What defensive player stats do you like/use?

[otherdave]

Since it is August...

Was just wondering what defensive stats you prefer when evaluating a player?


Why do you like this stat?


And where can the rest of us find this stat?

Thanks in advance.

[Ilikesports17]

I google DRPM, then I head over to bball ref and look at all the other advanced and standard defensive stats.

I dont trust any of them individually. I like to look at all of them and see if they all point in a cerain direction. If I only have a minute I look at Drtg.

[hwangjini_1]

i am interested in reading what is posted here. thanks for the thread. tp.

i have seen so many different posters post so many different stats as important, while simultaneously criticizing the defensive stats another poster posted. it gets confusing.

from what i can tell, there is no single defensive stat that is reliable for telling more than what that particular defensive stat measures.

i see lots of number, but not much in terms of a convincing argument.

i would welcome a battery of stats that helps us understand a player's defensive abilities.

go to it everyone, i would like to learn. 

[kozlodoev]

Unless there's something that makes use of the Spor**** Player Tracking data that I'm not aware of, I wouldn't trust any metric to provide conclusive individual defensive stats. At best, I occasionally use DPER as a rough guide for the productivity of players that are (supposedly) directly matched with a player of interest. But that's a pretty rough guide.

[ThePaintedArea]

Was just wondering what defensive stats you prefer when evaluating a player?

Interested in a wide range and not comfortable using one measure. But DRPM (Defensive Real Plus-Minus) is a key one, and I'd hesitate to make any kind of broad claim about a player that wasn't consistent with it.

Why do you like this stat?

It's an access to the whole defensive picture for a player, not just when he's on the ball. It accounts for how good the actual opponents are on the floor, as well as how good his teammates are. Thus it solves the problem that raw plus-minus poses.

And where can the rest of us find this stat?

http://www.espn.com/nba/statistics/rpm/_/sort/DRPM

The methodology is easily accessible, but the sorting and search options are limited.

[ViolentGhandi]

by only looking at the stats or certain stats one can easily miss a lot of information in order to avaluate a player - imo you have to take apart a game, look at every subsequence then cataloge & rate it.

Its the more scientific version of what you do while just watching the game at regular speed. Stats just tell you the pure numbers but don't take into account a lot of things you see while watching a game of basketball.

I will do an example of what I'm refering to. Imagine Avery picking up Harden on D. He manages to stay in front of him then fights over a pick just in time to cut the angle enough to stop him from completing a straight line drive to the basket. Instead Harden (who is easy to defend as there are no passing lanes the defender has to worry about) is picking up his dribble for a floater from, I don'T know maybe 11 feet away. The shot hits the rim, the backboard and finally gets rebounded by Al but a late foul call from the young ref, Harden had allready worked on, awards him 2 shots & gets AB into early foul trouble.

Stat wise you will have ABs playing time he accumulated during this sequence, 1 foul and thats about it. While watchin the game you get the impression that AB did pretty good in that sequence & that the foul call was a bad call. Let's go one step further. Imagine you are BS (I know he looks like he still isnt allowed to buy beer - remember its a example). As BS you look at the video footage to analyze the last game. So now you come to that scene and you analyze it. AB picks up Harden and as Harden uses the pick he (AB) has a certain amount of possible things he can do. He could just give up (yeah that Harden style D), he could run into the pick and lose Harden, he could fight thru the pick or pass behind the pick. Each Option you can think of you assign a number/mark to (sorted from best to worst option). This allows you to evaluate your players performance from a single event up to his whole carrier. The mark I give on such an event could further be affected by the opposing player (defending Harden is not the same as defending Perkins). AB beeing called for a foul he didnt commit affects the mark as well as the game situation. Imagine C's have a foul to give and lead with 1 point in the 4th and 5sec left to play. Again you would rate ABs play differently. This very time & work consuming approach allows you to come up with stats that go far beyond the +/--stat that everyone eigther is hating or loving. I don't say the +/- is completely useless or its the holy grail of stats. For me its just not giving me the amount of detail to make me a complete picture of a players performance.

Also if I want to I can create a biased picture by selecting a only special subset of stats. That allows for trolling, hating, calling players a bust and so on as well as unjustified praise. Citing stats doesnt proof my opinion right - it merly proofs I can read & repeat.

So whats my fav D stat? Its none in particular but a huge collection I can filter & check

[greece66]

larbrd33 had opened a similar thread some time ago - there was some good discussion there

also, I doubt this topic belongs to Celtics Talk - maybe Around the NBA?

[Monkhouse]

I honestly just go by my gut, despite statistics. You can't really put a stat to how hard a player can play, because defense is well, by effort 50% chance of it.

You either play to win defensively locked in mentally during a game, with the intent to shut down the man, and just make it tough for the offensive player to score.

I've seen a lot of defenders that are excellent defenders despite the statistics, and when they put in the effort, and how they excel in the playoffs or during a tough series.

[Denis998]

This right here is my favorite stat.
http://stats.nba.com/defensivehub/#!/tracking/player/?sort=PCT_PLUSMINUS&dir=-1

[Snakehead]

I honestly just go by my gut, despite statistics. You can't really put a stat to how hard a player can play, because defense is well, by effort 50% chance of it.

You either play to win defensively locked in mentally during a game, with the intent to shut down the man, and just make it tough for the offensive player to score.

I've seen a lot of defenders that are excellent defenders despite the statistics, and when they put in the effort, and how they excel in the playoffs or during a tough series.

The eye test still matters on defense to me too.  It's not that I don't look at stats but they don't tell it all.  Steals I think for example are often not really an indicator of the best defense being played especially on the perimeter.  And a lot of the bigs that really go after blocked shots can take themselves out of position doing so.  It's pretty complicated.

When I watch Smart for example I just think he's one of the very best defenders in the league already with his positioning, how he gets steals without gambling and being out of position, how he strikes at opportune times, his ability to get through screens, his hustle, his physicality and I don't think a stat can get all that.

[jacigar]

I use this formula: DRB+STL + BLKS =TOTAL X36 Divided by MPG + DRPG
 example Harford is 5.5 drb + .83 Stl + 1.48 Blls=7.81 tot x36= 281.16 div by 32.2 mpg = 8.76
+ drpg 2.08 = 10.84
Noel is 5.8+1.76+1.49= 9.05x36=325.8 div by 29.3=11.12+drep 2.21= 13.33

[jacigar]

This formula is to determine the defensive potential of a player over 36 mpg. actual poduction is this number divided by 36 x mpg. that is the actual def pg production. Other examples
Cousins 19.4 D.Jordan 24.18 C.Aldridge 18.83 Mpg Cousins 26.9,DJ 33.7 Aldridge 13.3
per game production Cousins 14.5 ,DJ 22.6,Aldridge 7.0 ,Harford 9.7, Noel 10.9
Kelly 9.67 production 5.4

[loco_91]

DRPM is probably the least bad, but defensive stats just aren't very good. Consensus reputation is better. That's in stark contrast to offensive stats, which are significantly better than league consensus imo. Another thing to look at is team defense.

[Roy H.]

I don't buy into any of the defensive formulas.  I mean, how can you trust a stat like DRPM, which considers a player's historic production, age and physical attributes like height in its formula?

I like the synergy based stats. Looking at how much a player holds opponents below their averages, or who allows the lowest percentage close to the basket, are interesting to me.

[ThePaintedArea]

I mean, how can you trust a stat like DRPM, which considers a player's historic production, age and physical attributes like height in its formula?

Trust? You don't need it, as long as you're clear what's being measured. The map is not the territory; a defensive stat is not defense.

As for historic production: why would you not want a bigger picture, a larger sample size for a player? Things change every day at that level; they get fatigue and nagging injuries, play hurt and play while recovering; and stamina and foot speed fluctuate. They get too low after a loss - or too high after a win.

As for age: it's a good proxy for experience, and on the other end, for athleticism and health.

As for height, it's simply harder to shoot or pass around a taller player; height is a good proxy for size more generally, and even for strength - and those are assets for defense.

Those measures are the kinds of things that are included in actuarial assessments, or, to pick a current example, Nate Silver's 'Polls-Plus' estimation of presidential candidates' chances. They are part, in other words, of a more nuanced statistical assessment.

The more relevant question is: should those things be weighted more or less?

I like the synergy based stats. Looking at how much a player holds opponents below their averages, or who allows the lowest percentage close to the basket, are interesting to me.

Yes, very interesting. Teams are getting some proprietary data from them; when all that eventually becomes generally available we may hope for more sophisticated discussions.

I see two big problems with people rejecting statistics out of hand, or only going by the 'eye test':

1. Most fans are just watching the ball. But any coach will tell you that team defense is way more important than individual defense; so most people only see individual defense, and mostly only when there's a breakdown for one or the other player. So most fans are missing the big picture about defense. To paraphrase the late great Chuck Daly, defense is spacing, and spacing defense.

2. Frankly, there's a big 'echo chamber' effect, where fans repeat whatever they've heard.

This doesn't stop people, on occasion, from being convinced that they're right, or even dismissing contrary arguments.

In the end, there's just no substitute for rewinding and looking at the whole play again.