So I've been thinking about our perimeter players and why they are perhaps not playing at an optimal level. My focus for this discussion is on offense and whether or not each player is operating in the best role for him. Specifically, I'm looking at the percent of a player's baskets that are assisted compared to the player's assist% (passing to others) and eFG%.
Let's start by looking at the team as a whole. Here's a table showing Duke's assist% (which for the team is the same as %assisted) vs. eFG% in each of the past 10 seasons:
Code:
Year %asstd eFG%
2018 57.6 55.9
2014 54.8 53.8
2015 53.5 56.6
2019 52.4 53.6
2013 52.1 53.9
2020 52.1 52.6
2021 50.4 51.3
2016 48.6 53.7
2012 48.4 52.8
2017 47.5 54.8
As you can see, the Duke teams with the highest assist% tend to shoot well, but there may not be a true correlation since the Duke team with the lowest assist% has the 3rd-best eFG%. It may be worth noting that this year's team has a fairly low assist% as well as the lowest eFG% over (at least) the past 10 years.
My analysis today involves the performance of individual players. I'm looking at perimeter players only, mostly from the past 5 seasons, with a few others from the previous 5 years thrown in for comparison.
First, a look at point guards:
Code:
Player %asstd asst% eFG%
Tre Jones 2019 43.2 24.0 45.6
Quinn Cook 2013 41.7 29.0 49.2
Tyus Jones 2015 37.8 27.5 48.9
Jeremy Roach 2021 30.6 17.8 48.8
Trevon Duval 2018 29.7 30.3 47.3
Tre Jones 2020 22.8 31.4 47.4
Because PGs have the ball in their hands a lot, by necessity fewer of their FGs are going to be assisted. Perhaps because of that, as well as the fact that PGs tend to focus more on distribution than on shooting, their eFG% tends to be lower, but they make up for that with a higher assist%.
Jeremy Roach is in the middle, as far as %assisted (perhaps a bit on the low side), and his eFG% is also in the middle (a bit on the high side, even). But what painfully stands out is his assist%, way too low to be compared to the others in this group. With this team, I think it would be best if he focused on distributing the ball. Alternatively, he could focus more on catching the ball where he can shoot it and not trying to create his own shot, because he's just not so good at that right now. But his hanging out in the "no man's land" could be a big reason why our offense tends to sputter.
Next, a quick look at players who just want to do it all by themselves:
Code:
Player %asstd asst% eFG%
Jayson Tatum 2017 38.2 12.4 50.7
Wendell Moore 2020 28.1 12.6 42.9
Austin Rivers 2012 19.5 13.0 50.5
Some players tend to handle the ball a lot, but mostly for their own purposes. These guys are high-end passers for shooting guards, but rarely catch-and-shoot. Austin Rivers in 2012 was perhaps the most extreme example of this kind of player, but as you can see, Wendell Moore followed this pattern last season. The thing is, from our small sample, to succeed at this you have to have an eFG% higher than 50%, and Wendell didn't make that cut. As many have noted this season, he really can't be effective in this role.
Next, a look at some catch-and-shoot players:
Code:
Player %asstd asst% eFG%
Andre Dawkins 2014 78.6 5.6 61.4
Alex O'Connell 2018 78.6 7.8 60.2
Gary Trent 2018 68.9 6.8 52.8
Alex O'Connell 2019 67.3 6.4 56.9
Joey Baker 2020 64.0 10.7 52.5
Wendell Moore 2021 55.6 9.8 39.1
Alex O'Connell 2020 55.0 9.9 47.6
Joey Baker 2021 42.9 3.7 34.0
I put Andre Dawkins in here as an extreme example, but for this type player, the more your shot is assisted, the better you shoot. For example, look at Alex O'Connell. When 79% of his baskets were assisted, his eFG% was outstanding. But each year he tried to do more, and by 2020 when he was only assisted on 55% of his baskets, his eFG% became unacceptable.
Wendell Moore seems to be learning this lesson. Early on in the season he was trying to be a DIY player and his shooting numbers were awful. His season numbers are still awful, but he's been catching and shooting enough to start pulling them up. Joey Baker has only made 7 field goals this season, so his numbers might be a little deceptive, but he needs to learn that same lesson. The more he shoots on a catch, the better he'll be.
Next, let's look at "regular" shooting guards:
Code:
Player %asstd asst% eFG%
Quinn Cook 2015 62.8 13.2 56.9
Matt Jones 2017 58.9 9.7 50.4
Cam Reddish 57.8 10.7 45.9
Cassius Stanley 2020 50.8 6.6 53.1
Seth Curry 2013 47.2 9.7 57.7
Frank Jackson 2017 45.9 12.6 56.4
DJ Steward 2021 44.7 12.0 50.0
Luke Kennard 2017 40.5 13.6 57.9
These guys pass a little bit, plus create shots for themselves (though not nearly so much as the DIY folks), and also catch-and-shoot. It's possible the guys at the top of this table belong more in the catch-and-shoot group and it's possible the guys at the bottom might be categorized in the do-it-yourself group. One fascinating thing to me is how different Quinn Cook's profile was in 2015, when he played shooting guard, then it was in 2013 as a point guard. Obviously his assist% dropped from 29% to 13%, but much more interesting, his %assisted number jumped from 42% to 63% and his eFG% made a corresponding jump from 49% to 57%. To me, that's very cool and illustrates how a shift in role might benefit some of our guards this season.
As far as DJ Steward is concerned, he appears to be attempting to create his own shot like Frank Jackson and Luke Kennard, but is only shooting as well as Matt Jones. Maybe next season he can be Kennard, but this season I suspect he needs to be more of a catch-and-shoot guy to optimize his efficiency. Either that or go full-on Austin Rivers and try to take advantage of his driving ability. Personally, I'd prefer to see the former.
Finally, a look at some "true" combo guards:
Code:
Player %asstd asst% eFG%
Grayson Allen 2018 58.0 21.2 53.6
Grayson Allen 2017 57.4 21.5 50.8
RJ Barrett 2019 45.1 23.5 50.6
Jordan Goldwire 2021 44.4 19.7 47.0
Jordan Goldwire 2020 36.2 14.4 55.9
These are players who find open shooters almost as well as PGs, but also look for their own. I was surprised that RJ Barrett fit in this category, but with that assist% it would be hard to put him anywhere else. This table also makes me think all of Grayson Allen's shenanigans made him underappreciated. He has a very high catch-and-shoot rate as well as a consistently high assist rate. Which seems to me to be pretty hard to accomplish.
As for Jordan Goldwire, the fact that he can be in the same conversation as those other guys is quite an achievement for someone was wasn't even ranked in the top 250 coming out of high school. His numbers also appear backwards, as he shot better with fewer assisted shots. Even after a deep dive into his numbers, I can't explain it.
Anyway, no great revelations here, but if Roach and Steward can fit more into roles than trying to do everything, and Moore and Baker focus on being pure catch-and-shoot guys, I suspect the overall offense will be better off.