Biggest regressions in opponent FG% from last yr: Thornton, Lawson, Howard, Brewer, and Jones

Here is the same list when sorted by GP >= 22, and DFGA >= 7

Puts into a bit of context just how good defensively Josh Smith was for us last year, when he allowed -4.6

 

Whats the definition for this stat?

 

Here is our team last year vs this year while defending 3 pointers. A shot is defended when I think the player is within 4 feet of the shooter.

A couple of things jump out, mainly that everyone is noticeably worse again.

Brewer goes from allowing 29% last year to like 42% this year, which is like a 13-14 point swing.

The only player doing well this year is really Thornton, no idea why he isn't getting more minutes.

 

Are you sure about these stats? I looked up one player on their NBA.com's Defense Dashboard, and it did not agree with your numbers at all. It says Corey Brewer gave up a 42.6% FG% last year (http://stats.nba.com/player/#!/201147/tracking/defense/?Season=2014-15&SeasonType=Regular Season). Your numbers really don't pass the reasonableness test, as they are so dang high.

 

I have looked at this stat a lot for a couple years now in an attempt to be able to quantify how good a player is on defense, and I have to point out it has some quirks. For example, everybody on here went on and on about Ariza's defense last year. This stat reports Ariza as giving up a very, very bad 4.1% higher FG% in 2014-2015, while Harden is recorded as a plus defender. You'll find weird things like that with this stat all the time, which makes me kind of question it.

 

Well heres the thing. A lot of good defense is not allowing your opponent to shoot in the first place. Like last year, Ariza allowed 9 FGA a game, while Harden is defending 12.

A lot goes into why this is. Opponents might think Ariza is a better defender so they avoid him. He could deny a good shot. Then you need to normalize the actual attempts to what the player usually attempts and adjust the FGA number. Harden could simply lose his man entirely, and then his bad defense got shifted on to someone else. Then you need to normalize for minutes. There needs to be a "deterrence index" for perimeter players.

There's a lot of context. This year, Kawhi Leonard allows 9.5 shots a game while usually always guarding the very best perimeter player and he does this limiting his opponents to -6.5. He gives up 3.5 makes a game...

Harden gives up almost 14 shots a game. and almost 7 makes.


The truth is last year, Harden played good defense for a long time, which is why we were in the top 10 in defense. Ariza just got a lot of the credit for it, but Harden pulled his weight and more.

but The point of this isn't to say Harden is a better defender than anyone else, It's to say 2014 Harden and Brewer were better defenders than 2015 versions of themselves, which is the case here.

 

Harden's defense last year was OK, lol.

 

Don't waste your time anymore. These stats are mostly worthless.
Don't take my word fine. Take his.

<blockquote class="twitter-tweet" lang="el"><p lang="en" dir="ltr">I don't think he's having an impact so much as he's there when they miss. <a href="https://t.co/V9P2WK4SYJ">https://t.co/V9P2WK4SYJ</a></p>&mdash; Seth Partnow (@SethPartnow) <a href="https://twitter.com/SethPartnow/status/676957428139839488">Δεκέμβριος 16, 2015</a></blockquote>
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>

<blockquote class="twitter-tweet" lang="el"><p lang="en" dir="ltr">Those individual &quot;shot defense&quot; stats on jumpers is almost wholly noise. <a href="https://t.co/Kahc7sYYwk">https://t.co/Kahc7sYYwk</a></p>&mdash; Seth Partnow (@SethPartnow) <a href="https://twitter.com/SethPartnow/status/676999966066417665">Δεκέμβριος 16, 2015</a></blockquote>
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>

<blockquote class="twitter-tweet" lang="el"><p lang="en" dir="ltr">Allocation of who is closest, yes. Credit/blame for &quot;forcing misses&quot; and &quot;allowing makes&quot; is far more complex. <a href="https://t.co/js0mvglfSf">https://t.co/js0mvglfSf</a></p>&mdash; Seth Partnow (@SethPartnow) <a href="https://twitter.com/SethPartnow/status/677120616709808128">Δεκέμβριος 16, 2015</a></blockquote>
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>

This guy is the founder of NylonCalculus the best public bball analyst (not in FO).

 

I agree that SportsVU-based defensive stats has a long way to go to produce much meaningful info for individual defense.

But when it says what you want it to about your team and theories, despite inexplicable outliers, people love it.

 

That's exactly what I use it for!!!! haha

 

Yeah, I tried to follow the path you gave, and I got to this page and all of the opponent FG%'s are different than these you gave. They are all in the 40's, as expected. You look at the DFG% column. There is no Opp FG%. I don't know where the disconnect was, but all those numbers are wrong. Just think about a typical NBA player's FG%. If our whole team was giving up >50% for FG's, we wouldn't even be at .500.

http://stats.nba.com/league/player/...6&SeasonType=Regular Season&TeamID=1610612745

 

With respect to defense, yes I will and I always will and no amount if blind homerism will change my mind. He is literally the worst defender in the entire National Basketball Association.

 

Rim protection stats are different than generic stats than these. RIM protection is actually pretty reliable because the shot location and who exactly is defending the fg and forcing or not the miss can be isolated.
BUT mostly when it comes to bigs. For the rest it doesn't matter almost at all (wings, perimeter players). Except if it is a far outlier which is remarkable (for example KJ McDaniels).

These stats are pretty much luck and noise. One bad defender who misses his man and isn't even close to defend,and his man passes by him and penetrates will show up as good or mediocre because he won't be counted as the primary defender.
One guy who is a bad defender and doesn't even put up his hands to contest a shot, so his team tries to hide him by putting him against a bad offensive player, will show as a good defender, if his man who is so bad at offence miss a shot (which he would miss anyway).

IF you want to have a small idea at least try to see the plays.
For example here is the Houston Rockets in pnr defensive roll man
http://stats.nba.com/playtype/#!/ro...er&OD=offensive&CF=TeamNameAbbreviation*E*hou

Howard: 60 percentile Capela 53.3 percentile

Here is defensive post ups
http://stats.nba.com/playtype/#!/post-up/?dir=1&PT=player&OD=offensive&CF=TeamNameAbbreviation*E*hou

ALWAYS look at the possesions and the freq. For example in his 7 games played thsi year DMO has defended post up 19 possesions (already more than Capela) and he is 8.1 percentile.
Howard is 15.6 percentile with a freq of 27.9.

Perimeter players:
PNR ball handler :
http://stats.nba.com/playtype/#!/ba...er&OD=offensive&CF=TeamNameAbbreviation*E*hou

Defensive iso
http://stats.nba.com/playtype/#!/ba...er&OD=offensive&CF=TeamNameAbbreviation*E*hou
Unfortunately for defensive transition the league has no stats.:grin:

Even those stats don't show well when a player lets someone blow by him for example, or that he has messed up a switch by not communicating and puts his teammate in a bad position which allows the attacker to score.

There are some articles from time to time from Vantage sports that have some stats.
They had one on Lawson where they had all the REALLY advanced stats (not publicly available by nba.com or nyloncalclulus or any other source ofc ) about his offence in defence.

Here is one more from the first half of last year where they talked about close outs.They say this about Harden's defence in the first half of last year :

http://www1.vantagesports.com/Articles/archive_article_view/VTzrGyYAACgAXrIe

AND this is what they said last month about Lawson and his fit with the Rockets thus far

This is what they said last week about the Rockets

http://www1.vantagesports.com/Articles/article_view/VmxVJSAAAB8AuSVZ

If anyone want to to sign up to VantageSports and have access to all the advanced stats like boxing out, reb chances, deflected pass rate etc etc and then let us know about some stats be my guest.

 

If you read the description at the top of your page, it clearly says at the rim. If I read correctly, you misrepresented the stats to mean shots from anywhere. Is there a way to completely delete this thread from the annals of Clutchfans history?

 

Why did we promote our defensive coordinator to head coach when defense is the problem.

 

Exactly, what I was thinking. Why are any of the staff left over from McHale's tenure still here?