Do high turnover % point guards create better offense?

[penbeast0]

I read this in the Harden/Brunson thread and it seems questionable to me. The assumption is that the riskier pass is the one that leads to higher team offense rather than just being lower level decision making.

So, let's make a list (I don't have time to search it right now but if anyone wants to please go for it, if not I will probably do it tomorrow), of the 10 highest TO% and lowest TO% PGs. Let's limit it to players who have made at least 2 all-star teams to weed out the chaff from the grain. Then we can look at their team ORTG numbers relative to their leagues and see if this statement is true.

[VanWest82]

Nash is the poster boy for this argument and I do believe that in his case part of the reason why his offenses were so successful is because he took risks to get guys great looks and the cumulative effect of those risky plays was that defenses freaked out more often and messed up in coverage when Nash had the ball.

In a larger context, I'd be skeptical that high TOV% for high AST% play makers was in any way correlated to better offenses than low TOV% guys. Maybe I'm wrong there but it's definitely not the way I'd handicap it.

[capfan33]

I haven't read the thread but I would guess, being reasonable, that the intent behind the argument is that turnovers aren't absolutely bad in the way that most people think, but instead are a necessary byproduct of being an aggressive passer. The idea is that in order to create high value shots, put pressure on the defense and create playoff resiliency, you need to throw borderline passes that can more easily be picked off compared to your garden variety, lets say, pocket pass, but if successful are almost a guaranteed 2 points.

And philosophically, especially as the game evolves, that's usually the correct approach. I think the approach you're advocating for could be improved by only looking at guards who cross a certain assist% threshold to get rid of black hole guards who turn the ball over a lot but don't actually pass much, and also look at BBall references section on turnovers caused by bad passes vs mishandling the ball.

[tsherkin]

TOV% can't be taken in a vacuum.

What's their AST% like? What's their shooting volume like? Where are the turnovers COMING from? Are they ball-handling turnovers aggressively pressing the defense with dribble penetration, or are they passing errors? Some context is required.

Magic and Stockton are, respectively, 40.9/19.4 and 50.2/20.8 AST%/TOV% players, right? That's objectively high TOV%, but when you consider how many assists they were producing, it matters less. Magic rocked a 2.89:1 AST:TOV ratio and Stockton 3.72:1. Magic, of course, shot more and was a more aggressive post player and what-not, so a lot of those turnovers had less to do with his passing than his scoring, particularly once LA grew accustomed to his passing style.

Food for thought. I don't think there is necessarily an importance level to be placed on TOV%. If Chris Paul were 6'5 instead of 5'11, his low-turnover style would still be quite useful, coupled to his scoring efficiency and so forth, you know?

[OhayoKD]

TOV% can't be taken in a vacuum.

What's their AST% like? What's their shooting volume like? Where are the turnovers COMING from? Are they ball-handling turnovers aggressively pressing the defense with dribble penetration, or are they passing errors? Some context is required.

Magic and Stockton are, respectively, 40.9/19.4 and 50.2/20.8 AST%/TOV% players, right? That's objectively high TOV%, but when you consider how many assists they were producing, it matters less. Magic rocked a 2.89:1 AST:TOV ratio and Stockton 3.72:1. Magic, of course, shot more and was a more aggressive post player and what-not, so a lot of those turnovers had less to do with his passing than his scoring, particularly once LA grew accustomed to his passing style.

Food for thought. I don't think there is necessarily an importance level to be placed on TOV%. If Chris Paul were 6'5 instead of 5'11, his low-turnover style would still be quite useful, coupled to his scoring efficiency and so forth, you know?

The biggest piece of context missing(largely because it isn't typically counted for box-score purposes), is how much a player is handling the ball. A 10 percent tov-rate from someone who is bringing up the ball nearly every possession is much better turnover economy than a 10 percent tov rate form someone who is splitting with a teammate if all else is equal.

I read this in the Harden/Brunson thread and it seems questionable to me. The assumption is that the riskier pass is the one that leads to higher team offense rather than just being lower level decision making.

So, let's make a list (I don't have time to search it right now but if anyone wants to please go for it, if not I will probably do it tomorrow), of the 10 highest TO% and lowest TO% PGs. Let's limit it to players who have made at least 2 all-star teams to weed out the chaff from the grain. Then we can look at their team ORTG numbers relative to their leagues and see if this statement is true.

Not a terrible idea but we'll have to pick a consistent time-frame(5-years?). Also not sure there's a point restricting it to PG's. Also might be nice to add offensive with/without for context

[tsherkin]

The biggest piece of context missing(largely because it isn't typically counted for box-score purposes), is how much a player is handling the ball.

Yeah and exactly what they're doing. Pounding the rock at the top of the circle doesn't mean a lot, but if you're attacking in the PnR and with dribble isolations and what-not, you expect a higher TOV%. Turnovers happen. Particularly for guys who aren't lofting 20+ FGA/g, if you're doing stuff with the ball, you're going to often generate a meaty number of turnovers. That's why guys like MJ and Dirk, and Kobe, were remarkable in those regards.

[OhayoKD]

The biggest piece of context missing(largely because it isn't typically counted for box-score purposes), is how much a player is handling the ball.

Yeah and exactly what they're doing. Pounding the rock at the top of the circle doesn't mean a lot, but if you're attacking in the PnR and with dribble isolations and what-not, you expect a higher TOV%. Turnovers happen. Particularly for guys who aren't lofting 20+ FGA/g, if you're doing stuff with the ball, you're going to often generate a meaty number of turnovers. That's why guys like MJ and Dirk, and Kobe, were remarkable in those regards.

Well the three players you listed handled the ball less relative to their box-usage. Not sure those are the examples I'd draw up of players being undersold in terms of "turnover economy". The first example that comes to mind is Nash who would dribble out entire defenses before making passes. Magic probably second.

[tsherkin]

Well the three players you listed handled the ball less relative to their box-usage. Not sure those are the examples I'd draw up of players being undersold in terms of "turnover economy". The first example that comes to mind is Nash who would dribble out entire defenses before making passes. Magic probably second.

Right, they posted a bunch and lofted tons of jumpers, so possessions that ended on those jumpers were FGA and not TOV, even if they missed and the other team nabbed a defensive rebound, I agree. Jordan, of course, rocked a "high" TOV of 11.9% during his 89 season averaging 8 apg, and apart from his rookie season was always a super-low TOV guy, so he's maybe not the hottest example because he was ridiculous even when he WAS handling the ball a lot.

EDIT: But to your point, he also made quick decisions and then burst like a lightning bolt to the basket, so his time of dribble was lower than Magic and Nash.

[eminence]

TO% as presented on BBref as an individual stat I don't see much value, but turnovers are obviously great to track. Having the ball is valuable, duh.

There are certainly strong playmakers who had moderately high individual turnovers, but successfully led team offenses, and in some cases despite their individual turnovers helped to keep team turnovers quite low (Nash in his first 3 years in Phoenix). Magic of course led strong offenses, but the teams were generally fairly turnover prone.

Steph probably the closest I could think of to a star guard who's legitimately turnover prone but still has led strong offenses. But he plays a very non-traditional role.

[OhayoKD]

Well the three players you listed handled the ball less relative to their box-usage. Not sure those are the examples I'd draw up of players being undersold in terms of "turnover economy". The first example that comes to mind is Nash who would dribble out entire defenses before making passes. Magic probably second.

Right, they posted a bunch and lofted tons of jumpers, so possessions that ended on those jumpers were FGA and not TOV, even if they missed and the other team nabbed a defensive rebound, I agree.

It's partially that. But there are other trade-offs, from a team-level, as well. One of the big hurdles handling the ball alot is defenses will put more focus on you. This was especially true in the 90's with the rules rendering off-ball defense pretty limited, though it's still true today(See: KD in OKC vs GSW). If you're doing less of that, then someone else has to do more of it.

Also, "pounding the rock" itself can be valuable depending on the context. Controlling tempo is itself a skill. I think a big reason the Warriors underperformed in 2015 vs the cavs and 2018/2019 vs the Rockets is the opposing "point" slowing the game to a crawl(notably they gained separation when the game sped up again). The best playmakers also sometimes use time pounding the rock to organizers their receivers so to speak and create optimal routes.

Jordan, of course, rocked a "high" TOV of 11.9% during his 89 season averaging 8 apg, and apart from his rookie season was always a super-low TOV guy, so he's maybe not the hottest example because he was ridiculous even when he WAS handling the ball a lot.

Definitely a low turnover-rate, but 12% in a peak year is an outlier for someone who was rocking sub-10 percent for basically their whole career and sub 9 percent for the entirety of their prime. Jordan still looks like one of the best in terms of turnover economy, but he's no longer the least turnover prone.
 

EDIT: But to your point, he also made quick decisions and then burst like a lightning bolt to the basket, so his time of dribble was lower than Magic and Nash.

He made quick decisions in terms of shooting, but I don't know the same is necessarily true with passing reads:
https://forums.realgm.com/boards/viewtopic.php?f=64&t=2247097&hilit=lamelo&start=140
I don't think Jordan was, on a whole, a quicker decision-maker than those two. He was quicker to shoot but that's not the same thing(and carried trade-offs as I'm sure you're aware)

Steph probably the closest I could think of to a star guard who's legitimately turnover prone but still has led strong offenses. But he plays a very non-traditional role.

Steph is also the first example of someone who can create at volume without the ball(though the amount he is creating is probably overstated given the shot-quality differentials).

He can, to a degree, avoid that trade-off to a degree everyone else can't.

[LukaTheGOAT]

I haven't read the thread but I would guess, being reasonable, that the intent behind the argument is that turnovers aren't absolutely bad in the way that most people think, but instead are a necessary byproduct of being an aggressive passer. The idea is that in order to create high value shots, put pressure on the defense and create playoff resiliency, you need to throw borderline passes that can more easily be picked off compared to your garden variety, lets say, pocket pass, but if successful are almost a guaranteed 2 points.

And philosophically, especially as the game evolves, that's usually the correct approach. I think the approach you're advocating for could be improved by only looking at guards who cross a certain assist% threshold to get rid of black hole guards who turn the ball over a lot but don't actually pass much, and also look at BBall references section on turnovers caused by bad passes vs mishandling the ball.

Being that I'm the poster that made the post, this is exactly what I meant.

[falcolombardi]

Correlation is not causation

[Owly]

I read this in the Harden/Brunson thread and it seems questionable to me. The assumption is that the riskier pass is the one that leads to higher team offense rather than just being lower level decision making.

So, let's make a list (I don't have time to search it right now but if anyone wants to please go for it, if not I will probably do it tomorrow), of the 10 highest TO% and lowest TO% PGs. Let's limit it to players who have made at least 2 all-star teams to weed out the chaff from the grain. Then we can look at their team ORTG numbers relative to their leagues and see if this statement is true.

(For context I haven't followed the other thread, so this won't be about the specifics going on there.)

Wouldn't the All-Star qualification bias things in favor of high tunover guys, at least if the assumption was meant to hold more generally?

They would be reckoned to have reached a notional threshold despite presumably being known to have had those high turnovers (baked in), perhaps because they're doing more other stuff, not necessarily high leverage passing. Then again perhaps this problem is always there to some degree ... if someone is playing a guy they presumably rate him somewhat (at least above the immediate alternatives) and will have some knowledge of tendencies and expected outputs (less so in younger players where there's less sample, perhaps greater change).


Overall I'd guess there's some (I'd guess slight) correlation (between value of average assist and an appropriate turnover measure) in that holding all else equal a player will fall on a continuum that if sliced up will put you in a bucket that either does or doesn't tend to throw high-risk, high-reward passes when the opportunity to do so comes up. That said I'd imagine it's pretty weak because there's a lot of other ways to generate turnovers. So I (really) wouldn't think one could leap from high turnovers to this guys is throwing high value passes and leading to a higher team offense.


Also as noted by others there's the question of how to best measure turnovers. TOV% will, IIRC, will calibrate towards turnovers per shot attempt but not assists so will favor "shoot-first" guards.

[penbeast0]

Ok, I won't post any numbers as there are a lot of questions about what constitute high turnovers. What I'm basically looking for is whether a high assist/low turnover primary playmaker (PG or point forward or even Jokic) tends to create better or worse offense than a high assist/high turnover primary playmaker. We might have to factor out shot attempts somehow, not sure.

But it would seem logical that a guy who produces a lot of assists with relatively low turnovers (Chris Paul is the poster boy for this) would create better offenses than a guy who produces a similar number of assists with relatively high turnovers UNLESS the shots that are generally created are generally better across that type of player or it is an unusual effect only found with one or two point guards. Does a James Harden type produce better than a Chris Paul type (and does it also apply to less players like a Tyus Jones). So how do we set up the problem to eliminate variables that will skew this away from that question.

[Texas Chuck]

Being a high turnover PG can be very very misleading. Take Nash and Luka. Both seen as high turnover PG's but when you look at the little Mavericks during their tenures these are two of the best ball security teams in the league. And why is that? Because Nash and Luka two of the all-time great decision makers and guys you 100% want the ball in their hands as much as possible. Now Nash was lower usage and had the benefit of playing with ball-security legend Dirk, but the principle holds. The Suns were also a good ball security team.


So its not as cut and dried a question as it appears at first glance.

[DSMok1]

Correlation is not causation

This.

Many things can lead to high turnover percentage. Most of them are not good.

However, one of the things that can lead to high turnover percentage is attempting many high leverage passes that also benefit the team substantially when they succeed.

So a player can have a high turnover percentage and still be a tremendous offensive creator.

But all things being equal, a high turnover percentage is not good.

[WintaSoldier1]

I haven't read the thread but I would guess, being reasonable, that the intent behind the argument is that turnovers aren't absolutely bad in the way that most people think, but instead are a necessary byproduct of being an aggressive passer. The idea is that in order to create high value shots, put pressure on the defense and create playoff resiliency, you need to throw borderline passes that can more easily be picked off compared to your garden variety, lets say, pocket pass, but if successful are almost a guaranteed 2 points.

And philosophically, especially as the game evolves, that's usually the correct approach. I think the approach you're advocating for could be improved by only looking at guards who cross a certain assist% threshold to get rid of black hole guards who turn the ball over a lot but don't actually pass much, and also look at BBall references section on turnovers caused by bad passes vs mishandling the ball.

Fantastic Post.

[Lost92Bricks]

I came in here to say Chris Paul was the best of his era at running high ranked offenses.

People love using the low turnovers against him on here. Out of "theory" alone.

No one wants to actually look at the result of his teams.

Multiple Clipper teams were #1 offenses with terrible FT shooting dragging their numbers down.

[Doctor MJ]

I read this in the Harden/Brunson thread and it seems questionable to me. The assumption is that the riskier pass is the one that leads to higher team offense rather than just being lower level decision making.

So, let's make a list (I don't have time to search it right now but if anyone wants to please go for it, if not I will probably do it tomorrow), of the 10 highest TO% and lowest TO% PGs. Let's limit it to players who have made at least 2 all-star teams to weed out the chaff from the grain. Then we can look at their team ORTG numbers relative to their leagues and see if this statement is true.

List of guys listed as "point guards" with at least 2 all-star appearances by career RS TO%.

Top 10 guys on this:

1. Kemba Walker 10.5
2. Shai Gilgeous Alexander 10.8
3. Kyrie Irving 11.4
4. Gus Williams 11.6
5. Damian Lillard 11.7
6. Sidney Moncrief 12.2
7. Derrick Rose 12.7
8. Isaiah Thomas 12.8
9. Gary Payton 12.8
10. Terrell Brandon 12.9

Bottom 10:

1. Norm Van Lier 226.9
2. Rajon Rondo 21.7
3. Charlie Scott 21.2
4. John Stockton 20.8
5. Ben Simmons 20.3
6. Tiny Archibald 20.2
7. Steve Nash 19.5
8. Magic Johnson 19.2
9. Jason Kidd 18.9
10. Maurice Cheeks 17.5

[Doctor MJ]

I haven't read the thread but I would guess, being reasonable, that the intent behind the argument is that turnovers aren't absolutely bad in the way that most people think, but instead are a necessary byproduct of being an aggressive passer. The idea is that in order to create high value shots, put pressure on the defense and create playoff resiliency, you need to throw borderline passes that can more easily be picked off compared to your garden variety, lets say, pocket pass, but if successful are almost a guaranteed 2 points.

And philosophically, especially as the game evolves, that's usually the correct approach. I think the approach you're advocating for could be improved by only looking at guards who cross a certain assist% threshold to get rid of black hole guards who turn the ball over a lot but don't actually pass much, and also look at BBall references section on turnovers caused by bad passes vs mishandling the ball.

So I'll piggy back on this along with the basic data I presented here:

The goal is not to optimize TO%, because that can be achieved by hurting the team's offense - pass the ball to teammates who aren't in scoring position, dodge whenever they try to pass you the ball. Thus, the best point guards are always looking to find some balance between risk and reward, and they do so influenced by their strengths and weaknesses, as well as those of their teammates.

So take two elite point guards, the one with the lower TO% is not necessarily going to be the better of the two, and of course neither is the opposite true.

But with all that said, I think it putting a particular perspective here makes sense:

When a point guard chooses to make a pass that will either a) lead to a transition bucket or b) a turnover, the a:b ratio determines whether this is the right move in general or not. And what would we call a player who excels in this sort of decision making? Well, a great passer, no?

So then, if great passers can do this with a higher ratio...then they should do it more than other guys, and in doing so, they'll end up committing more turnovers than someone who was rightfully more cautious because he was less skilled.

That's no proof of the trend we should expect, but it doesn't show how a counterintuitive correlation between TO% and better offense could come into being.