Databallr Releases a New Stat: zTS%

Moderators: bwgood77, Clav, zimpy27, KingDavid, cupcakesnake, bisme37, Domejandro, ken6199, infinite11285, Dirk

User avatar
E-Balla
RealGM
Posts: 36,902
And1: 26,078
Joined: Dec 19, 2012
Location: The Poster Formerly Known As The Gotham City Pantalones
   

Re: Databallr Releases a New Stat: zTS% 

Post#101 » by E-Balla » Wed Apr 29, 2026 4:53 pm

Warriors Analyst wrote:
E-Balla wrote:Re: Steph's playtype and Draymond - Yes Draymond isn't an outlier passer, but I'd argue he's about as much of an outlier screener as there is. Is it totally legal? Maybe not, but Steph's playtype was easy from 2015 to 2019 largely because the Warriors as a team took nothing but easy shots and did not use hard play types. Steph in 2016 had a +0.2 playtype difficulty adjustment and led the team. Livingston was a -1.2. Dray was a -1.6. Barbosa a -0.8. Klay a -0.3. Iguodala a -1.4. Barnes a -0.7. This seems completely and totally accurate to me.


I get what you are saying, to an extent. But I think the off-ball movement and mileage Steph accrued to get a lot of those off-ball shots should be considered here in terms of "difficulty." Steph took a lot of off-ball and motion shots that were absolutely, by any definition, "open", but those aren't shots many guys in the league were capable of making. Most star players don't put in the off-ball work to have a shot diet similar to Steph. Maxey might be the closest I've seen, in terms of off-ball numbers.

You're right and because of this Steph has a +10.6 zTS% in 2015, a +13.0 in 2016, +7.3 in 2017, +12.4 in 2018, and +9.2 in 2019. He has more top 5 finishes in zTS% than TS% and a 6th place finish in zTS% in 2014.
Doctor MJ
Senior Mod
Senior Mod
Posts: 54,711
And1: 23,562
Joined: Mar 10, 2005
Location: Cali
     

Re: Databallr Releases a New Stat: zTS% 

Post#102 » by Doctor MJ » Wed Apr 29, 2026 5:19 pm

E-Balla wrote:Re: Steph's playtype and Draymond - Yes Draymond isn't an outlier passer, but I'd argue he's about as much of an outlier screener as there is. Is it totally legal? Maybe not, but Steph's playtype was easy from 2015 to 2019 largely because the Warriors as a team took nothing but easy shots and did not use hard play types. Steph in 2016 had a +0.2 playtype difficulty adjustment and led the team. Livingston was a -1.2. Dray was a -1.6. Barbosa a -0.8. Klay a -0.3. Iguodala a -1.4. Barnes a -0.7. This seems completely and totally accurate to me.


Focusing in on this "hard playtype" concept:

If the Warriors did not use hard playtypes, that seems smart. So why are other teams so dumb as to use hard playtypes?

My wording is purposefully simple-minded here and not intended as an actual insult to other teams, who I'm certainly not going to call "dumb" en masse.

I'll also say: As an advocate for Kerr's coaching, and for read & react systems generally, it makes a lot of sense to me that some measure would say the Warriors are using a playtype that they can be more effective with than with other approaches, but that gets to one of the key rebuttals to those of us who tend to advocate for read & react:

A lot of teams have tried read & react approaches and failed, so what is it about the ones that succeed that is separating them? because it seems unlikely that the difference is actually about those teams failing to call upon similar playtypes to the successful team.

One way of answering this might be that we're talking about two distinct but related phenomena:

1. The expected success in a particular scheme (with specific playtype footprint) with a random group of league-appropriate talents.
2. The actual success using a particular scheme among the set of players on teams that choose & coach players that can thrive in that playtype footprint.
Getting ready for the RealGM 100 on the PC Board

Come join the WNBA Board if you're a fan!
User avatar
E-Balla
RealGM
Posts: 36,902
And1: 26,078
Joined: Dec 19, 2012
Location: The Poster Formerly Known As The Gotham City Pantalones
   

Re: Databallr Releases a New Stat: zTS% 

Post#103 » by E-Balla » Wed Apr 29, 2026 5:45 pm

Doctor MJ wrote:Focusing in on this "hard playtype" concept:

If the Warriors did not use hard playtypes, that seems smart. So why are other teams so dumb as to use hard playtypes?

My wording is purposefully simple-minded here and not intended as an actual insult to other teams, who I'm certainly not going to call "dumb" en masse.

I'll also say: As an advocate for Kerr's coaching, and for read & react systems generally, it makes a lot of sense to me that some measure would say the Warriors are using a playtype that they can be more effective with than with other approaches, but that gets to one of the key rebuttals to those of us who tend to advocate for read & react:

A lot of teams have tried read & react approaches and failed, so what is it about the ones that succeed that is separating them? because it seems unlikely that the difference is actually about those teams failing to call upon similar playtypes to the successful team.

One way of answering this might be that we're talking about two distinct but related phenomena:

1. The expected success in a particular scheme (with specific playtype footprint) with a random group of league-appropriate talents.
2. The actual success using a particular scheme among the set of players on teams that choose & coach players that can thrive in that playtype footprint.

The successful team, because let's be real here the Warriors are the only real successful team that played a heavy read and react style in this way (the Spurs always had Tony and Manu PNR and most teams that historically have low playtype difficulty it is because they thrive in transition) and it took a perfect combination of players and coaching to do so. Without 2 of the 5-10 best shooters ever, without a ton of illegal screens, without high level passers/thinkers/processors like Dray, Iggy, Bogut, Livingston it's DOA. That's why the Warriors can't recreate that and Kerr had to dumb down the system post 2020. It's hard to get that strong of a collection of passers, screeners, and cutters. I'd say it's only a handful of teams all time with more than 3 guys I trust to consistently make good reads and the Warriors had 6.
Doctor MJ
Senior Mod
Senior Mod
Posts: 54,711
And1: 23,562
Joined: Mar 10, 2005
Location: Cali
     

Re: Databallr Releases a New Stat: zTS% 

Post#104 » by Doctor MJ » Wed Apr 29, 2026 6:26 pm

E-Balla wrote:
Doctor MJ wrote:Focusing in on this "hard playtype" concept:

If the Warriors did not use hard playtypes, that seems smart. So why are other teams so dumb as to use hard playtypes?

My wording is purposefully simple-minded here and not intended as an actual insult to other teams, who I'm certainly not going to call "dumb" en masse.

I'll also say: As an advocate for Kerr's coaching, and for read & react systems generally, it makes a lot of sense to me that some measure would say the Warriors are using a playtype that they can be more effective with than with other approaches, but that gets to one of the key rebuttals to those of us who tend to advocate for read & react:

A lot of teams have tried read & react approaches and failed, so what is it about the ones that succeed that is separating them? because it seems unlikely that the difference is actually about those teams failing to call upon similar playtypes to the successful team.

One way of answering this might be that we're talking about two distinct but related phenomena:

1. The expected success in a particular scheme (with specific playtype footprint) with a random group of league-appropriate talents.
2. The actual success using a particular scheme among the set of players on teams that choose & coach players that can thrive in that playtype footprint.

The successful team, because let's be real here the Warriors are the only real successful team that played a heavy read and react style in this way (the Spurs always had Tony and Manu PNR and most teams that historically have low playtype difficulty it is because they thrive in transition) and it took a perfect combination of players and coaching to do so. Without 2 of the 5-10 best shooters ever, without a ton of illegal screens, without high level passers/thinkers/processors like Dray, Iggy, Bogut, Livingston it's DOA. That's why the Warriors can't recreate that and Kerr had to dumb down the system post 2020. It's hard to get that strong of a collection of passers, screeners, and cutters. I'd say it's only a handful of teams all time with more than 3 guys I trust to consistently make good reads and the Warriors had 6.


Sounds like you're saying:

"It's incredibly hard for a team to succeed at running an offense focused on taking advantage of not-hard playtypes"

Would you agree with that? I would say there's definitely some truth in that.

Re: That's why Warriors can't recreate. But, is the stat actually saying they can't recreate this?

When I look at Curry's zTS page, the orange "role" line is pretty level compared to the blue "rTS" line and in general it doesn't seem like orange line explains much at all about why the blue is changing like it is.

Again, doesn't mean the stat is useless, but it doesn't seem to be correlating with your explanation of it seems to indicate it should.
Getting ready for the RealGM 100 on the PC Board

Come join the WNBA Board if you're a fan!
User avatar
FrodoBaggins
Veteran
Posts: 2,884
And1: 4,635
Joined: Dec 25, 2013

Re: Databallr Releases a New Stat: zTS% 

Post#105 » by FrodoBaggins » Wed Apr 29, 2026 6:31 pm

Teams that try to generate more "easy" play types (transition, cuts, rolls, etc) do so generally at the expense of turnovers. Increased passing, ball, and player movement is typically going to result in more of them. That's the tradeoff for a higher shooting percentage, and these types of teams do usually have higher FG%, 2PT%, eFG%, TS%. Sometimes the gamble works and results in a higher ORTG, and sometimes it doesn't.

It worked for Golden State, and they had their fears about turnovers early on:

https://www.espn.com/espn/feature/story/_/page/enterpriseWarriors/how-steve-kerr-revolutionized-golden-state-warriors-offense-charcuterie-board

"CAN WE EVEN do this?" Myers wonders.

It's early November, and doubt is creeping into the mind of the Warriors' GM. His team has begun the season 5-0, but basketball-wise, it's a disaster. The Warriors are racking up turnovers like they're storing them for winter -- averaging 21.6 per game. That's not only the worst mark in the league, it's about five turnovers per game more than the worst team in the prior season and only a few off the worst mark in NBA history.

After each game, Gelfand has been feeding Kerr postgame statistical reports, and the first stats listed are always passes: passes per game, secondary assists, free throw assists, the number of possessions with zero to two passes, three to five, six-plus. In morning film sessions, while coaches show players 15 to 20 clips from the previous game, they also post passing totals. And indeed, the Warriors are hitting that 300-per-game mark -- averaging 320.8, in fact, through the first five games, eighth best in the league.

The good news, then? The team is passing. The bad news? They're overpassing.

"Don't pass for the sake of it," Kerr implores his team. "If you're open, shoot it. If not, pass it. But don't be stationary. Move!"

Still, it's a struggle, like a classical flutist trying to learn to play jazz flute -- onstage, in real time. In a Nov. 9 loss to Phoenix, the Warriors tally 26 turnovers -- 10 by Curry alone -- after also notching 26 the game before against Houston.

Curry, for his part, is relying on what Kerr calls "horrible tendencies" -- careless left-handed hook passes over the top of defenses -- but also doing something far worse: remaining stationary after making passes. Defenses are manhandling Curry, and Kerr tells his star to run from pressure, not fight it, that even a back cut without getting the ball is a productive play because he's taking the defense with him. Instead, Curry is, as Kerr came to call it, "dancing" in place -- and stopping their offense as a result. Meanwhile, Draymond Green, former second-round draft pick, is trying too hard to establish himself as one of the team's top playmakers. He'll show potential, then become frustrated if he fails. "Keep it simple," the staff tells the third-year forward. "You can make plays, but make the simple play."

For weeks, Kerr has harped on the turnovers. In August, Kerr had visited Seattle Seahawks coach Pete Carroll during his team's training camp and had seen, in the Seahawks' defensive meeting room, a football on a rubber handle attached to a wall; as players came in and out, they'd hit the ball, trying to knock it loose. Carroll believed the habit would cause more fumbles. Ball possession, Carroll preached, is everything.

For the first six games, in this regard, the Warriors have been a white-hot mess -- like a race car with a wobbly wheel. Game 7, Nov. 11, would be a date with the Spurs, defending NBA champions and the gold standard for ball movement. Kerr had played four seasons (and won two titles) under Spurs coach Gregg Popovich and had admired how the Spurs' passing helped foster a selfless, team-first culture.

"It wasn't just play your best five guys to death," Kerr says. "It was play everybody. You go deep into your rotation, even if it means losing a couple of games in the regular season, just empower everybody. It's kind of the beauty of basketball, the old cliché about the total being greater than the sum of its parts -- I believe in all of that. Five guys have to operate together, but the other seven on the bench, or nine, however many, they've got to feel part of it."

Revenge would also be a factor. The Spurs had beaten the Warriors in the playoffs two seasons earlier, then swept them in the regular season the year before Kerr arrived. "The reason we made all these changes," Gelfand says, "was to get on their level." From day one, Curry recalls, Kerr had talked about the Spurs and their legacy. Though the season is young, the Warriors know this game will be a measuring stick.

They do not measure up.

“It's kind of the beauty of basketball, the old cliché about the total being greater than the sum of its parts -- I believe in all of that.”

- Steve Kerr

It begins well enough, the Warriors clinging to a 38-34 lead midway through the second quarter. But then the Spurs hit their stride. With less than a minute before halftime, forward Boris Diaw pump-fakes two Warriors on the perimeter, drives and zips the ball to Manu Ginobili on the right wing, who catches it with his left hand and in the same motion whips it to the right corner, where Tony Parker has enough time to do his taxes before swishing a 3-pointer. Six seconds of perfection.

After halftime, the Spurs cash in on more sloppy Warriors turnovers -- an errant pass by Curry, a fumble by Green. By this point in the season, Kerr has seen so many mistakes that he's been repeating, repeatedly, the phrase: "We're just slingin' the ball around out there!" It's like a mantra. Or a koan. He's saying it so much that his wife, Margot, has begun chiding him for it. And that's what Kerr sees against the Spurs: more carelessness, more slingin' the ball.

The Spurs, who feature the same Big Three -- Tim Duncan, Ginobili and Parker -- as they had when Kerr played beside them a dozen years before, cruise to a 113-100 win. In the locker room, Kerr explains to his deflated team that it doesn't matter that the Warriors had outshot the Spurs. Not only had they lost the turnover battle, 19-8, they'd lost their focus. "Look, guys," Gentry adds, "you don't want to say it, but this is how we want to play. This is who we want to emulate."

It's an enigma -- and a conundrum. They need to play with pace but protect the ball. They need to play unselfishly but not too unselfishly. Pass the ball, but don't turn down a great shot.


"Can we do that?" Kerr asks.

"The main goal," Steph Curry says, "is to just make the defense make as many decisions as you can so that they're going to mess up at some point with all that ball movement and body movement and whatnot."

IT'S JUNE 13, 2017, 24 hours after the Warriors have won their second championship in three years. They've humiliated the Cavaliers in five games, and Vino Volo in Oakland International Airport is abuzz, as usual. Wine unites everyone, Volo's staffers like to say. But it's seats C1 and C2 at the end of the bar, where Kerr and Fraser sat on that August afternoon, that to them are now legend.

Lawrence Flores, a 36-year-old assistant manager, was working the floor that day, stealing peeks at Kerr's demonstration. He's told the story a dozen times to friends and family. "That could've been the creation of this offense," he tells them, "That could've been the start." When Ninkovich, now 32, watches the Warriors, he sometimes sees not players but cranberries and almonds.

After their loss three seasons ago to the Spurs and inspired by the manner in which San Antonio had filleted them, the Warriors went on to win their next 16 games. "It was," Kerr says of that Spurs loss, "the best thing that could've ever happened to us." Pre-Spurs loss, the Warriors had ranked last in turnover percentage, with Green amassing more turnovers than assists. The rest of the season, they would rank sixth in turnover percentage, with Green averaging twice as many assists as turnovers.

And the epiphany arrived just five days after that Spurs defeat. By virtue of a schedule quirk, the Warriors were granted a four-day break after a road game against the Lakers, and when Kerr entered the visitors locker room at Staples Center before tip-off, he proffered a deal: "Play the way we've been talking about and play the right way -- take care of the ball, defend, do all that stuff -- and I'll give you the next two days off." The players literally gasped in disbelief.

That night, there wasn't one moment, or a singular play, but a river of them -- a constant flow, the ball pinballing around the court, side to side, to the tune of 343 passes. "Beautiful," Kerr says, thinking back on it. The Warriors scored a season-high 136 points.

In the days prior, what Kerr had most wanted was to know that his words were being heeded. "You just want to know the ship is heading in the right direction," he says. And as he watched the rout unfold, he saw everything he had been preaching, his players carrying out his vision with focus and flair.

The transformation was radical -- and ruthlessly effective. By the end of the season, the Warriors ranked second in offensive efficiency and first in defensive efficiency. They averaged 315.9 passes per game, nearly 70 more than the season before -- the second-biggest leap in the league. They had the highest increase that season in assists per game and secondary assists per game, and the second-highest jump in assist-to-turnover ratio. They would go on to win an NBA-record 73 games the next season, falling one win shy of a second consecutive NBA title, and Curry would win his second straight NBA MVP award, just as Nash had done in Phoenix exactly one decade earlier in the offense that so inspired Kerr.

The Warriors ultimately found that if defenses were panicked about the first pass, by the time their third pass arrived, they were rewarded with a wide-open corner 3. "The main goal," Curry says, "is to just make the defense make as many decisions as you can so that they're going to mess up at some point with all that ball movement and body movement and whatnot. But it took awhile for us to kind of get the understanding of where each other was going to be without having to call a set play or whatnot. So it took awhile."

Actually, it took eight regular-season games.

It took. And it held: The Warriors today claim the three highest assist-per-game averages of the past two decades. And all have come in the past three years. "I can't sit here and say we knew this was going to happen," Fraser says, "but if I go back and read Steve's thesis on what he wished for, it's very close to what happened."

Consider: Since the start of the 1995-96 season, nine of the 10 best teams in offensive efficiency were either the mid-'90s Bulls (where Kerr played), the Nash-led Suns (where Kerr managed) or Kerr's modern-day Warriors. Kerr's basketball journey weaved through offensive greatness, and then he built his own.

"It was like it was destiny to have Steve come in and try to coach that way," says Luke Walton, former Warriors assistant coach, "because they were built to play that way."

And after two seasons alongside Kerr, soon after Walton agreed to take over the rebuilding Lakers, the new coach announced to his team that he wanted to create a nightly goal. He wanted something to establish a culture. Something to make everyone feel a part of a whole. Luke Walton wanted 300 passes a game.


zTS% isn't concerned with turnovers, so it's not going to show that consequence. It's purely concerned with scoring efficiency. I'd say shooting efficiency, actually, as turnovers are a part of scoring. But it's hard to separate scoring turnovers and passing turnovers in the Synergy play type data, as databallr talked about in the article.

Still a clear upgrade over TS% and rTS%/TS+ for the majority of use cases.
User avatar
E-Balla
RealGM
Posts: 36,902
And1: 26,078
Joined: Dec 19, 2012
Location: The Poster Formerly Known As The Gotham City Pantalones
   

Re: Databallr Releases a New Stat: zTS% 

Post#106 » by E-Balla » Wed Apr 29, 2026 7:12 pm

Doctor MJ wrote:
E-Balla wrote:
Doctor MJ wrote:Focusing in on this "hard playtype" concept:

If the Warriors did not use hard playtypes, that seems smart. So why are other teams so dumb as to use hard playtypes?

My wording is purposefully simple-minded here and not intended as an actual insult to other teams, who I'm certainly not going to call "dumb" en masse.

I'll also say: As an advocate for Kerr's coaching, and for read & react systems generally, it makes a lot of sense to me that some measure would say the Warriors are using a playtype that they can be more effective with than with other approaches, but that gets to one of the key rebuttals to those of us who tend to advocate for read & react:

A lot of teams have tried read & react approaches and failed, so what is it about the ones that succeed that is separating them? because it seems unlikely that the difference is actually about those teams failing to call upon similar playtypes to the successful team.

One way of answering this might be that we're talking about two distinct but related phenomena:

1. The expected success in a particular scheme (with specific playtype footprint) with a random group of league-appropriate talents.
2. The actual success using a particular scheme among the set of players on teams that choose & coach players that can thrive in that playtype footprint.

The successful team, because let's be real here the Warriors are the only real successful team that played a heavy read and react style in this way (the Spurs always had Tony and Manu PNR and most teams that historically have low playtype difficulty it is because they thrive in transition) and it took a perfect combination of players and coaching to do so. Without 2 of the 5-10 best shooters ever, without a ton of illegal screens, without high level passers/thinkers/processors like Dray, Iggy, Bogut, Livingston it's DOA. That's why the Warriors can't recreate that and Kerr had to dumb down the system post 2020. It's hard to get that strong of a collection of passers, screeners, and cutters. I'd say it's only a handful of teams all time with more than 3 guys I trust to consistently make good reads and the Warriors had 6.


Sounds like you're saying:

"It's incredibly hard for a team to succeed at running an offense focused on taking advantage of not-hard playtypes"

Would you agree with that? I would say there's definitely some truth in that.

Yeah that sounds right.

Re: That's why Warriors can't recreate. But, is the stat actually saying they can't recreate this?

No but that's not the point of the stat is it? That's what Kerr said when he explained how he simplified the system, and it shows in the stat specifically for Curry, but isn't too big a deal here I don't think.

When I look at Curry's zTS page, the orange "role" line is pretty level compared to the blue "rTS" line and in general it doesn't seem like orange line explains much at all about why the blue is changing like it is.

Again, doesn't mean the stat is useless, but it doesn't seem to be correlating with your explanation of it seems to indicate it should.

If you compare Curry in zTS to Curry in rTS yeah it's similar. His efficiency drops in a year and that would be represented in both, same for it going up. Here's where it really shows - his rank in rTS/zTS by year:
10: 65/61
11: 27/14
12 (didn't qualify but if he did): 14/9
13: 35/15
14: 12/6
15: 4/3
16: 1/1
17: 15/15
18: 1/1
19: 10/3
21: 13/3
22: 58/29
23: 18/6
24: 58/23
25: 48/22
26: 29/12

Steph ranks a lot higher by zTS% excluding his rookie and 15-18 years. I don't think this says anything particularly about 15-19 other than saying Steph's playtypes selection was built off a lot of actions associated with motion shooters as opposed to PGs.
User avatar
zimpy27
Forum Mod
Forum Mod
Posts: 46,860
And1: 45,353
Joined: Jul 13, 2014

Re: Databallr Releases a New Stat: zTS% 

Post#107 » by zimpy27 » Wed Apr 29, 2026 7:50 pm

Do you actually want a much higher zTS% than TS%?

Doesn't it suggest that you are taking bad shots? Sure you make some but seems like you could be more efficient without them.
"Let's play some basketball!" - Fergie
Doctor MJ
Senior Mod
Senior Mod
Posts: 54,711
And1: 23,562
Joined: Mar 10, 2005
Location: Cali
     

Re: Databallr Releases a New Stat: zTS% 

Post#108 » by Doctor MJ » Wed Apr 29, 2026 8:18 pm

FrodoBaggins wrote:Teams that try to generate more "easy" play types (transition, cuts, rolls, etc) do so generally at the expense of turnovers. Increased passing, ball, and player movement is typically going to result in more of them. That's the tradeoff for a higher shooting percentage, and these types of teams do usually have higher FG%, 2PT%, eFG%, TS%. Sometimes the gamble works and results in a higher ORTG, and sometimes it doesn't.

It worked for Golden State, and they had their fears about turnovers early on:

https://www.espn.com/espn/feature/story/_/page/enterpriseWarriors/how-steve-kerr-revolutionized-golden-state-warriors-offense-charcuterie-board

"CAN WE EVEN do this?" Myers wonders.

It's early November, and doubt is creeping into the mind of the Warriors' GM. His team has begun the season 5-0, but basketball-wise, it's a disaster. The Warriors are racking up turnovers like they're storing them for winter -- averaging 21.6 per game. That's not only the worst mark in the league, it's about five turnovers per game more than the worst team in the prior season and only a few off the worst mark in NBA history.

After each game, Gelfand has been feeding Kerr postgame statistical reports, and the first stats listed are always passes: passes per game, secondary assists, free throw assists, the number of possessions with zero to two passes, three to five, six-plus. In morning film sessions, while coaches show players 15 to 20 clips from the previous game, they also post passing totals. And indeed, the Warriors are hitting that 300-per-game mark -- averaging 320.8, in fact, through the first five games, eighth best in the league.

The good news, then? The team is passing. The bad news? They're overpassing.

"Don't pass for the sake of it," Kerr implores his team. "If you're open, shoot it. If not, pass it. But don't be stationary. Move!"

Still, it's a struggle, like a classical flutist trying to learn to play jazz flute -- onstage, in real time. In a Nov. 9 loss to Phoenix, the Warriors tally 26 turnovers -- 10 by Curry alone -- after also notching 26 the game before against Houston.

Curry, for his part, is relying on what Kerr calls "horrible tendencies" -- careless left-handed hook passes over the top of defenses -- but also doing something far worse: remaining stationary after making passes. Defenses are manhandling Curry, and Kerr tells his star to run from pressure, not fight it, that even a back cut without getting the ball is a productive play because he's taking the defense with him. Instead, Curry is, as Kerr came to call it, "dancing" in place -- and stopping their offense as a result. Meanwhile, Draymond Green, former second-round draft pick, is trying too hard to establish himself as one of the team's top playmakers. He'll show potential, then become frustrated if he fails. "Keep it simple," the staff tells the third-year forward. "You can make plays, but make the simple play."

For weeks, Kerr has harped on the turnovers. In August, Kerr had visited Seattle Seahawks coach Pete Carroll during his team's training camp and had seen, in the Seahawks' defensive meeting room, a football on a rubber handle attached to a wall; as players came in and out, they'd hit the ball, trying to knock it loose. Carroll believed the habit would cause more fumbles. Ball possession, Carroll preached, is everything.

For the first six games, in this regard, the Warriors have been a white-hot mess -- like a race car with a wobbly wheel. Game 7, Nov. 11, would be a date with the Spurs, defending NBA champions and the gold standard for ball movement. Kerr had played four seasons (and won two titles) under Spurs coach Gregg Popovich and had admired how the Spurs' passing helped foster a selfless, team-first culture.

"It wasn't just play your best five guys to death," Kerr says. "It was play everybody. You go deep into your rotation, even if it means losing a couple of games in the regular season, just empower everybody. It's kind of the beauty of basketball, the old cliché about the total being greater than the sum of its parts -- I believe in all of that. Five guys have to operate together, but the other seven on the bench, or nine, however many, they've got to feel part of it."

Revenge would also be a factor. The Spurs had beaten the Warriors in the playoffs two seasons earlier, then swept them in the regular season the year before Kerr arrived. "The reason we made all these changes," Gelfand says, "was to get on their level." From day one, Curry recalls, Kerr had talked about the Spurs and their legacy. Though the season is young, the Warriors know this game will be a measuring stick.

They do not measure up.

“It's kind of the beauty of basketball, the old cliché about the total being greater than the sum of its parts -- I believe in all of that.”

- Steve Kerr

It begins well enough, the Warriors clinging to a 38-34 lead midway through the second quarter. But then the Spurs hit their stride. With less than a minute before halftime, forward Boris Diaw pump-fakes two Warriors on the perimeter, drives and zips the ball to Manu Ginobili on the right wing, who catches it with his left hand and in the same motion whips it to the right corner, where Tony Parker has enough time to do his taxes before swishing a 3-pointer. Six seconds of perfection.

After halftime, the Spurs cash in on more sloppy Warriors turnovers -- an errant pass by Curry, a fumble by Green. By this point in the season, Kerr has seen so many mistakes that he's been repeating, repeatedly, the phrase: "We're just slingin' the ball around out there!" It's like a mantra. Or a koan. He's saying it so much that his wife, Margot, has begun chiding him for it. And that's what Kerr sees against the Spurs: more carelessness, more slingin' the ball.

The Spurs, who feature the same Big Three -- Tim Duncan, Ginobili and Parker -- as they had when Kerr played beside them a dozen years before, cruise to a 113-100 win. In the locker room, Kerr explains to his deflated team that it doesn't matter that the Warriors had outshot the Spurs. Not only had they lost the turnover battle, 19-8, they'd lost their focus. "Look, guys," Gentry adds, "you don't want to say it, but this is how we want to play. This is who we want to emulate."

It's an enigma -- and a conundrum. They need to play with pace but protect the ball. They need to play unselfishly but not too unselfishly. Pass the ball, but don't turn down a great shot.


"Can we do that?" Kerr asks.

"The main goal," Steph Curry says, "is to just make the defense make as many decisions as you can so that they're going to mess up at some point with all that ball movement and body movement and whatnot."

IT'S JUNE 13, 2017, 24 hours after the Warriors have won their second championship in three years. They've humiliated the Cavaliers in five games, and Vino Volo in Oakland International Airport is abuzz, as usual. Wine unites everyone, Volo's staffers like to say. But it's seats C1 and C2 at the end of the bar, where Kerr and Fraser sat on that August afternoon, that to them are now legend.

Lawrence Flores, a 36-year-old assistant manager, was working the floor that day, stealing peeks at Kerr's demonstration. He's told the story a dozen times to friends and family. "That could've been the creation of this offense," he tells them, "That could've been the start." When Ninkovich, now 32, watches the Warriors, he sometimes sees not players but cranberries and almonds.

After their loss three seasons ago to the Spurs and inspired by the manner in which San Antonio had filleted them, the Warriors went on to win their next 16 games. "It was," Kerr says of that Spurs loss, "the best thing that could've ever happened to us." Pre-Spurs loss, the Warriors had ranked last in turnover percentage, with Green amassing more turnovers than assists. The rest of the season, they would rank sixth in turnover percentage, with Green averaging twice as many assists as turnovers.

And the epiphany arrived just five days after that Spurs defeat. By virtue of a schedule quirk, the Warriors were granted a four-day break after a road game against the Lakers, and when Kerr entered the visitors locker room at Staples Center before tip-off, he proffered a deal: "Play the way we've been talking about and play the right way -- take care of the ball, defend, do all that stuff -- and I'll give you the next two days off." The players literally gasped in disbelief.

That night, there wasn't one moment, or a singular play, but a river of them -- a constant flow, the ball pinballing around the court, side to side, to the tune of 343 passes. "Beautiful," Kerr says, thinking back on it. The Warriors scored a season-high 136 points.

In the days prior, what Kerr had most wanted was to know that his words were being heeded. "You just want to know the ship is heading in the right direction," he says. And as he watched the rout unfold, he saw everything he had been preaching, his players carrying out his vision with focus and flair.

The transformation was radical -- and ruthlessly effective. By the end of the season, the Warriors ranked second in offensive efficiency and first in defensive efficiency. They averaged 315.9 passes per game, nearly 70 more than the season before -- the second-biggest leap in the league. They had the highest increase that season in assists per game and secondary assists per game, and the second-highest jump in assist-to-turnover ratio. They would go on to win an NBA-record 73 games the next season, falling one win shy of a second consecutive NBA title, and Curry would win his second straight NBA MVP award, just as Nash had done in Phoenix exactly one decade earlier in the offense that so inspired Kerr.

The Warriors ultimately found that if defenses were panicked about the first pass, by the time their third pass arrived, they were rewarded with a wide-open corner 3. "The main goal," Curry says, "is to just make the defense make as many decisions as you can so that they're going to mess up at some point with all that ball movement and body movement and whatnot. But it took awhile for us to kind of get the understanding of where each other was going to be without having to call a set play or whatnot. So it took awhile."

Actually, it took eight regular-season games.

It took. And it held: The Warriors today claim the three highest assist-per-game averages of the past two decades. And all have come in the past three years. "I can't sit here and say we knew this was going to happen," Fraser says, "but if I go back and read Steve's thesis on what he wished for, it's very close to what happened."

Consider: Since the start of the 1995-96 season, nine of the 10 best teams in offensive efficiency were either the mid-'90s Bulls (where Kerr played), the Nash-led Suns (where Kerr managed) or Kerr's modern-day Warriors. Kerr's basketball journey weaved through offensive greatness, and then he built his own.

"It was like it was destiny to have Steve come in and try to coach that way," says Luke Walton, former Warriors assistant coach, "because they were built to play that way."

And after two seasons alongside Kerr, soon after Walton agreed to take over the rebuilding Lakers, the new coach announced to his team that he wanted to create a nightly goal. He wanted something to establish a culture. Something to make everyone feel a part of a whole. Luke Walton wanted 300 passes a game.


zTS% isn't concerned with turnovers, so it's not going to show that consequence. It's purely concerned with scoring efficiency. I'd say shooting efficiency, actually, as turnovers are a part of scoring. But it's hard to separate scoring turnovers and passing turnovers in the Synergy play type data, as databallr talked about in the article.

Still a clear upgrade over TS% and rTS%/TS+ for the majority of use cases.


Excellent point Frodo!

So we maybe can characterize this situation as one where in order for an aggressive-passing team to be worthwhile you need a combo of 2 things:

a) Effective passers who can succeed at a relatively high rate when they try hard playmaking for others
b) Effective shooters who can succeed at a relatively high rate when the passes come to them

The more you have (a), the more it makes sense to aim for these easier playtypes.
The more you have (b), the better in all situations.

This is not a contradiction of Databallr's work, but something to be recognized alongside of it.

Then back to the Warriors:

I think the fact that they had such high turnovers is telling us that the playmaking was not in fact THAT good. Not saying it wasn't "good" as in above NBA-average, but not so good that there wasn't a major cost to it.

Now, one interpretation of that is essentially that the playmakers are acting as the sin-eaters for the shot takers, and there's some truth in that, but if the competitive advantage is having great shooters we should take care not overrate a playmaker's general capacity simply because he has outstanding targets any more than we'd want to underrate the playmaker because he plays with bad shooters.
Getting ready for the RealGM 100 on the PC Board

Come join the WNBA Board if you're a fan!
Doctor MJ
Senior Mod
Senior Mod
Posts: 54,711
And1: 23,562
Joined: Mar 10, 2005
Location: Cali
     

Re: Databallr Releases a New Stat: zTS% 

Post#109 » by Doctor MJ » Wed Apr 29, 2026 8:31 pm

E-Balla wrote:
Doctor MJ wrote:
E-Balla wrote:The successful team, because let's be real here the Warriors are the only real successful team that played a heavy read and react style in this way (the Spurs always had Tony and Manu PNR and most teams that historically have low playtype difficulty it is because they thrive in transition) and it took a perfect combination of players and coaching to do so. Without 2 of the 5-10 best shooters ever, without a ton of illegal screens, without high level passers/thinkers/processors like Dray, Iggy, Bogut, Livingston it's DOA. That's why the Warriors can't recreate that and Kerr had to dumb down the system post 2020. It's hard to get that strong of a collection of passers, screeners, and cutters. I'd say it's only a handful of teams all time with more than 3 guys I trust to consistently make good reads and the Warriors had 6.


Sounds like you're saying:

"It's incredibly hard for a team to succeed at running an offense focused on taking advantage of not-hard playtypes"

Would you agree with that? I would say there's definitely some truth in that.

Yeah that sounds right.

Re: That's why Warriors can't recreate. But, is the stat actually saying they can't recreate this?

No but that's not the point of the stat is it? That's what Kerr said when he explained how he simplified the system, and it shows in the stat specifically for Curry, but isn't too big a deal here I don't think.

When I look at Curry's zTS page, the orange "role" line is pretty level compared to the blue "rTS" line and in general it doesn't seem like orange line explains much at all about why the blue is changing like it is.

Again, doesn't mean the stat is useless, but it doesn't seem to be correlating with your explanation of it seems to indicate it should.

If you compare Curry in zTS to Curry in rTS yeah it's similar. His efficiency drops in a year and that would be represented in both, same for it going up. Here's where it really shows - his rank in rTS/zTS by year:
10: 65/61
11: 27/14
12 (didn't qualify but if he did): 14/9
13: 35/15
14: 12/6
15: 4/3
16: 1/1
17: 15/15
18: 1/1
19: 10/3
21: 13/3
22: 58/29
23: 18/6
24: 58/23
25: 48/22
26: 29/12

Steph ranks a lot higher by zTS% excluding his rookie and 15-18 years. I don't think this says anything particularly about 15-19 other than saying Steph's playtypes selection was built off a lot of actions associated with motion shooters as opposed to PGs.


Glad we agree on the first.

Re: not what the stat is for. Well, you brought it up in this thread about the stat so I expected you thought it was related.

Re: rTS/zTS. A worthwhile thing to list side by side, but I'll just emphasize that the variance in rTS is way bigger than any changes in role in the graph, which means that the bulk of the explanation for rTS differences from season to season has to be explained by means other than what zTS can capture.

Again, doesn't make zTS not worthwhile. Databallr emphasized he was looking to distinguish between playmaking & shooting impact, and one way of framing this is that larger deviation if "Playmaking Impact" by the Warriors and its inevitable wax/wane over time.
Getting ready for the RealGM 100 on the PC Board

Come join the WNBA Board if you're a fan!
Doctor MJ
Senior Mod
Senior Mod
Posts: 54,711
And1: 23,562
Joined: Mar 10, 2005
Location: Cali
     

Re: Databallr Releases a New Stat: zTS% 

Post#110 » by Doctor MJ » Wed Apr 29, 2026 8:33 pm

zimpy27 wrote:Do you actually want a much higher zTS% than TS%?

Doesn't it suggest that you are taking bad shots? Sure you make some but seems like you could be more efficient without them.


So, sounds like you're experience similar confusion to what a lot of us did.

zTS is not about shot quality but about adjusting for playtype role. Read Databallr's site to get the full explanation.
Getting ready for the RealGM 100 on the PC Board

Come join the WNBA Board if you're a fan!
User avatar
E-Balla
RealGM
Posts: 36,902
And1: 26,078
Joined: Dec 19, 2012
Location: The Poster Formerly Known As The Gotham City Pantalones
   

Re: Databallr Releases a New Stat: zTS% 

Post#111 » by E-Balla » Thu Apr 30, 2026 1:11 am

zimpy27 wrote:Do you actually want a much higher zTS% than TS%?

Doesn't it suggest that you are taking bad shots? Sure you make some but seems like you could be more efficient without them.

No it suggests you're creating more of your offense. Again this has nothing to do with shot quality. The types of players with a much higher zTS than TS are passing point guards and dynamic guard playmakers and the types of players with a much lower zTS than TS are lob threat zero skill bigs that are easily much more replaceable scoring options.
f4p
Veteran
Posts: 2,551
And1: 2,545
Joined: Sep 19, 2021
 

Re: Databallr Releases a New Stat: zTS% 

Post#112 » by f4p » Thu Apr 30, 2026 2:07 am

Doctor MJ wrote:
zimpy27 wrote:Do you actually want a much higher zTS% than TS%?

Doesn't it suggest that you are taking bad shots? Sure you make some but seems like you could be more efficient without them.


So, sounds like you're experience similar confusion to what a lot of us did.

zTS is not about shot quality but about adjusting for playtype role. Read Databallr's site to get the full explanation.


And maybe just to put numbers to it. Let's say the average off the dribble long 2 is taken at 40% or 0.8 points per shot (PPS). And the average off the dribble 3 is taken at 35% (1.05 PPS). And the average isolation is 0.9 PPS.

If all your isolations result in long 2s that you make at 40%. You would look average by comparison to long 2s but you would get a negative zTS for isolations (0.8 PPS vs 0.9 PPS expected).

If all your isolations result in 3s that you make at 35%. You would look average by comparison to 3s but you would get a positive zTS for isolations (1.05 PPS vs 0.9 PPS expected).
f4p
Veteran
Posts: 2,551
And1: 2,545
Joined: Sep 19, 2021
 

Re: Databallr Releases a New Stat: zTS% 

Post#113 » by f4p » Thu Apr 30, 2026 2:18 am

Doctor MJ wrote:Now, one interpretation of that is essentially that the playmakers are acting as the sin-eaters for the shot takers, and there's some truth in that, but if the competitive advantage is having great shooters we should take care not overrate a playmaker's general capacity simply because he has outstanding targets any more than we'd want to underrate the playmaker because he plays with bad shooters.


I think there's a lot of truth to it. People often say Steph should have more assists when you think about what he creates for others ,but that would come with a side of more turnovers as well. And it's part of why off ball motion offenses are tough. Because you essentially outsource the playmaking and decision making to someone other than your best player.

Now if you have Draymond green (and Iggy) that can be a wonderful thing and create synergy. But the reason you usually let Nash/magic or LeBron/harden/Luka have the ball all the time is they are better than everyone else at making the reads and running everything and actually finding a second player who can do that (at a high level) is tough.

And for the warriors specifically, while Steph is obviously part of giving the warriors easy shots, I think the reason it's easy to think the decision makers, specifically Draymond, are a huge part of it, is the fact that Steph has a better efficiency when Draymond is on the court, the exact opposite of what we expect from a defensive specialist and also not to be expected if the shooter and not the decision maker is creating most of the benefit.
Doctor MJ
Senior Mod
Senior Mod
Posts: 54,711
And1: 23,562
Joined: Mar 10, 2005
Location: Cali
     

Re: Databallr Releases a New Stat: zTS% 

Post#114 » by Doctor MJ » Thu Apr 30, 2026 2:43 pm

f4p wrote:
Doctor MJ wrote:Now, one interpretation of that is essentially that the playmakers are acting as the sin-eaters for the shot takers, and there's some truth in that, but if the competitive advantage is having great shooters we should take care not overrate a playmaker's general capacity simply because he has outstanding targets any more than we'd want to underrate the playmaker because he plays with bad shooters.


I think there's a lot of truth to it. People often say Steph should have more assists when you think about what he creates for others ,but that would come with a side of more turnovers as well. And it's part of why off ball motion offenses are tough. Because you essentially outsource the playmaking and decision making to someone other than your best player.

Now if you have Draymond green (and Iggy) that can be a wonderful thing and create synergy. But the reason you usually let Nash/magic or LeBron/harden/Luka have the ball all the time is they are better than everyone else at making the reads and running everything and actually finding a second player who can do that (at a high level) is tough.

And for the warriors specifically, while Steph is obviously part of giving the warriors easy shots, I think the reason it's easy to think the decision makers, specifically Draymond, are a huge part of it, is the fact that Steph has a better efficiency when Draymond is on the court, the exact opposite of what we expect from a defensive specialist and also not to be expected if the shooter and not the decision maker is creating most of the benefit.


So, when it comes to the scale of said truth, it would sure be nice if it were tracked somewhere passes-to turnovers.

Closest relevant thing I can thin of is actually Databallr's oTOV value data. Obviously there's the rub that if the team keeps playing with the same level of playmaking recklessness with or without a player, that data is going to look very different from if they change things up when he goes to the bench. Arguably the ideal way to play with a player who requires dangerous passes-to is to go from "hare" mode to "turtoise" mode when he rests, but of course there are complications to switching schemes dramatically with routine substitutions.

If we look at the classic obvious dangerous pass-to avatar of Shaq, what we see is this:

In the earliest 4-year study nbarapm, Shaq rates as a slight positive +0.2 from '00-03, and after that he goes negative, and continues to get more negative mostly unabated for the rest of his career. He's a -0.3 from '01-04, and -1.9 from '08-11. Worth noting that in 2001 they removed the Illegal Defense rule which was artificially making entry passes to the low post easier.

With Curry we see him generally as having positive oTOV indicators. He peaks from '13-16 with a +1.0, which is interesting because that includes half Jackson years and half Kerr years. He doesn't go negative until '20-23 with a -0.5, which then goes to his nadir of -0.6 in '21-24.

By contrast if we look at the classic positive case here, Chris Paul, we see him north of +1.0 his whole career, peaking at +2.6 from '08-11.

A foil to Paul is Steve Nash who also had absolutely top tier passing ability but generally played a considerably more dangerous (for TOs) style of play. He's also positive the whole time, and peaks at +1.8 from '03-06, which splits two years in Dallas and two years in Phoenix.

Back to the Warriors, something interesting with Curry's signature assister Draymond: He's mostly negative by oTOV value. He peaks at +0.2 from '15-18 (Kerr's first 4 years), but of his career he posts a value worse than -0.5.

I would chalk that up to Dray being more aggressive in his passing than his teammates, which I'd generally consider to be a good thing, but of course, it would be a better thing if he could do it while reducing turnovers too.

For contrast there, the king of maximum-passing is Jokic, who is also a big who volume scores. I'd expect there's at least as heavy of a "tax" that Jokic is paying for this by oTOV value compared to Dray. Jokic's numbers are pretty close to zero here, but I will say they've generally been slightly positive, running only slightly negative in '19-22 & '20-23.
Getting ready for the RealGM 100 on the PC Board

Come join the WNBA Board if you're a fan!
dhsilv2
RealGM
Posts: 52,901
And1: 28,597
Joined: Oct 04, 2015

Re: Databallr Releases a New Stat: zTS% 

Post#115 » by dhsilv2 » Thu Apr 30, 2026 3:45 pm

Doctor MJ wrote:
f4p wrote:
Doctor MJ wrote:Now, one interpretation of that is essentially that the playmakers are acting as the sin-eaters for the shot takers, and there's some truth in that, but if the competitive advantage is having great shooters we should take care not overrate a playmaker's general capacity simply because he has outstanding targets any more than we'd want to underrate the playmaker because he plays with bad shooters.


I think there's a lot of truth to it. People often say Steph should have more assists when you think about what he creates for others ,but that would come with a side of more turnovers as well. And it's part of why off ball motion offenses are tough. Because you essentially outsource the playmaking and decision making to someone other than your best player.

Now if you have Draymond green (and Iggy) that can be a wonderful thing and create synergy. But the reason you usually let Nash/magic or LeBron/harden/Luka have the ball all the time is they are better than everyone else at making the reads and running everything and actually finding a second player who can do that (at a high level) is tough.

And for the warriors specifically, while Steph is obviously part of giving the warriors easy shots, I think the reason it's easy to think the decision makers, specifically Draymond, are a huge part of it, is the fact that Steph has a better efficiency when Draymond is on the court, the exact opposite of what we expect from a defensive specialist and also not to be expected if the shooter and not the decision maker is creating most of the benefit.


So, when it comes to the scale of said truth, it would sure be nice if it were tracked somewhere passes-to turnovers.

Closest relevant thing I can thin of is actually Databallr's oTOV value data. Obviously there's the rub that if the team keeps playing with the same level of playmaking recklessness with or without a player, that data is going to look very different from if they change things up when he goes to the bench. Arguably the ideal way to play with a player who requires dangerous passes-to is to go from "hare" mode to "turtoise" mode when he rests, but of course there are complications to switching schemes dramatically with routine substitutions.

If we look at the classic obvious dangerous pass-to avatar of Shaq, what we see is this:

In the earliest 4-year study nbarapm, Shaq rates as a slight positive +0.2 from '00-03, and after that he goes negative, and continues to get more negative mostly unabated for the rest of his career. He's a -0.3 from '01-04, and -1.9 from '08-11. Worth noting that in 2001 they removed the Illegal Defense rule which was artificially making entry passes to the low post easier.

With Curry we see him generally as having positive oTOV indicators. He peaks from '13-16 with a +1.0, which is interesting because that includes half Jackson years and half Kerr years. He doesn't go negative until '20-23 with a -0.5, which then goes to his nadir of -0.6 in '21-24.

By contrast if we look at the classic positive case here, Chris Paul, we see him north of +1.0 his whole career, peaking at +2.6 from '08-11.

A foil to Paul is Steve Nash who also had absolutely top tier passing ability but generally played a considerably more dangerous (for TOs) style of play. He's also positive the whole time, and peaks at +1.8 from '03-06, which splits two years in Dallas and two years in Phoenix.

Back to the Warriors, something interesting with Curry's signature assister Draymond: He's mostly negative by oTOV value. He peaks at +0.2 from '15-18 (Kerr's first 4 years), but of his career he posts a value worse than -0.5.

I would chalk that up to Dray being more aggressive in his passing than his teammates, which I'd generally consider to be a good thing, but of course, it would be a better thing if he could do it while reducing turnovers too.

For contrast there, the king of maximum-passing is Jokic, who is also a big who volume scores. I'd expect there's at least as heavy of a "tax" that Jokic is paying for this by oTOV value compared to Dray. Jokic's numbers are pretty close to zero here, but I will say they've generally been slightly positive, running only slightly negative in '19-22 & '20-23.


Ben Taylor had a video breaking down his thoughts on the value add of risky passes, I can't recall the name to pull it up right now. He kinda shows that there's a value to risk to a point and a cost of being conservative (CP3) at a point too. Trying to pull out stats on assists is just difficult...much like why this zTS% sounds good but it's going to create as much noise imo as it'll resolve. That isn't to say we drop it. Just like it's useful to track these oTOV or to look at turnover to pass rates...but I'm not sure they'll be conclusive.
User avatar
FrodoBaggins
Veteran
Posts: 2,884
And1: 4,635
Joined: Dec 25, 2013

Re: Databallr Releases a New Stat: zTS% 

Post#116 » by FrodoBaggins » Thu Apr 30, 2026 5:49 pm

zTS% isn't trying to compete with or replace one-number/all-in-one impact metrics like EPM, LEBRON, XRAPM, and DARKO. It's also not in league with adjusted +/- flavors (APM, RAPM) or their subcomponents (6Factor). Not box-score derived advanced statistics like Win Shares, BPM, or PER either. It's a shooting efficiency statistic, in line with FG%, eFG%, TS%, and rTS%/TS+. That's what it should be compared to.

Assists, passing, and playmaking are as relevant to it as they are to true shooting percentage.
dhsilv2
RealGM
Posts: 52,901
And1: 28,597
Joined: Oct 04, 2015

Re: Databallr Releases a New Stat: zTS% 

Post#117 » by dhsilv2 » Thu Apr 30, 2026 6:08 pm

FrodoBaggins wrote:zTS% isn't trying to compete with or replace one-number/all-in-one impact metrics like EPM, LEBRON, XRAPM, and DARKO. It's also not in league with adjusted +/- flavors (APM, RAPM) or their subcomponents (6Factor). Not box-score derived advanced statistics like Win Shares, BPM, or PER either. It's a shooting efficiency statistic, in line with FG%, eFG%, TS%, and rTS%/TS+. That's what it should be compared to.

Assists, passing, and playmaking are as relevant to it as they are to true shooting percentage.


Ironically, I think the actual likely end/best use for zTS% is going to be mixing it into some all in one box metric like PER where it could replace TS% and change up how it leans on usage. Time and far smarter people than I will have to figure that out. But I see that as a long term place for it. That said...I do look forward to see when player tracking data (play type here) actually is functionally incorporated. So far things like RAPTOR I felt did a poor if not terrible job with that (not that RAPTOR was terrible just the integration side with on ball defensive metrics).
User avatar
FrodoBaggins
Veteran
Posts: 2,884
And1: 4,635
Joined: Dec 25, 2013

Re: Databallr Releases a New Stat: zTS% 

Post#118 » by FrodoBaggins » Thu Apr 30, 2026 6:14 pm

dhsilv2 wrote:
FrodoBaggins wrote:zTS% isn't trying to compete with or replace one-number/all-in-one impact metrics like EPM, LEBRON, XRAPM, and DARKO. It's also not in league with adjusted +/- flavors (APM, RAPM) or their subcomponents (6Factor). Not box-score derived advanced statistics like Win Shares, BPM, or PER either. It's a shooting efficiency statistic, in line with FG%, eFG%, TS%, and rTS%/TS+. That's what it should be compared to.

Assists, passing, and playmaking are as relevant to it as they are to true shooting percentage.


Ironically, I think the actual likely end/best use for zTS% is going to be mixing it into some all in one box metric like PER where it could replace TS% and change up how it leans on usage. Time and far smarter people than I will have to figure that out. But I see that as a long term place for it. That said...I do look forward to see when player tracking data (play type here) actually is functionally incorporated. So far things like RAPTOR I felt did a poor if not terrible job with that (not that RAPTOR was terrible just the integration side with on ball defensive metrics).

It's an interesting thought. PER, BPM, and the like use position-based weighting, which I believe was determined based on RAPM regressions.
dhsilv2
RealGM
Posts: 52,901
And1: 28,597
Joined: Oct 04, 2015

Re: Databallr Releases a New Stat: zTS% 

Post#119 » by dhsilv2 » Thu Apr 30, 2026 6:47 pm

FrodoBaggins wrote:
dhsilv2 wrote:
FrodoBaggins wrote:zTS% isn't trying to compete with or replace one-number/all-in-one impact metrics like EPM, LEBRON, XRAPM, and DARKO. It's also not in league with adjusted +/- flavors (APM, RAPM) or their subcomponents (6Factor). Not box-score derived advanced statistics like Win Shares, BPM, or PER either. It's a shooting efficiency statistic, in line with FG%, eFG%, TS%, and rTS%/TS+. That's what it should be compared to.

Assists, passing, and playmaking are as relevant to it as they are to true shooting percentage.


Ironically, I think the actual likely end/best use for zTS% is going to be mixing it into some all in one box metric like PER where it could replace TS% and change up how it leans on usage. Time and far smarter people than I will have to figure that out. But I see that as a long term place for it. That said...I do look forward to see when player tracking data (play type here) actually is functionally incorporated. So far things like RAPTOR I felt did a poor if not terrible job with that (not that RAPTOR was terrible just the integration side with on ball defensive metrics).

It's an interesting thought. PER, BPM, and the like use position-based weighting, which I believe was determined based on RAPM regressions.


PER only used positions when Hollinger converted PER into his value add metrics which are mostly gone outside of I think ESPN might still just have them. But yes BPM does based on the regression to RAPM. The idea(s) with PER was that John thought TS% was highly important but he also believed that a coach would have say a Kobe who had a perfectly OK TS% take extra shots because coaches know more than a "spread sheet" if you will so he created a huge bias for usage% which I've always questioned on guys once their TS% dropped too much (not kobe but perhaps a melo in a bad year). By putting zTS% in vs TS%, you'd effectively be accounting for the "coaching decision". You'd still have the black box that is...is the player taking bad shots within a play type that still are low value or forcing shots in a tough play type where passing and resetting is better. In my mind this could be solved to a degree with some kind of formula looking for outliers where zTS is too much greater than TS. Similarly, when TS% is too much bigger than zTS, I think we'll have found an exceptional player within an "easy" play type. Think elite lob threat vs a guy who's just a lob threat. The first guy is really valuable. The second guy could very easily be replaced with little loss in TS% (where zTS would be better). Meanwhile zTS% might over correct for the first guy who's not just creating elite results but may also create elite volume (this again is where perhaps we can get a PER type metric?).
User avatar
SkyHook
Head Coach
Posts: 6,875
And1: 5,475
Joined: Jun 24, 2002
 

Re: Databallr Releases a New Stat: zTS% 

Post#120 » by SkyHook » Thu Apr 30, 2026 7:39 pm

I'm not sure that I buy the framing of the "hard playtype" — which, to me, sounds as if it's meant to be something noble — rather than simply referring to it as a "less efficient playtype". There are certainly players who prefer an iso-focused play style from an aesthetic standpoint, often to the detriment of their team and teammates. Is someone being marginally more effective than average at that inefficient playtype more valuable than someone whose focus is on a far higher efficiency playtype while perhaps being marginally worse than average at it? I don't know the answer, just asking the question.

Doctor MJ wrote:Now, one interpretation of that is essentially that the playmakers are acting as the sin-eaters for the shot takers, and there's some truth in that, but if the competitive advantage is having great shooters we should take care not overrate a playmaker's general capacity simply because he has outstanding targets any more than we'd want to underrate the playmaker because he plays with bad shooters.

Interesting from a philosophical perspective. I'll have to give this some thought.
"When the mob and the press and the whole world tell you to move, your job is to plant yourself like a tree beside the river of truth, and tell the whole world...

... NO, YOU MOVE."

Return to The General Board