Do high turnover % point guards create better offense?

[rrravenred]

Some great, thoughtful discussion. Another thing I'd add is that the impact of turnovers will also be mediated by how good the transition defence is. How much a turnover hurts a team (as opposed to a missed shot or 24 violation) isn't going to be a team-insensitive constant.

[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.

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.

Good points, and I think to sum up the counterintuitive part, really great passers can see opportunities that a lot of players wouldn't, much less have the ability to attempt, and these types of passes usually lead to more turnovers. So by labeling turnovers as bad, you're essentially downgrading said player for having better vision/intuition/physical passing ability compared to the average passer.

[penbeast0]

And that's where I'm just not sure. A high opportunity, high risk pass might look better when it works and might be the right move in some cases but the question is how often it works and how much of an efficiency improvement it generates since it is creating turnovers as well. Intuitively, I would guess that high turnovers are NOT a good thing. That doesn't mean a player with 5 turnovers per 48 minutes is more turnover prone than one with 4 as it depends on how much he is handling the ball. So someone like Magic, Nash, or Stockton who has very high assist numbers but also high turnover numbers because they have the ball a lot might not be generating more turnovers in a team sense. Nor are they necessarily making more of the high risk, high reward passes as a percentage of passes, it just might appear that way because they are making so many of the team's passes.

[rrravenred]

And that's where I'm just not sure. A high opportunity, high risk pass might look better when it works and might be the right move in some cases but the question is how often it works and how much of an efficiency improvement it generates since it is creating turnovers as well. Intuitively, I would guess that high turnovers are NOT a good thing. That doesn't mean a player with 5 turnovers per 48 minutes is more turnover prone than one with 4 as it depends on how much he is handling the ball. So someone like Magic, Nash, or Stockton who has very high assist numbers but also high turnover numbers because they have the ball a lot might not be generating more turnovers in a team sense. Nor are they necessarily making more of the high risk, high reward passes as a percentage of passes, it just might appear that way because they are making so many of the team's passes.

There's probably also a threat perception aspect to it (the offensive version of hearing footsteps) which can inhibit defensive rotations and drive bad decision defensive making. i.e. You don't have to constantly throw high-risk/reward passes, you just have to make the defense THINK you can...

[lessthanjake]

It’s plausible to me that there could at least be a correlation between higher turnover % PGs and offensive efficiency, because high turnover % may correlate with taking more risks and therefore correlate with creating more great looks. The offensive efficiency gains from the better looks could in many cases outweigh the offensive inefficiency from higher turnovers. That said, high turnover % also will correlate with worse defensive ratings. So I can imagine a scenario where higher turnover % correlates with higher offensive efficiency but does not correlate with higher net rating.

[Jaivl]

I'd wager that the graph with "TOV%" as the x and "team offense" as the y would resemble a flipped フ

[Ancalagon]

It’s plausible to me that there could at least be a correlation between higher turnover % PGs and offensive efficiency, because high turnover % may correlate with taking more risks and therefore correlate with creating more great looks. The offensive efficiency gains from the better looks could in many cases outweigh the offensive inefficiency from higher turnovers. That said, high turnover % also will correlate with worse defensive ratings. So I can imagine a scenario where higher turnover % correlates with higher offensive efficiency but does not correlate with higher net rating.

Really important point that I don’t think is being acknowledged enough so far on this thread.

The net effect of a turnover to an OFFENSE is no different than a missed shot - a possession ended without points. But turnovers result in the opponent scoring much more frequently than missed shots.

So the negative impact of turnovers is not to the offensive rating but to the defensive rating.

Edit: quick view of the highest total turnover seasons by players vis-a-vis Net Rating (this was easier to pull than TO %).

James Harden 2016-17
Houston 55-27, 5.84 SRS, +5.47 Net Rating

Russell Westbrook 2016-17
OKC 47-35, 1.14 SRS, +0.8 Net Rating

George McGinnis 1974-75
INA 45-39, 1.14 SRS, +1.0 Net Rating

A couple other notables by HOF players:

Barkley in 85-86 on a 55 win team, Lebron in 17-18 on a 50 win team, AI in 04-05 on a 43 win team, Isiah in 86-87 on a 52 win team, Kidd in 95-96 on a 26 win team.

So at least at the upper margins, the teams are usually still pretty good - which is likely why the high turnovers are being tolerated.

[Doctor MJ]

It’s plausible to me that there could at least be a correlation between higher turnover % PGs and offensive efficiency, because high turnover % may correlate with taking more risks and therefore correlate with creating more great looks. The offensive efficiency gains from the better looks could in many cases outweigh the offensive inefficiency from higher turnovers. That said, high turnover % also will correlate with worse defensive ratings. So I can imagine a scenario where higher turnover % correlates with higher offensive efficiency but does not correlate with higher net rating.

Really important point that I don’t think is being acknowledged enough so far on this thread.

The net effect of a turnover to an OFFENSE is no different than a missed shot - a possession ended without points. But turnovers result in the opponent scoring much more frequently than missed shots.

So the negative impact of turnovers is not to the offensive rating but to the defensive rating.

Edit: quick view of the highest total turnover seasons by players vis-a-vis Net Rating (this was easier to pull than TO %).

James Harden 2016-17
Houston 55-27, 5.84 SRS, +5.47 Net Rating

Russell Westbrook 2016-17
OKC 47-35, 1.14 SRS, +0.8 Net Rating

George McGinnis 1974-75
INA 45-39, 1.14 SRS, +1.0 Net Rating

A couple other notables by HOF players:

Barkley in 85-86 on a 55 win team, Lebron in 17-18 on a 50 win team, AI in 04-05 on a 43 win team, Isiah in 86-87 on a 52 win team, Kidd in 95-96 on a 26 win team.

So at least at the upper margins, the teams are usually still pretty good - which is likely why the high turnovers are being tolerated.

This is a great point but makes me want to go to one more level of detail:

A turnover that results in a steal will on average have considerable negative impact to your defense.

A turnover from the ball going out of bounds will not.

Would be good to see data clarifying how many of a player’s turnovers result in the former vs the latter.


Sent from my iPhone using Tapatalk

[Krodis]

Even more granularly, I would suspect turnovers in the paint are much less likely to result in easy points the other way than turnovers on the perimeter.

Sent from my SM-N981U using Tapatalk

[trex_8063]

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.

Some off-cuff-statements (haven't read other replies yet).....

As worded [particularly in title], short answer "no". That is: there's not a cause-and-effect, where higher turnover rates = better offense. This, I think, is intuitive and obvious.

But PG's with relatively high(ish) turnover rates nonetheless lead really good offenses, though the reasons can be varied:
*maybe their indvidual scoring acumen is elite
**maybe "luck" in teammates
***and then there's the quetion which you allude to re: risky passes (which has sub-questions: a) are all risky passes "better" [higher reward] passes? b) how often is a PG attempting them? c) How GOOD at completing them is he?

Are risky passes "better" passes? Not necessarily. Some passes simply are dumb and high risk, without much of a bigger upside.
However, there ARE a lot of other passes which truly are High Risk/High Reward.
How often a PG is going for these is variable, and wrt to "c" above: how often he can successfully complete them is also pretty variable......and might be the biggest variable in determining if they're worth it.

High risk/reward passes with someone like Nikola Jokic are obviously worth it, because his placement is so precise (SO precise, in fact, that they're hardly high risk......they're almost ONLY high reward).
Someone like Russell Westbrook, otoh......I'm not sure.


As to your suggestion of a correlation statistical model: I'm mildly intrigued, and might try to tackle that if I find the time.

NOTE: However, I think using TOV% alone for the purposes of any such model is an awful idea. TOV% only factors shots and turnovers, NOTHING else. It will as consequence tell you that PG's who err heavily toward pass-first mentalities are TERRIBLE in terms of ball-control (John Stockton looks awful according to TOV%, for example). Was he actually poor in terms of turnover economy? NO, of course not. That's the pitfall of TOV%, though; and consequently it's only useful in comparing players of similar type and play-style (and era, perhaps).
Were I to do such a study, I'd probably go with my own mTOV%, fwiw.

[capfan33]

And that's where I'm just not sure. A high opportunity, high risk pass might look better when it works and might be the right move in some cases but the question is how often it works and how much of an efficiency improvement it generates since it is creating turnovers as well. Intuitively, I would guess that high turnovers are NOT a good thing. That doesn't mean a player with 5 turnovers per 48 minutes is more turnover prone than one with 4 as it depends on how much he is handling the ball. So someone like Magic, Nash, or Stockton who has very high assist numbers but also high turnover numbers because they have the ball a lot might not be generating more turnovers in a team sense. Nor are they necessarily making more of the high risk, high reward passes as a percentage of passes, it just might appear that way because they are making so many of the team's passes.

Doing granular studies on this would be interesting albeit hard. But my interpretation of this is more philosophical and game theory oriented. Essentially as any competitive endeavor becomes more "solved', evolved and matured, which the NBA definitely is at this point, gaining separation from the competition becomes exceedingly difficult. This is doubly true in the playoffs.

As such, in order to gain a meaningful competitive advantage, you really have to push the limits of the sport, and so attempting these high risk passes is a necessity if you're serious about beating good opponents.

Now the caveat in the case of a sport such as basketball is that, there are other ways to eek out a meaningful competitive advantage, usually through overwhelming size + athleticism, or if you're lucky, an exceptionally good system and players to fit that system. But absent of those things, and even with those things, playing on a "knifes edge" so to speak has a lot of benefits in the long run even if it might cause some hiccups here and there. And moreover, I would say pushing limits is a bit more practical for most players compared to growing a few inches taller or gaining a few inches on their vertical. That's why high leverage passing is beneficial, not just because it can lead to an almost guaranteed two points.

And to the point on Magic and Nash, you're right that turnover numbers don't look as bad for a lot of these players when you consider the proportion of their teams possessions their eating up. But on your second point, as me and Doc went over earlier, Magic and Nash may not look like their making as many high risk passes as they actually are because they're simply better at converting them and make it look routine. Also, I'm not even sure guys like Magic and Nash approach the level of heliocentric ball-dominance that today's stars like Trae and Luka achieve, so even that may not really be as relevant a factor as you might think.

[penbeast0]

The trouble with high risk, high reward comes when the risk outweighs the reward. As a former HS coach, I may prefer the consistent safe option because at my level it works well. I see your point when you are outgunned or outmanned, you need to change the equation and high risk, high reward may do so. But if I am a successful team (contender), I would generally be the team that outguns or outmans the competition and my play style would probably be more conservative.

Again, philosophic difference. If I have a player that makes and succeeds with those passes to the point where they aren't really high risk anymore (Magic, Nash, Stockton, etc.) then that becomes a more conservative, less high risk option. Like always, systems are great but have to be tailored to fit the personnel you are building around.

[Doctor MJ]

The trouble with high risk, high reward comes when the risk outweighs the reward. As a former HS coach, I may prefer the consistent safe option because at my level it works well. I see your point when you are outgunned or outmanned, you need to change the equation and high risk, high reward may do so. But if I am a successful team (contender), I would generally be the team that outguns or outmans the competition and my play style would probably be more conservative.

Again, philosophic difference. If I have a player that makes and succeeds with those passes to the point where they aren't really high risk anymore (Magic, Nash, Stockton, etc.) then that becomes a more conservative, less high risk option. Like always, systems are great but have to be tailored to fit the personnel you are building around.

Great discussion here.

Something that's always stuck in my mind is that while in the NBA defense has pretty much always been dominated by shotblockers rather than thieves:

a) Historically this was not generally the case until the arrival of true big men.
b) It really didn't seem to be the case when I played pre-high school.
c) It hasn't generally been true in the WNBA.
d) It's not how it is in most video games.

The last one might seem trite, but to me it just hammers in that there are effectively "settings" in basketball like there are in video game basketball. In your league, how easy is it for the talent to block shots compared to snatch steals? The answer goes along way to indicating which is more valuable as a skillset.

Guys like Magic, Nash & Jokic are able to do things that if most tried to emulate, it would be for the worse. But if you're good enough at it, well, it's magical.

[Doctor MJ]

So I wanted to look into some data on the distinction between turnovers in general and turnover-by-steal. Unfortunately bkref doesn't make it easy to aggregate so I'm just going to pick key years for the players I think to check - if you'd like to see me do specific others, or would like to do it yourself, please go ahead and ask/do.

For each player season, I'll be listing out the On/Off for TO% and Opponent Steal%, sorted by the latter.

I'm also going to include non-point guards because I was curious about them as well.

Below, with each of these numbers used in this context for offensive players, negative On/Off is what you want.

Jordan '97: -2.8 -2.3
Nash '05: -4.5 -2.1
Tatum '23: -2.5 -2.1
Garnett '04: -3.2 -2.0
LeBron '09: -3.2 -2.0
Malone '97: -3.2 -1.8
Duncan '03: -2.7 -1.4
Paul '15: -2.2 -1.4
Nowitzki '07: -3.5 -1.3
Kawhi '16: -2.4 -1.3
Reggie '98: -3.4 -1.2
Stockton '97: -3.1 -1.2
Hali '23: -24. -1.2
Giannis '19: -0.8 -1.2
Kobe '06: -3.5 -1.1
Westbrook '17: -1.5 +1.0
Lillard '19: -3.0 -0.9
Manu '05: -1.0 -0.8
Amare '05: -2.3 -0.7
Iverson '01: -2.0 -0.7
Shaq '00: 0.0 -0.6
Ray '01: -0.5 -0.4
Curry '16: 0.2 -0.1
Harden '17: -0.3 0.2
Embiid '23: -0.6 0.3
Kidd '02: -1.1 0.4
Jokic '23: -2.4 0.5
Durant '14: 0.1 0.5
Luka '23: -0.9 0.7
Wade '09: -1.6 0.9

And then some guys from this year:

Shai: -4.5 -2.6
Maxey: -3.4 -2.2
Giannis: +0.2 -1.0
Embiid: -1.0 -0.5
Brunson: -1.9 -0.2
Harden: -0.7 -0.2
Lillard: -0.5 -0.2
Hali: -1.7 0.0
Tatum: -0.7 0.0
Kawhi: -0.7 +0.3
Jokic: -2.3 +0.5

I feel like my biggest takeaway here is that point guards don't seem to have much, if any, more control on team turnovers than other players do.

Beyond that, Jordan, Nash, as well as this year's Shai & Maxey really stand out.