It's a bit unfair to characterize Durant as a high turnover player. He has the same career TO% as Lebron James at 12.5. Lebron was the lead playmaker for his two championship Miami teams, or rather every team he's ever been on. The secondary playmaker in Miami, Wade, has an even higher TO% at 13.8. The main reason for OKC's lack of playoff success since Harden was traded is because of injuries to both Westbrook and Durant nearly every playoffs. When both were healthy, the Thunder were an elite team. Even this year, they win 70% of their games. GS and San Antonio just happen to be historically good and make them look bad in comparison.
I'm not sure what difference their position makes. They are both primary ballhandlers for their team. How is that relevant in your mind? I'm missing the point. As far as a single stat telling the story, it really depends on the question. If the question is "who's more turnover prone", then that's fairly black and white. Rondo turns the ball over on a higher percentage of possessions.
You're overthinking this. I've already been over this with bbholic in another thread. Yes a turnover ends a possession but it's not a point lost. At least anymore than a missed shot is lost points. Likewise, a turnover isn't a point for the opponent. Other than end of quarter situations, the other team is still going to get their next possession regardless of whether you score, miss a shot or turn it over. If we had scored a basked instead of turning the ball over, would the opposition then get a possession with the chance to score? Of course, they would. We didn't give them an extra possession. The only time a TO is going to give the other team extra points is when the TO directly leads to a scoring play such as a steal and a breakaway. The NBA tracks points off turnovers but that includes any score following a TO. If I throw a bad pass and you intercept it, walk it up the floor, pass the ball 6 or 7 times and hit a shot at the buzzer then that's points off of a TO by NBA definition. Now in reality, that TO didn't lead to any higher probability of scoring than a missed shot or made shot would have. There's certainly TO that directly lead to points for the opponent but it's nowhere close to 2 points per turnover and the number of those per game isn't that high. What you were describing is potential for points but if you're going to count loss of potential points then you also need to deduct for every missed shot and every offensive rebound you allow and every offensive rebound that you don't get, every FT you miss, ever foul you commit,etc... Van Gundy and Popovich are defensive minded coaches. They are always going to want to depend on their defense first and foremost. They won't do anything that puts their defense in a bad situation. It's no different than NFL teams with a defensive minded head coach. They typically use more conservative offenses and depend on their defense to win games. All you have to do is to look at GS, they average 1 fewer TO per game than the Rockets do. They're more of a wide open offense than the structured Spurs offense. There's more than one system that can be successful.
In my humble opinion, turnover percentage is a very deceptive metric. I probably like it the least of all the metrics I look at. And here's my reasoning. 1. TOV% by definition is the number of turnovers a player commits in relation to the total possessions that player uses. So TOVs/(FGA +.44FTA +TOVs). This doesn't take into account how much of a teams possessions the player chews up.......especially when you also count the amount of times they make a play and an assist. And it's the number of turnovers in relation to the team's total possessions that makes the difference. For instance a player can have a relatively high turnover percentage of 20% but he very rarely shoots the basketball...so he doesn't use up a lot of team possessions. So a player with a 20% TOV rate could actually turn the ball over a lot less of his teams total possessions than a player with a 12% TOV rate that shoots the ball a lot. Case: Player 1: TOV 20%, commits 4 TOVs per 100 of his team's possessions, is the primary playmaker, generates 14.5 assists per 100 of his teams possessions. Player 2: TOV 16%, commits 6 TOVs per 100 of his team's possessions, is the primary playmaker, generates 9.6 assists per 100 of his team's possessions. Which player is the more efficient playmaker? Which player should be running the offense? 2. Guys that play defense and get steals and draw offensive charges on the defensive end can carry a higher turnover percentage and it doesn't negatively effect the team as much because they are offsetting their own turnovers by getting back possessions on the other end. Case in point: Player 1: TOV% = 16%, 6 TOVs per 100 team possessions, 2.2 steals per 100 team possessions, 1 offensive rebound per 100 team possessions, 9.6 assists per 100 team possessions Player 2: TOV% = 25%, 5.4 TOVs per 100 team possessions, 2.6 steals per 100 team possessions, 1.9 offensive rebounds per 100 team possessions, 16 assists per 100 team possessions Player 1's turnovers cost his team more than Player 2 because Player 2 offsets more of his turnovers with steals, offensive rebounds, and defense. Player 2's TOV% is simply high because it is a function of the low number of shots he takes and turnovers he makes. While Player 2's TOV% is artificially low because he takes a lot of shots and uses a lot of possessions. But....the net effect on the team is what is important to winning basketball teams. Scorers are gonna score. We want them to score. Harden is an excellent, efficient scorer. But we want efficient playmaking to go along with that great, efficient scoring. And Harden is not efficient enough as a playmaker. His playmaking hurts us because he forces the play, and relies on getting bailed out with foul calls, and he throws to the same areas of the court over and over allowing the defense to anticipate, steal the pass, and run it down our throats. --------------------------------------------------------------------------------------- This is why TOV% does not tell the story of how efficient a distributor is and how much or little their turnovers mean in the scheme of winning and losing. Of all the playmakers in the Association that have played in excess of 1000 minutes this season, James Harden has a lower STL/TOV ratio (per 100 team possessions) at 36.7% than everybody except these players: JJ Barea Damian Lillard Derrick Rose Reggie Jackson Emmanual Mudiay Brandon Knight That's a pretty poor list of defenders right there. And Barea and Knight aren't even true playmakers. Those are players that do not offset the turnovers they make by getting the ball back on the defensive end. That simply means their turnovers HURT MORE and it means one of their teammates has to pick up the slack for them. By contrast a guy like CP3 is at 82.5%. Paul makes up for practically all his turnovers by playing D and getting steals. Throw in all his offensive fouls drawn on the defensive end and he's probably a net positive....meaning he more than makes up for his turnovers on the defensive end by getting them back. In addition James Harden's Assist/Turnover ratio is 1.6. Find me a primary playmaker of a playoff team with a lower assist/turnover ratio. No, don't. Because you won't. He has the lowest Assist/turnover ratio of any primary playmaker on a team going to the playoffs. And he has one of the lowest assist/turnover ratios of any primary playmaker in the league that has played over 1000 minutes. These are the reasons why Harden is not good enough as the primary playmaker to carry this team. Harden is a SCORER. And one of the best in the business. As a playmaker.....he is way below average. We need a primary playmaker that will spoon-feed Harden and make his job so much easier. Doesn't mean Harden can't still pass the ball. But Harden is at his best when scoring the basketball. He should be averaging over 30 ppg alongside a point guard who is generating 10 assists a game by creating and throwing dimes to both him and Dwight. But Harden will have to see that and accept that fact. Until he does...we're stuck in the mud and can only go so far.
You're correct that Harden hasn't played as well as he did last season but that doesn't mean that he isn't still playing really well. He's still one of the 5 most efficient scorers in the league and he's posting a career high in pts, assists and rebounds. His turnovers are also at a career high but they are only 0.6 higher than they were last season. Sure it would be great if his TOs would be lower but our falloff this season isn't because Harden is averaging an extra 0.6 TO/game. If you really believe that you can't build around a guy who scores 28.5 pts/game on 59.4% TS% and with 7.4 ast and 6.3 reb then good luck trying to find a player to build around. The differences between this season and last season aren't that great if you look at Harden's numbers. Usage +0.6, TS% -.011, PTS +1.1, AST +0.4, REB +0.6, STL -0.2, TOV +0.6,FTA +0.2. As far as any of the floor time stats, everyone on the team's numbers are way down due to the fact that we've sucked for most of the season. There's a very good chance that next year Steph Curry won't have as good a season as he's having in '15-'16 but that doesn't mean that he still won't be great. Likewise, Durant isn't having a season equal to his MVP year but he's still a guy to build around.
aelliot a turnover is a lost shot for us and an extra shot for the opponent. Figure out how many points we make per shot we take and how many points our opponents make for every shot they take. The number each way will be somewhere around 1. Probably a little more except if we're playing totally crappy teams like the Sixers/Lakers. So the point differential is 2 points at least. Argue it 9 ways to sunday, it's a 2 point differential. Argue it with analytics that get paid to do basketball analysis for a living. It's a 2-point differential.
Interesting discussion, as always. I agree that there is a bad turnover vs others. A definition that would be easy to track is one that didn't result in the clock being stopped, ie play continued, and it puts you in transition defense (often at a rather big disadvantage). And yeah, I agree about a TO doesn't give the opps an extra possession, but I thought that was a minor point when the larger point is how efficient your Off and Def is. TO-free possessions and Defense against Bad TOs So, what I'd really like to see is a stat for pts from (TO-free possessions) and opponent pts from (possessions from bad TOs -- ones that didn't stop the clock) One thing we can probably measure for sure is the effect of Net Diff Pts has on playoff seeding. A 1 pt differential appears to make a big difference. Even a .5 differential can. So, we aren't talking about huge numbers here. We are mainly talking about how much TOs make you less eff on Off and Def. There is a Net in there somewhere that we can pin down to show why TOs are a big deal to coaches. And yes, I understand that Net Pt Differential will be caused by other factors, too. But it would be interesting to see what TO-free possession stats look like. btw: GSW is #12 in the league for TOs/possession, due to their pace. Rockets are 3rd. Interesting system fact: Some defenses don't care about TOs/gm and Some great offenses choose to accept more risk for TOs Don't bother looking, TO differential is not a very good indicator of anything regarding W/Ls. Spurs / Hawks are in the negative and 5th and 6th lowest in the league. Probably due to less interest in forcing TOs vs bad shots. Which brings up an interesting discussion, why do the best defenses not always use TOs/gm as an indicator? And why are certain offensive systems and plays designed to limit TOs, where others (like motion) accept more risk for turnovers?
TOV% is an estimate not an actual stat. It's an estimate because BBReference doesn't have the possession information for each player. The NBA does have the possession information and they track turnovers per actual possession. That's why you should use TO Ratio rather than TOV%. As far as STL/TOV, that's a pretty arbitrary ratio. So Harden's STL/TOV ratio is .367 this season. Tony Parker's STL/TOV ratio during the Spurs 2014 championship season was .243. So what does that prove other than STL/TOV isn't a meaninful stat?
It's a lost shot (not neccessarily points) for sure but it's certainly not an extra shot for the other team. If I turn the ball over then the opponent gets a possession. Now if instead I had made a basket or missed a shot, then does the opponent still get the next possession? Of course they do. Regardless of if I turn the ball over, score or miss a shot, my opponent is going to get the ball and an opportunity to score. Offensively, a turnover ends a possession just as a missed shot does or a made shot does. It doesn't cost you points but you do lose the opportunity to score. What your talking about is the opportunity for points. That doesn't neccessarily equate to actual points. I could have a 1 point lead and turn the ball over 4 straight possessions and still have the lead if the opponent doesn't score. In that scenario, those turnovers didn't cost me anything. It's like I said in my reply to HP, if you want to discuss lost opportunity for points, then you also need to include missed shots, offensive rebounds allowed to opponents, every rebound or loose ball that your team didn't get, fouls committed , etc.
Ummmm, TO ratio is simply the number of turnovers a player commits per 100 possessions that they use. That's not how the game is played. It's a team game. It's team possessions and turnovers per team possessions that counts. That's why I look at TOVs per 100 of the team's possessions. No, it's not an arbitrary ratio. It's a very simply peek at how much a player defends and gets back what he lost on the offensive end. The problem with your example is Tony Parker only committed 2.2 turnovers per game or 3.3 turnovers per 100 of his team's possessions. That's basically half what Harden is committing this season. So while Parker did have a deficit of 2.9 turnovers to steals per 100 possessions in 2014, Harden currently has a deficit of 3.8 turnovers to steals per 100 possessions. (And remember, you're using the extreme outlier on the bell curve for comparison here. And it's still not close.) Not to mention Parker's A/TO ratio was 3.0 in 2014 while Harden's A/TO ratio is at 1.6. By the way, you're pulling the outlier on the bell curve and trying to establish your point. The Spurs run a different system than we do. Pops actually makes them D up mano y mano and play honest D, not just gamble for the steal and leave Timmy naked exposed underneath. And they played very disciplined ball-sharing offense that season. Probably the most beautiful passing offense ever in the modern era. That's a lot different than a ball dominant Harden ramming the ball down to the lip or juking and firing every possession. Tell me how many teams won a championship in the 3-point era with their primary playmaker having an assist/turnover ratio equal or less than 1.6. Tell me how many teams won a championship with their primary playmaker having a stl/turnover ratio of less than .5. Go back and do the same exercise for teams that make the playoffs in the 3-point era. Now go back and combine those 2 factoids and tell us how many teams won a championship or made the playoffs when their primary playmaker had a A/TO ratio equal to or less than 1.6 combined with a STL/TO ratio of less than .4. Then try to convince us those ratios are meaningless.
Can I get some concrete numbers here? What IS the acceptable assist to turnover and steal to turnover ratio? Or does it even matter? Just whatever Harden has isn't good enough?
https://www.reddit.com/r/nba/comments/4bn8ck/nba_passer_rating_20152016/ Harden is 6th in the League in passer rating. 2nd amongst all non point guards behind one of the greatest passers to ever play the game.
If you're trying to look at Harden's TOs, then the only thing that matters is the possessions where he has the ball. Possessions where Harden isn't involved have no effect on his turnovers. Say a team has 50 possessions and player A consumes 10 of those possessions and turns the ball over 5 times. In those same 50 possessions player B consumes 20 possessions and also turns the ball over 5 times. If you're going by team possessions then they both have a turnover rate of 5/50 = .10. In reality though, player A (5/10) is turning the ball over at twice the rate of player B (5/20). It's a completely arbitrary number that you're trying to use to show that we can't win a title with Harden as the primary ballhandler. I simply showed you that as recently as 2014 a team won the title with a primary ballhandler who had a much lower ratio than Harden does. That shoots down your argument. Do you realize that things change most every season? Not too many years ago people would have laughed at you if you claimed that a championship team would shoot the amount of 3 pointers that today's teams shoot. Same with small ball lineups. There's always exceptions and different circumstances. You're quick to point out that the Spurs play a different style as a way to overlook the fact that Parker's ratio goes directly against your claim. You don't however mention the system that Harden plays in. It's a fast pace high risk system that depends on him generating offense for himself and others. How many players can you find that have been in that situation? Comparing some arbitrary stat for a player in one system with those of in another system is meaningless. Offense isn't our problem, it's the defense. We start 3 players who's strength is suppose to be defense but this season they've been average at best on defense. The bottom line is that we have way too many players in our rotation that are average or worse on both sides of the ball. That's the problem. Who's our 2nd best offensive player after Harden? Beasley? A guy that we just picked up off the street? What does that tell you? Despite that we're still the 7th most efficient offense in the league. Offense isn't the problem. We need players who are above average on at least one side of the ball. If they can be above average on both sides then great but average to below average on both sides of the ball is a recipe for disaster. My guess is that we'll try and add at least one more plus offensive player to take the pressure off of Harden. Beasley's shown that you can play with Harden and still have a high usage (26.1%). Another offensive option would help but the more important thing is that we need some plus defenders. Whoever we put on the court next year need to be above average on at least one side of the ball. Our offense is good enough,that's not the problem.
For some reason he thinks he is a point guard... Just shoot the ball and stop trying to get everyone else involve that is the point guard job geez..
This has been a very interesting discussion. Nyloncalculus had written an article just recently about this. How liveball turnovers are much more harming than other kind of turnovers. http://nyloncalculus.com/2016/02/15/we-know-better-now-a-note-on-outdated-statistics/ Here is the interesting part : As for liveball turnovers(the turnovers that are the most harmfull because they lead to fastbreaks) till 2/9/16 Houston was at 58.3%. That's no. 5 most in the league. At the same time no.3 at most total turnovers in the league and no.2 at most steals allowed. So all in all Houston is one of the teams who hurts the MOST in the league by turnovers. As for Harden himself Nylon has Harden as no.1 at TOV USAGE in all the league and positions and as no.9 at true tov% among guards. There are no starters infront of him except Mudiay a rookie and MCW who got benched (coz of his turnovers in large part). Turnover Usage: Percentage of plays during which the player commits a turnover. True Turnover %: The percentage of turnovers committed by the player relative to all “contributing” possessions.
Artis Gilmore's record for most TOs in an NBA season is 366. Harden is averaging a robust 4.6 TOs a game (#1 in the NBA) with a total of 332 TOs thus far for the season (also #1 in the NBA). With 10 games left, Harden is still on pace to break this dubious record. #Pursuit
