Read on Twitter
I really like this question and clip from Brett.
There's a few things that I would highlight that are at play:
1. Shot selection and relation to the shot clock
2. Live reads and relation to the shot clock
3. Game flow and gamesmanship
https://twitter.com/KornHoops/status/1349058915632340993
There's a few things that I would highlight that are at play:
1. Shot selection and relation to the shot clock
2. Live reads and relation to the shot clock
3. Game flow and gamesmanship
https://twitter.com/KornHoops/status/1349058915632340993
1) Shot selection and relation to the shot clock
My benchmark for teams, and especially clients, is always 1.00 PPP. In almost every situation, save for some low-usage, poor shooting forwards in a one-a-game catch-and-shoot scenario, I look at any sub-1.00 PPP shot as needless.
My benchmark for teams, and especially clients, is always 1.00 PPP. In almost every situation, save for some low-usage, poor shooting forwards in a one-a-game catch-and-shoot scenario, I look at any sub-1.00 PPP shot as needless.
Floaters, on average, are sub-1.00-PPP shot attempts.
In 2019-20, there were 12,130 floaters taken during the season, and the cumulative PPP of the shots was a hair under 0.86. So, on average, a decision to take a floater results in basically a 43% shot from two.
Not great.
In 2019-20, there were 12,130 floaters taken during the season, and the cumulative PPP of the shots was a hair under 0.86. So, on average, a decision to take a floater results in basically a 43% shot from two.
Not great.
Obviously, not all floaters are created equal. Some of these floaters are sub-0.60 PPP shots. Some are, on average, 1.00 PPP or better shots.
All that we know is that, as a collective, there is a predisposition to take too many floaters, as evident by their below average PPP.
All that we know is that, as a collective, there is a predisposition to take too many floaters, as evident by their below average PPP.
This bias towards taking too many floaters is very evident when comparing the volume and PPP of possessions occurring in the final 4 seconds of the shot clock
Late in Clock: 23,848 possessions, 0.82 PPP
Floaters: 12,130 possessions, 0.86 PPP
Nearly identical efficiencies.
Late in Clock: 23,848 possessions, 0.82 PPP
Floaters: 12,130 possessions, 0.86 PPP
Nearly identical efficiencies.
So, without divulging exactly what I teach my clients to look for before they should be considering taking a floater on a given possession that's outside of late clock scenarios, and, more importantly, the messaging as to why this should be the case...
...It's safe to assume that for all but the players who are proven to be above 1.00 PPP on floaters (and there were 19 of them with at least 50 attempts last year), the average NBA (or college) player shouldn't be hunting for them in actual game reps.
And De'Andre isn't here.
And De'Andre isn't here.
Without going into specifics, the most evident parts of my criteria are being met here:
De'Andre is not hunting a floater.
De'Andre is quite close to the rim.
De'Andre is going up against a good rim protector in good position.
De'Andre has the space to get a floater off.
But...
De'Andre is not hunting a floater.
De'Andre is quite close to the rim.
De'Andre is going up against a good rim protector in good position.
De'Andre has the space to get a floater off.
But...
2. Live reads and relation to the shot clock
And this is what I think is fantastic about this situation:
De'Andre is not only not actively hunting his own floater, he is hunting and scanning to try to figure out if he can get another player (Clint Capela) a dunk.
I love this.
And this is what I think is fantastic about this situation:
De'Andre is not only not actively hunting his own floater, he is hunting and scanning to try to figure out if he can get another player (Clint Capela) a dunk.
I love this.
During his final, Sophomore year at Virginia, De'Andre had 43 tracked PnR + pass-out possessions. A shade above 1 a game.
He was excellent against that compeition on that small sample:
Individual possessions: 22 possessions, 1.18 PPP
Overall: 43 possessions, 1.05 PPP
He was excellent against that compeition on that small sample:
Individual possessions: 22 possessions, 1.18 PPP
Overall: 43 possessions, 1.05 PPP
When De'Andre was drafted, I don't expect anybody really expected him to be anything more than 3&D...
Well, I did.
I loved his analytical profile/techniques, I love that he works with @DaminAltizer (basically a cheat code for valuable skills). https://twitter.com/JosephGillMA/status/1131327637509558274
Well, I did.
I loved his analytical profile/techniques, I love that he works with @DaminAltizer (basically a cheat code for valuable skills). https://twitter.com/JosephGillMA/status/1131327637509558274
And while I don't think he's exactly a natural for these guard skills, he's worked hard on his technique and it shows in the angle he takes off the screen, the dribbling until he's forced into a pick-up, and most importantly, his eyes.
This entire drive, his head is up, and his speed isn't exactly a sprint, but it's quick, it's under control, and he can always put a burst on if he needs to. This is, my educated guess, on how the progression went for De'Andre:
1. Space after the screen, continuing to drive is the right call.
2. Identify the primary help defender (Embiid), and the coverage (deep drop).
--> What does this mean?
---->Likely my man is trialing
---->Strong corner won't need to help
3. Glance strong corner. No overhelp.
2. Identify the primary help defender (Embiid), and the coverage (deep drop).
--> What does this mean?
---->Likely my man is trialing
---->Strong corner won't need to help
3. Glance strong corner. No overhelp.
No easy pass strong corner for a three.
4. Live passing read is Embiid. Will be guess and recover too soon?
--> After a beat, nope, Embiid is remaining in good, centerfield position. Unlikely I'll be able to take this myself.
5. Deep drop, centerfield... Lob pass..?
-->New read
4. Live passing read is Embiid. Will be guess and recover too soon?
--> After a beat, nope, Embiid is remaining in good, centerfield position. Unlikely I'll be able to take this myself.
5. Deep drop, centerfield... Lob pass..?
-->New read
At this point, Embiid is no longer the primary read, it's the weak-side tagger (Scott).
Unless this is a perfect anticipation by the last remaining defender (assuming Hunter's man doesn't gamble and abandon him completely), 4 of the 5 defenders are safely accounted for mentally.
Unless this is a perfect anticipation by the last remaining defender (assuming Hunter's man doesn't gamble and abandon him completely), 4 of the 5 defenders are safely accounted for mentally.
This is now a binary passing read (if floater isn't being considering and, hint, check De'Andre's feet, he was never thinking floater at any point on this drive):
1. Scott doesn't tag, alley-oop to Capela.
2. If Scott tags, even if Collins isn't in position for CaS, kick-out
1. Scott doesn't tag, alley-oop to Capela.
2. If Scott tags, even if Collins isn't in position for CaS, kick-out
IF there was 4 seconds or less on the shot clock, I would be advocating for a more aggressive, and nuanced read:
1. Scott doesn't tag, alley-oop to Capela.
2. If Scott tags... Embiid's the live read again
2a: Embiid's cheating up: Strong rim attack
2b. Embiid's back: Floater
1. Scott doesn't tag, alley-oop to Capela.
2. If Scott tags... Embiid's the live read again
2a: Embiid's cheating up: Strong rim attack
2b. Embiid's back: Floater
So, as played, I think De'Andre makes the correct read for the vast majority of the situation.
Maybe this would be a 1.10 PPP or so floater. Maybe a little better.
But, kicking out could lead to:
1. Collins open CaS three (1.27 PPP last 2 years)
2. Kick to Trae, CaS (1.34)
Maybe this would be a 1.10 PPP or so floater. Maybe a little better.
But, kicking out could lead to:
1. Collins open CaS three (1.27 PPP last 2 years)
2. Kick to Trae, CaS (1.34)
I'm never going to advocate a player routinely leaving their feet on a binary read, but De'Andre was in full control of this situation, appears to make all the right reads, and he's either passing into a dunk, an open CaS three, or a possession reset when he executes this read.
Basically, the passing upside is massive, the downside is every once in a while a defender makes a great individual gamble and picks this ball off. That's rare. A natural outcome is a mid-clock reset.
For taking the floater, the upside is, maybe he takes a 1.10 PPP shot here and there. Also, remember, collectively NBA players overestimate their abilities on game rep floaters, though.
The downside is a potentially a telegraphed, then contested, sun-1.00 PPP floater.
The downside is a potentially a telegraphed, then contested, sun-1.00 PPP floater.
And that shot would come at the expense of potentially making the odd-on correct play.
Not to mention, basketball is a social game. Passing the ball in 50/50 situations can inspire teammates to do the same, sometimes passing out of 45/55 situations they otherwise wouldn't.
Not to mention, basketball is a social game. Passing the ball in 50/50 situations can inspire teammates to do the same, sometimes passing out of 45/55 situations they otherwise wouldn't.
Also, this decision to default to a pass is probably in De'Andre's best long-term interest due to...
3) Game flow and gamesmanship
First and foremost: This wasn't a simple decision tree, but it appears De'Andre followed some pretty binary positional reads. He didn't have to once attempt to read a weight distribution or anticipate future movement.
That makes life much easier.
First and foremost: This wasn't a simple decision tree, but it appears De'Andre followed some pretty binary positional reads. He didn't have to once attempt to read a weight distribution or anticipate future movement.
That makes life much easier.
Spatial awareness of opponents is natural for some hoops players, it's learned behavior for others, it's an alien concept to a lot of those who never played at a high level.
Not exactly easy, but doable.
Trying to juggle it with reads of weight distributions... That's hard.
Not exactly easy, but doable.
Trying to juggle it with reads of weight distributions... That's hard.
Maybe this was one of those situations where a floater could've been gotten off cleanly from a short distance. If I had guess, yeah, guess he could've.
But the majority of the time, there isn't space from close enough to make it worth it against a defender you shouldn't attack.
But the majority of the time, there isn't space from close enough to make it worth it against a defender you shouldn't attack.
But, while these reads are easy, they're able to be executed quickly, accurately, and consistently.
More variables to weight, the harder that is.
It's also worth mentioning that De'Andre has been making good decisions in 2020-21. He's 4 for 8 on his floaters, above average!
More variables to weight, the harder that is.
It's also worth mentioning that De'Andre has been making good decisions in 2020-21. He's 4 for 8 on his floaters, above average!
De'Andre is sacrificing a bit of his ability to potentially make the perfect read every single time to make a very accurate read almost every time.
In the long-run, this approach makes his life easier, and makes him and his team more efficient as he's right more often.
In the long-run, this approach makes his life easier, and makes him and his team more efficient as he's right more often.
In a situation where a coin flip pays uneven odds, De'Andre is basically deciding he won't try to use syntax clues to guess which flip will be heads or tails, he's noticed that heads pays out at better than even odds, so he's guessing heads, profitably, every time.
He's smart.
He's smart.
And also, he's sending his opponents a message in the innergame play that the 76ers, and any opponents who are scouting, aren't able to bait him into potentially inefficient decisions easily.
A floater should almost always be a counter, depending on how efficient a player is.
A floater should almost always be a counter, depending on how efficient a player is.
The problem with a game that moves quickly is that most players can't counter an opponent's counter on a shot attempt (post-moves being an exception).
If De'Andre uses his counter here, on the first play of the game, Embiid knows he might be able to bait, then counter, later.
If De'Andre uses his counter here, on the first play of the game, Embiid knows he might be able to bait, then counter, later.
Think about these two possessions:
1. On this play, De'Andre takes this floater, and it's 1.10 PPP shot.
De'Andre isn't exactly quick into his floaters. He's not Steve Nash, that's totally fine for his position/archetype. Not yet worth spending developmental time on a niche.
1. On this play, De'Andre takes this floater, and it's 1.10 PPP shot.
De'Andre isn't exactly quick into his floaters. He's not Steve Nash, that's totally fine for his position/archetype. Not yet worth spending developmental time on a niche.
Embiid understands that if he gives De'Andre a certain look, he can force a certain shot, potentially.
Later in the same game, Embiid gives De'Andre the same read, yet he's lying in wait. De'Andre, on the move, starts an upwards motion. Joel pounces on a contest: 0.60 PPP shot.
Later in the same game, Embiid gives De'Andre the same read, yet he's lying in wait. De'Andre, on the move, starts an upwards motion. Joel pounces on a contest: 0.60 PPP shot.
This is the problem with an overreliance on counters, even when they work in single instances, they can be manipulated against you at later times, unless you can also hit at a high rate against a dedicated defense against the counter (few, few can).
All the sudden, instead of trusting the process, De'Andre is potentially engaging in mental warfare against Embiid.
He isn't making reads, he's trying to anticipate what Embiid will do, fakes or not, before he does.
Dangerous predicament.
Those 2 possessions, xPPP: 0.85 PPP.
He isn't making reads, he's trying to anticipate what Embiid will do, fakes or not, before he does.
Dangerous predicament.
Those 2 possessions, xPPP: 0.85 PPP.
