NBA introduces "hustle stats" using SportVu data for playoffs.

[bringWallyBack]

What are some of these stats?

It will be a single combined "hustle" number, like PER or something, based on the following elements.

According to the NBA, the following will be counted through a combination of SportVU tracking and video analysis performed at the Secaucus replay center:

Deflections – Defensive player touches the ball on a non-shot attempt
Loose Balls Recovered – Player gains sole possession of a live ball not in the control of either team
Contested 2’s – Defensive player closes out and raises a hand to contest a 2 point shot prior to its release
Contested 3’s – Defensive player closes out and raises a hand to contest a 3 point shot prior to its release
Charges Drawn – Defensive player draws a charge
Screen Assists – Offensive player sets a screen for a teammate that directly leads to a made field goal by that teammate

The first five were the items tracked at Summer League, while “Screen Assists” are a new category.

Eh, lots of flaws in those. How well will "contested" be tracked? A lot of players just "contest" but not in a way that is enough to possibly disturb the shooter--it's more for show. Also, players who guard players that shoot more often will have more "contested" 2's/3's, but that doesn't mean they hustle more than that they guard volume shooters. Which, though valuable to know, would be better captured by things like defensive rtg and eFG% against. Screen assists...well if you're screening for Curry you'll have a lot of field goals off your screens. If you're screening for a lot of other players...not so much. All of these could be addressed if the stats are normalized in some way, but it doesn't look like they plan to.

I remember some years ago reading an article about UCLA having some sort of way tracking these types of hustling plays. They were able to really see who was making the effort in a defensive play rather than just a lazy hands up.

[brownsmith89]

lou amundson might be one of those players who hustles a lot without putting up stats.

[young_frogger]

Eh, lots of flaws in those. How well will "contested" be tracked? A lot of players just "contest" but not in a way that is enough to possibly disturb the shooter--it's more for show. Also, players who guard players that shoot more often will have more "contested" 2's/3's, but that doesn't mean they hustle more than that they guard volume shooters. Which, though valuable to know, would be better captured by things like defensive rtg and eFG% against. Screen assists...well if you're screening for Curry you'll have a lot of field goals off your screens. If you're screening for a lot of other players...not so much. All of these could be addressed if the stats are normalized in some way, but it doesn't look like they plan to.

Also 'loose balls recovered' sounds super subjective. If a loose ball happens to roll in your direction and you grab it, you didn't necessarily 'hustle' for it IMO

Just like rebounds? Some are of great use for your team and some just fall in your hands with three team mates standing around as well....

Yeah but at least rebounds aren't subjective (although I'm still not entirely sure how 'team rebounds' work, if someone could explain that to me that'd be great). It's like, how do we even define a ball being 'loose'? What if the dude throws a bad pass which gets deflected but then he manages to get it back? He'd get credit for hustling even though he almost turned the ball over.

[King Ken]

Al HORFORD is about to be a MVP

Sent from my SM-N920V using Tapatalk

[stephanieg]

They should introduce an anti-hustle stat, like when a player complains to the ref while their man goes down the court and scores. Call it a Wade.

[Zubby]

I hope Daryl Moray doesn't come across this.

Beverley type players are going to get paid.

[mtron929]

Well, it is just confusing. The most commonly known stats are all quantities that can be readily observed while being quantifiable from watching the game. Who scored the points. Who provided the assists. Who got the rebounds. Who got the steals. Who blocked a shot.

ON the other hand, who hustled how much amount in that play just becomes something that cannot be readily tracked by a viewer and thus will not be something that will be taken seriously by the mainstream fans.

[EmperorLocky]

[homecourtloss]

Will be interesting to see these stats for an entire season and then lol at DRAPM stats.

[Novocaine]

It will be a single combined "hustle" number, like PER or something, based on the following elements.

Eh, lots of flaws in those. How well will "contested" be tracked? A lot of players just "contest" but not in a way that is enough to possibly disturb the shooter--it's more for show. Also, players who guard players that shoot more often will have more "contested" 2's/3's, but that doesn't mean they hustle more than that they guard volume shooters. Which, though valuable to know, would be better captured by things like defensive rtg and eFG% against. Screen assists...well if you're screening for Curry you'll have a lot of field goals off your screens. If you're screening for a lot of other players...not so much. All of these could be addressed if the stats are normalized in some way, but it doesn't look like they plan to.

I remember some years ago reading an article about UCLA having some sort of way tracking these types of hustling plays. They were able to really see who was making the effort in a defensive play rather than just a lazy hands up.

Virtually every team tracks them, including the shot contesting. Normally using tiers for level of shot "contestedness", eg: type I - contested shot with the defender rising up both arms outstretched, type II - both hands up, defender on the floor, type III - hand on the shooters face, etc. It's always been a big part of the job for scouts and video analysts/coordinators. Not on every single possession because that would take too much manpower and random representative samples are good enough.

I agree with narmerguy that a single contested/open statistic is useless and even misleading. Some shooters are bothered with even a half-hearted run on a close-out or a static big with his arms stretched under the rim; on others those things have basically no impact. But the degree that each shot is contested is hugely important.

[blau08]

Where are all these stats? I'm looking everywhere and I can't find it.

[Atmanne]

Where are all these stats? I'm looking everywhere and I can't find it.

http://stats.nba.com/hustle/#!/

[kodo]

This should be interesting. I wonder if this is something that the analytics departments of teams are already tracking. I bet the proprietary systems used by teams are even more fine-grained to capture the "intangibles."

Agreed. Coaching departments have been tracking this for many years, before the recent analytics revolution.

One of the jobs a lower tier assistant coach had in the past is to track these non-boxscore stats by hand during the game...I assume SportsVu is just making all that easier and they have access to league-wide data now.

[Pillendreher]

Just a quick overview:

Screen assists:

1.Biyombo 4.0
2.Bogut 4.0
3.Turner 3.0
4.Patterson 3.0
5.Crowder 3.0

Deflections:

1.Horford 6.0
2.Lowry 5.0
3.Felton 4.0
4.George 4.0
5.Ariza 4.0

Loose Balls recovered:

1.George 4.0
2.Turner 3.0
3.Horford 2.0
4.Curry 2.0
5.Felton 1.0

Charges drawn:

1.George 2.0
2.Crowder 1.0
3.Scola 1.0
4.Hardaway Jr. 1.0

Contested Shots:

1.Green 19.0
2.Horford 17.0
3.Turner 16.0
4.Ibaka 15.0
5.Milsap 13.0

[MystikSpiral]

I already don't even know what TS% and PER and other advanced stats mean, but hey, anything to help you avoid actually watching a game and just enjoying it, I guess...

[CodyB]

These teams are routinely paying $130+ million a year for players + coaching staff.

For another $5M, they can virtually have an army of statisticians going over every minor detail and identifying things that the eye won't see.

These "Advanced" advanced stat make a lot of sense. But I'm sure it's something that statisticians have been doing for a while anyway.

[wilt]

I know for a fact that there is a **** of data available to teams going still way beyond this, even in Europe (with much lower budgets). But that doesn´t mean it isn´t nice to have some of it getting available to us now and then

[abark]

Also 'loose balls recovered' sounds super subjective. If a loose ball happens to roll in your direction and you grab it, you didn't necessarily 'hustle' for it IMO

Just like rebounds? Some are of great use for your team and some just fall in your hands with three team mates standing around as well....

Yeah but at least rebounds aren't subjective (although I'm still not entirely sure how 'team rebounds' work, if someone could explain that to me that'd be great). It's like, how do we even define a ball being 'loose'? What if the dude throws a bad pass which gets deflected but then he manages to get it back? He'd get credit for hustling even though he almost turned the ball over.

Assist are subjective, but they still manage to reflect what they are intended too. A stat can require judgement calls and still be useful.

[Sixerscan]

Also 'loose balls recovered' sounds super subjective. If a loose ball happens to roll in your direction and you grab it, you didn't necessarily 'hustle' for it IMO

Just like rebounds? Some are of great use for your team and some just fall in your hands with three team mates standing around as well....

Yeah but at least rebounds aren't subjective (although I'm still not entirely sure how 'team rebounds' work, if someone could explain that to me that'd be great). It's like, how do we even define a ball being 'loose'? What if the dude throws a bad pass which gets deflected but then he manages to get it back? He'd get credit for hustling even though he almost turned the ball over.

I believe team rebounds are when the ball goes out of bounds off of a shot without anyone securing the rebound first.

[young_frogger]

Just like rebounds? Some are of great use for your team and some just fall in your hands with three team mates standing around as well....

Yeah but at least rebounds aren't subjective (although I'm still not entirely sure how 'team rebounds' work, if someone could explain that to me that'd be great). It's like, how do we even define a ball being 'loose'? What if the dude throws a bad pass which gets deflected but then he manages to get it back? He'd get credit for hustling even though he almost turned the ball over.

I believe team rebounds are when the ball goes out of bounds off of a shot without anyone securing the rebound first.

I see. Is there a distinction between offensive and defensive team rebounds? Meaning do offensive team rebounds count towards the team's total O-board count?

That could definitely be counted as a hustle stat then. Because a lot of the times putting pressure on someone going for a rebound makes him unable to corral the rebound properly, forcing him to tip it out of bounds, but they don't get credit for the board. It's kind of like when you force a guy into a bad pass or stepping out of bounds but you don't get credit for the steal.