Mannion.
In spite of all that, Poole has performed really well and has mostly exceeded my expectations. I think it’s likely he’s a career 35% three point shooter on high volume and high variance, if only because he’d be a borderline All-Star if he shot >37% from three on his diet of shots (Damian Lillard, for example, is a 37% career three point shooter), which speaks to how hard it is to be shot above league average from three if you’re shooting a lot of them off the dribble or on the move. But the things that I’m going to keep track of in this recurring POOLE WATCH BABY!!! section are his restricted area FG%, his floater FG%, and his diet of three point shots, which I think will ultimately determine his ceiling
Right now, Poole is shooting 66.2% from the restricted area. That’s... higher than what Steph Curry is shooting in the restricted area this year, albeit on much lower volume. I don’t think it’s reasonable that Poole will hover that high forever — I’ve been tracking that number all season and for a while he was in the high 70’s, which would put him in prime LeBron territory and obviously as the sample size has increased, his RA efficiency has come down to earth to merely elite percentages for a guard. But in 26 games after his G-League stint last year, Poole shot nearly 60% in the restricted area and he’s now given us 40 games at 66.2%, so we have almost an entire season’s worth data that shows Poole has borderline elite, if not elite finishing ability for a guard. As I mentioned earlier, Poole shot 50% on floaters last year. He’s upped that to 52.4% on 36 attempts this year.
Now things get interesting when we look at his shot splits and tracking shot dashboard on three point attempts. Poole is shooting 35.2% from three this year on nearly 200 attempts but only 24 attempts have come from either corner! He’s shooting 61.5% from the left corner and 54.5% from the right corner, but 174 of Poole’s attempts, nearly 88% of his attempts on the year, have come above the break, where he’s only shooting 31.6%. Poole is shooting 37% from three on the year when he gets a catch and shoot look and only 32% on pull-ups. That is, however, a marked improvement from last year when his catch and shoot percentage was... 29.9% and his pull-up percentage was 23.8% on threes. In his putrid rookie season, Poole shot 27.3% from the left corner 3 on 22 attempts, 40% from the right corner 3 on 15 attempts, and 27.4% on 226 attempts on above the break threes.
As I mentioned before, Poole’s ceiling is tied to his three point percentage. The high level finishing and touch on the floaters is super encouraging, but at the end of the day, if Poole can shoot above league average from three on a volume of difficult shots, his ceiling becomes extremely high. I also want to stress just how difficult this is to actually do! Damian Lillard has only cleared 40% from three once in his career and shoots it at 37% career, which is exactly league average. Donovan Mitchell is shooting a career high 38.6% this year (36.2% career). Luka Doncic shoots at 35.9% this year (33.2% career) and James Harden’s 36.1% is a shade below his 36.3% career three point percentage. Obviously I am not claiming that Poole will become better than any of those players if he becomes a >37% three point shooter next year, but my point is that he’s already shooting perfectly acceptable percentages for his role and shot diet and there’s plenty of reason to believe those percentages can improve as he matures.
One last Poole stat before we move onto the next section: the Warriors best two-man combo per net rating is Curry/Poole, which has a +23.1 net and a 96.9 DRTG (!). Playing two shooters and creators at once has its benefits, who could have guessed?
“The logic of Steve Kerr's coaching decisions pt. 2: Kelly Oubre’s role,” will be published tomorrow morning!