Great Statistical and Scouting Analysis of Austin Rivers [Archive]

Kedsy

11-21-2011, 12:58 PM

http://www.grantland.com/blog/the-triangle/post/_/id/10296/the-problems-with-austin-rivers

It's by Sebastian Pruitti of Basketball Prospectus fame, so it's really good stuff.

Yeah, pretty good stuff, although my guess is Grantland only posted it because it suggests Austin's not as good as the hype. All the problems seem fixable, however. I'm not worried about Austin at all. If anybody can help him with his decision making, it's Coach K.

CDu

11-21-2011, 01:01 PM

The last paragraph sums it up:

If Rivers starts to make the correct passes and starts using his pull-up jumper, he will live up to all the hype. His turnover rate will drop. Defenses will be forced into more decisions, which will open up the floor to hesitation dribbles, crossovers and drives to the basket. All the tools are there for Rivers, he just needs to figure out how to make the decisions that maximize his considerable talents.

Rivers has shown a very mature ability to beat his man off the dribble. He hasn't yet shown the ability to consistently make the best decision when the help defense arrives. I'm guessing that's largely due to (a) not being used to such organized help defense at the college level and (b) not yet being used to having such good teammates.

He seems to be great at taking step one in the attack (beat man), but iffy at making step two (read/react to help defense or lack thereof). When/if he figures out what's available after his initial move, he'll be devastating. As is, he's already very good.

slower

11-21-2011, 01:02 PM

Thanks, but 4 games into a freshman season is way too early for an overall analysis, IMO, especially on a statistical basis.

I'd call them preliminary observations.

Well, I don't think that these are any observations that many of us haven't already made. But I think Austin will be fine. Much better than fine, actually.

And, as with anything you read in Grantland concerning Duke, just remember that Grantland is Simmons' baby. Simmons will take any and every shot at Duke that he can. And even stuff written by Shane Ryan, their in-house "Duke guy", will always tend to have negative things in it. That's just the way it is over there. And, of course, anything written by UNC homers like Jay Kang will take as many shots at Duke as possible.

CDu

11-21-2011, 01:04 PM

Thanks, but 4 games into a freshman season is way too early for an overall analysis, IMO, especially on a statistical basis.

I'd call them preliminary observations.

It's still an analysis. It's just not a final analysis. And it doesn't proclaim to be a final analysis. It's just an analysis of what he has been at the moment and what he needs to improve to be truly dominant. And I think it's a pretty fair assessment of what he has done so far.

loran16

11-21-2011, 01:05 PM

Thanks, but 4 games into a freshman season is way too early for an overall analysis, IMO, especially on a statistical basis.

I'd call them preliminary observations.

I love how you basically ignored the article in this comment. It's more like you just read the title and felt necessary to attack it without reading it. Bad form really. The article doesn't make any overall assessment of Rivers as a failure (or success) you know.

@Kedsy, while I dislike some of the CBB stuff at Grantland, they've shown a record in other sports (Baseball, Football) of getting statistical-analysis guys to write pieces. I suspect this wouldn't have been written if the NBA was still going on (this writer is normally on the NBA side of Prospectus), but I don't think the piece was published simply because it was about a Duke player underperforming thus far.

NSDukeFan

11-21-2011, 01:14 PM

The last paragraph sums it up:

Rivers has shown a very mature ability to beat his man off the dribble. He hasn't yet shown the ability to consistently make the best decision when the help defense arrives. I'm guessing that's largely due to (a) not being used to such organized help defense at the college level and (b) not yet being used to having such good teammates.

He seems to be great at taking step one in the attack (beat man), but iffy at making step two (read/react to help defense or lack thereof). When/if he figures out what's available after his initial move, he'll be devastating. As is, he's already very good.

I thought it was an interesting article, though I agree with roywhite that it is early to be making judgements statistically.

I also found the clips used as examples were not fantastic as the first clip showing that Austin forced a shot, but in reality he got to the middle of the lane and the help side defender was not a factor, but the defensive player made a nice play to strip Austin. I wouldn't have called that a mental mistake at all, just getting stripped by the defensive player.

In the next clip, the author notes poor decision-making as Austin gets to a great spot on the floor for a short lay-up, but just misses avoiding the defensive player and gets called for a charge. I find it tough to be too critical of him on this play, though I agree with the call.

I agree with the basic statements the author makes, as there has been a trend of Austin not always making the best decisions when driving, but these clips don't necessarily highlight this that well, IMO. I have to commend the author for using clips though as I do better enjoy analysis that includes pictures and videos to show what he is trying to say.

Thanks for sharing this loran16 and I agree with CDu that Austin is having some challenges adjusting to what he can and cannot do vs. the second defender at the college level, but he is already doing well and I expect will get better as he gains more experience.

jamesfrommaiden

11-21-2011, 01:16 PM

Are you serious? What does the young man have to do to impress this guy? I think Austin is doing a very good job so far. Yes there is plenty of room for improvement, but he doesn't have to be the alpha and omega for this team. Fortunately for Austin he is surrounded by a cast that includes veteran and young talented players as well. I think you also have to consider the immense pressure placed on him as the number one overall recruit and going to Duke. A gaurd which means he has a long line of some of the greatest college players ever at the position to follow. I have seen a confident, self assured, cocky if you like, with that great Duke swagger. He knows how good he is. He doesn't need, or have to prove to anyone that he is the real deal. He can get to the rim at will. If he doesn't get a basket, he gets fouled, or both. We will see more of the latter in time as he gets stronger to finish around the rim. I leave you with this. UNC fans and people of their ilk like to ask the question "who is going to gaurd Harrison Barnes?". I am asking those people who is going to check Austin Barnes? GO DUKE!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!

COYS

11-21-2011, 01:22 PM

This is a great, detailed article that demonstrates with stats and examples what many on the board have also noticed. Rivers needs to learn when to take the ball all the way to the rim and when to kick it out or pull up for a jump shot rather than a leaning, spinning layup through traffic. Rivers has not been a ball hog as many have feared. He's a willing passer. He's just not a willing passer on his drives, which is the biggest reason he hasn't been an efficient passer. I would also add to the article that his three point shooting has yet to get going, either. Making a whole bunch of three pointers is THE quickest cure to offensive inefficiency. Since he's streaky from three, I'd imagine that his three point percentage will never be particularly awe-inspiring, but hitting anywhere from 35-37% will help open up the floor for him.

Anyway, I know I'll be watching to see if/how much Austin's decision-making on drives improves. If he starts producing more positive plays than negative ones (assist, free throw attempt, or made pull up jumper) when he can't get all the way to the rim for a score, it will DRAMATICALLY improve the teams overall offensive efficiency, which is already solid (8th in the country according to KenPom, even though this season's stats are not yet completely isolated to the current season yet).

gwlaw99

11-21-2011, 01:25 PM

I am not worried about Rivers at all at this point. He has a lot to learn about the college game and how best he can best use his talent, but there has already been a lot of progress from China. There is no reason to think he won't continue to develop and mature. Good thing he picked such a great teacher.

Duvall

11-21-2011, 01:29 PM

Solid piece. Glad to see Grantland get some CBB writing from an actual adult, even if they had to turn to an NBA writer to do it.

roywhite

11-21-2011, 01:35 PM

I love how you basically ignored the article in this comment. It's more like you just read the title and felt necessary to attack it without reading it. Bad form really. The article doesn't make any overall assessment of Rivers as a failure (or success) you know.

@Kedsy, while I dislike some of the CBB stuff at Grantland, they've shown a record in other sports (Baseball, Football) of getting statistical-analysis guys to write pieces. I suspect this wouldn't have been written if the NBA was still going on (this writer is normally on the NBA side of Prospectus), but I don't think the piece was published simply because it was about a Duke player underperforming thus far.

I read the analysis and made my comment.
Your response seems hostile.

Biscuit King

11-21-2011, 01:47 PM

So, to summarize: Rivers has no difficulty beating the first man off the dribble, but he needs to improve his decision-making when help defense comes.

For a freshman who has played 4 games at the D-I level, this does not seem to be advanced analysis. It is certainly correct. It's just not very insightful. The efficiency stats are also to be taken with a grain of salt in such a small sample size.

juise

11-21-2011, 01:48 PM

Regarding the question of the analysis being silly because we're four games into the season, the author agrees:

Is this to say that Rivers is a bust and a terrible player? No. To make any sweeping generalizations about a player after four games would be silly.

CDu

11-21-2011, 02:13 PM

I read the analysis and made my comment.
Your response seems hostile.

In fairness, your comment did sort of sound like you hadn't read the article. Because the author in fact uses words like "yet" and "to make any sweeping generalizations about a player after four games would be silly."

It's an analysis of what Rivers has done so far (and I think a fairly accurate one), not a final judgment on him as a player.

CDu

11-21-2011, 02:14 PM

Regarding the question of the analysis being silly because we're four games into the season, the author agrees:

He does not call his analysis silly. He says making declarations about Rivers as a player would be silly. Those are very different statements. One can analyze without making sweeping declarations.

juise

11-21-2011, 02:27 PM

He does not call his analysis silly. He says making declarations about Rivers as a player would be silly. Those are very different statements. One can analyze without making sweeping declarations.

You're right. That was poorly worded by me. I guess I was more referring more to the "final analysis" phrasing from earlier in the thread.

Overall, I agree that (1) the sample size is too low and that (2) it doesn't take PPP stats to say demonstrate that Austin has some room for growth in decision making. (Likewise, we don't need a larger sample size to demonstrate that Mason needs work on free throw consistency.) However, I think that it is useful to take a snapshot of Austin's season at this point, noting his current PPP, and compare the number to his efficiency in February. I think the comparison will be favorable, even with increased competition.

tele

11-21-2011, 02:49 PM

"Great Statistical and Scouting Analysis of Austin Rivers", so which one is silly at this point of the season? If an observtion is accurate you can find a statistic to support it 100% of the time, if an observation is inaccurate, it usually takes more analysis to find one. And no, I haven't read the article.

CDu

11-21-2011, 02:58 PM

basket1544

11-21-2011, 09:40 PM

Well done. I like the fact that the author includes some of his high school work to the story also. I agree that Austin has a terrific first step and will become a great basketball player as soon as he learns how to pass to his open teammates. I think it is awesome that teams respect him so much this early as to double team him.
He'll learn; he's got the greatest teacher in the game.

FellowTraveler

11-21-2011, 10:53 PM

DukieTiger

11-22-2011, 12:23 AM

To my eye, in his first several games (exhibition and real) Rivers demonstrated a fantastic first step, but his lack of elite athleticism frequently allowed the defense to recover by the time he shot, leading to a lot of drives that began with him making a great move to create separation and ended with a contested shot anyway. In the last two games, he’s had a few great drives where he’s successfully built on the initial separation, including a very nice drive tonight in which he incorporated a jump-stop and a couple of quick changes of direction. His first step is a huge asset. When he develops the ability to capitalize on it by maintaining separation to get clean shots -- and by finding teammates left open by a shifting defense -- he’ll be a truly elite offensive player.

Lack of elite athleticism? :confused:

I think it more likely that the separation closes simply because he is not used to the closing speed of opposing defenses at this level. Either that, or the separation closes because a help defender comes into the picture- and no matter how athletic you are, it's hard to avoid help defenders at all times.

That said, I agree with your point about his need to learn how to capitalize on the separation he is able to create. I think defenders can close on him or catch up to him because he tends to hold the ball too long before making a decision. Thus, he drives into the teeth of the defense or gets stripped or takes contested shots. He definitely needs (and will!) to learn to capitalize on his quickness by making decisions quickly as well. When his decision-making catches up to his first step, he will pull up for a jumper and still have that space, he will find the open man when the space is still available, or he will get to the basket when there is room for that as well. It's his mental quickness, not physical quickness, that is lacking at this point.

Des Esseintes

11-22-2011, 02:49 AM

In fairness, your comment did sort of sound like you hadn't read the article. Because the author in fact uses words like "yet" and "to make any sweeping generalizations about a player after four games would be silly."

It's an analysis of what Rivers has done so far (and I think a fairly accurate one), not a final judgment on him as a player.

I don't mean to put words in roy's mouth, but the argument could be made that statistical analysis after four games--even with the caveat that it is not a "final judgment"--is illegitimate based on sample size issues. I remember a spirited discussion from a few days back between you and Kedsy (and others) about whether plus-minus tells us anything meaningful. You came down on the side that it probably does not, at least in college, because there are simply is not enough information. That was over the course of an entire season. The present Rivers sample size is so miniscule as to be nonexistent right now. It's an honest question: how are the situations different? This, I think, is the issue. It isn't that the Pruiti is leaping to final conclusions. Of course he isn't. But isn't it a leap to make even provisional conclusions based on statistics when the sample size is so pathetic?

Perhaps some people on the board recall what kenpom's rankings used to look like early in the season before he adopted preseason weighting last year. It took a very long time for the results to tell us anything, and for quite a while many of the results were actively misleading. It would have been insane to say, "Well, the data to this point says Old Dominion is the best team in the country. This is by no means final, but I think we can ..." No. The correct thing at the time would have been to say that the data was not yet telling us anything upon which we could rely, full stop. Eventually, more data accrued, the rankings stabilized, and the stats became an immensely powerful tool. That point was well past the fourth game of the season.

Grey Devil

11-22-2011, 03:14 AM

I truly appreciate the time and effort that Sebastian Pruiti put into his analysis of Austin's "tendencies."

My only reaction is that Coach K and staff (along with their own high tech video equipment) have probably already gone over this with Austin, maybe even multiple times, before this insightful article was even posted.

It will just take some time before Austin incorporates the suggested skill set into his repertoire of moves. It is, after all, a big step up for him (to be playing at this level). But I have no doubt that he will catch on and eventually be using his newly learned skill to help the team win more games. The more interesting question to me is how soon will we be seeing him "get it" in games?

For me the fun of being a Duke fan is not in all the victories, altough I certainly enjoy each and every one, but in watching Coach K and staff help each player maximize their abilities and blend them all into one "fist."

Go Duke!

Grey Devil

juise

11-22-2011, 03:28 AM

The present Rivers sample size is so miniscule as to be nonexistent right now. It's an honest question: how are the situations different? This, I think, is the issue. It isn't that the Pruiti is leaping to final conclusions. Of course he isn't. But isn't it a leap to make even provisional conclusions based on statistics when the sample size is so pathetic?

As I mentioned earlier in the thread, I don't think that it is a leap. I think the bigger question is whether the conclusion is worth discussing.

Let's say you were to run a Chi-squared hypothesis test between Ryan's free throw results (0's and 1's) and Mason's. The sample size is still small, but I think the test would tell you there's a statistically significant difference between the data sets and the conclusion would be that Ryan has been shooting FT's better than Mason. Are we shocked? No.

Similarly, Austin's PPP results after four games rank rather low nationally. This small data set may be lead us to the conclusion that Austin is not using possessions very efficiently. Does this match up with the observations we've made thus far? Yeah, I think so. So, I think the question is the advanced metric Pruiti quotes tells us anything new?

Fortunately, basketball players are not normal distribution functions. This snapshot is not necessarily predictive of future results. I'm not sure that more data will validate or invalidate Pruiti's analysis because Austin will be a different player next week than he was last week as he climbs the learning curve of playing D1 ball. The analysis of four games "is what it is." It's a progress report of a freshman season in progress. I'm excited to see what the next grading period brings.

Let's go, Duke!

CDu

11-22-2011, 11:04 AM

I don't mean to put words in roy's mouth, but the argument could be made that statistical analysis after four games--even with the caveat that it is not a "final judgment"--is illegitimate based on sample size issues. I remember a spirited discussion from a few days back between you and Kedsy (and others) about whether plus-minus tells us anything meaningful. You came down on the side that it probably does not, at least in college, because there are simply is not enough information. That was over the course of an entire season. The present Rivers sample size is so miniscule as to be nonexistent right now. It's an honest question: how are the situations different? This, I think, is the issue. It isn't that the Pruiti is leaping to final conclusions. Of course he isn't. But isn't it a leap to make even provisional conclusions based on statistics when the sample size is so pathetic?

The short answer is that a single-game +/- (or a small-sample +/-) is based on an N of 1 (or 3-4, if it's a 3-4 game +/-). PPP after 4 games for a high-volume player like Rivers is about 60. That's still a small sample size, but it's a much less small sample size.

I agree that it's silly to make relative comparisons (e.g., "he's ranked Xth in efficiency") after 4 games. But using those 60ish possessions worth of data to support a qualitative statement ("he hasn't played great yet, and he's struggled with Y types of plays") is completely reasonable. And that was the crux of Pruiti's article. He used the data to support the point that Rivers hasn't been terribly efficient, and his inefficiency so far has stemmed from a particular type of play (not great decisionmaking off the drive).

Highlander

11-22-2011, 12:37 PM

I enjoyed the article. After watching Austin last night vs. Tennessee, I have to say that the author has a point. He was 1-7 at one point IIRC, and almost all of his misses were on drives where he beat his defender on the first step, but then failed to finish. He got going once he hit those two big three's, then altered his approach to start using a teardrop/giant killer shot and began to have some success. So kudos to Austin for realizing the drive wasn't working and going to some other weapons. That shows some maturity right there.

Personally, I believe Austin has the unenviable position of following the most NBA ready Freshman to ever play at Duke. He is already starting, and once he establishes a rhythm, I have no doubt he will mean as much to Duke as HB does to UNC.

burnspbesq

11-22-2011, 12:45 PM

jjasper0729

11-22-2011, 12:48 PM

Looks to me like he spent a lot of time hanging out around Reggie Miller. That's who his shot reminds me of.

My question about Austin is simple: how did a coach's kid end up with such an ugly, fundamentally unsound stroke? :confused:

As long as his elbow is sticking out at that crazy angle, he's going to be a streaky jump-shooter at best.

roywhite

11-22-2011, 12:50 PM

My question about Austin is simple: how did a coach's kid end up with such an ugly, fundamentally unsound stroke? :confused:

As long as his elbow is sticking out at that crazy angle, he's going to be a streaky jump-shooter at best.

Ask Chris Collins. His was hardly textbook, either.

My theory FWIW....a young Collins or Austin is probably the kid on the team of 8-year olds or 10-year olds who has the best chance of hitting long jump shots, so he gets the "green light" while the other kids don't.
Still, at that young age, it's a bit of a heave, and the form takes second place to getting the ball somewhere near the basket.
Then the bad habit sticks.

basket1544

11-22-2011, 07:50 PM

Des Esseintes

11-22-2011, 10:48 PM

The short answer is that a single-game +/- (or a small-sample +/-) is based on an N of 1 (or 3-4, if it's a 3-4 game +/-). PPP after 4 games for a high-volume player like Rivers is about 60. That's still a small sample size, but it's a much less small sample size.

I agree that it's silly to make relative comparisons (e.g., "he's ranked Xth in efficiency") after 4 games. But using those 60ish possessions worth of data to support a qualitative statement ("he hasn't played great yet, and he's struggled with Y types of plays") is completely reasonable. And that was the crux of Pruiti's article. He used the data to support the point that Rivers hasn't been terribly efficient, and his inefficiency so far has stemmed from a particular type of play (not great decisionmaking off the drive).

I hear you. But the 58 possessions Pruiti is writing about truly aren't much. 58 possessions comprise a bit more than three-quarters of your typical Duke game. Are you prepared to draw conclusions, even provisional conclusions, about this year's team based on something less than a single game? I am not. Take turnovers, for which Pruiti calls out Rivers. Over the first four games, Austin has turned it over too much. Fair enough. This is a fact. And maybe Austin does have a turnover problem. But you wouldn't call a team "turnover-prone" based on the evidence of most of one game. I wouldn't, at least.

(For what it's worth, which is nothing, Austin's been better with turnovers the two games since the article. Has he developed into a better player? Possibly, probably, in some ways. But maybe it was just a small sample size, and we are getting a better idea of his ability to control the basketball. I won't venture a numbers-based guess, and I don't think Pruiti should either.)

Nor do I think Pruiti is being as forthright as he's being given credit for here. He talks about the 44.8% of Austin's plays that are pick-and-roll and the 29.3% that are isos. "29.3%" sounds impressive--until you realize he's talking about 17 plays. 17 plays is NOTHING. Taking the percentage out to the tenth place is the definition of meretriciousness, and exists only to paper over the non-robustness of the data.

Grey Devil

11-23-2011, 02:14 AM

It was nice to see Austin drive the lane against Michigan tonight and pass the ball out to Seth, who promptly swished a three. Maybe he's been reading this thread... ;).

Hopefully we'll see more of that against Kansas.

Grey Devil

CDu

11-23-2011, 10:36 AM

I hear you. But the 58 possessions Pruiti is writing about truly aren't much. 58 possessions comprise a bit more than three-quarters of your typical Duke game. Are you prepared to draw conclusions, even provisional conclusions, about this year's team based on something less than a single game? I am not. Take turnovers, for which Pruiti calls out Rivers. Over the first four games, Austin has turned it over too much. Fair enough. This is a fact. And maybe Austin does have a turnover problem. But you wouldn't call a team "turnover-prone" based on the evidence of most of one game. I wouldn't, at least.

I wouldn't call a team turnover-prone based on one game. But I don't believe that Pruiti is doing that, either though. I'd say a team had too many turnovers in that game, and will need to do better in that regard moving forward. There's a difference between saying "has struggled with turnovers so far" and "is a turnover prone team/player." One is an analysis of what has happened to this point, the other is a conclusion about a quality of the player/team.

And that's what Pruiti did. He analyzed what Rivers has done to this point, and suggested where Rivers will need to do better to be as good as advertised. He hypothesizes as to why that might be the case, and you can quibble with that. But when he does so, he states that it's just his opinion. And in a few cases, he suggests counterarguments that are made.

You can quibble about the jump to the decimal place if you'd like, but that's nibbling at the margins. He could have been more general in his analysis (rounding to 30% instead of 29.3%, for example) and still made his point. And he could have done so without the national ranking in PPP. But I think that's focusing on the trees and missing the forest.