Harden has a better defensive RPM than Kawhi this year

Insecure fans
Might as well say Clint Capela is a great three point shooter

The only important thing about harden is that he makes the rockets legit contenders

 

Weren’t teams using our switching schemes against us to draw Capela away from the basket?

The team has been trying to address that, but maybe Capela’s defensive statistics are still sea-anchored by the sputtering start to the season.

 

All true, but in the last minute of an important game, I'm just about as comfortable with Harden guarding on the perimeter as I am with Kawhi. And I'm MORE comfortable with Harden guarding someone in the paint.

When Harden knows it's time to lock in, he's more than capable.

 

It’s just a horrible stat. All individual defensive box score stats are.

I mean, Kawhi is a 2016 DPOY, and a guy who won Finals MVP for he defense of Lebron.

But not only do fans continue referencing this stat ... like this thread ... but we spend a lot of time defending their absolutely Gross errors. It’s so weird, we defend a stat no professional coach, GM or scout would ever recommend using.

This isn’t even the first year defense stats got Kawhi wrong. Try his last full year, too. There was a thread in the Dish about this very same thing.

Kawhi Leonard is so great at D, he's actually hurting the Spurs

The entire premise of that article is some guy at CBSsports looking at a horrible stat, and deciding to use 3 gifs of Kawhi guarding Butler to try to make sense of what the defensive guru Popovich is doing with his DPOY, and saying it is bad defense,,,yet he opens the article saying there is nothing wrong with their record or team defensive rating.

At the beginning of the article, just like OP here, he admits these individual defensive stats are not reliable, too. But he chooses them over Pops here.

 

Numbers NEVER lie.

If the numbers don't support your programming, reject the programming not the numbers.

 

Rockets need to get 1st seed, otherwise they don't count harden

 

In the case of individual defensive stats, we do need to reject the programming (formulas), because the numbers are so horribly off with the eye test.

No professional coach, GM or scout would ever recommend using DRPM to compare players, but fans feel some weird need to not only defend the Numbers, but also the programming that spat out that Kawhi is a negative defender. Lulz

 

Everything on the line, I'm choosing Kawhi. :-D

BUT...

...absolutely. When Harden gets isolated and eyes are on him. He can D up. That D he played on Kyrie on his step back three (even if Kryie buried the shot) was fantastic.

 

Something wrong with those stats. They say they get them from Basketball Reference but when I look at the adjusted defensive +/-

https://www.basketball-reference.com/leagues/NBA_2019_advanced.html

Klay is actually 331

 

Kawhi's DRPM is not negative: https://www.basketball-reference.com/leagues/NBA_2018_advanced.html

I think there is a difference between what ESPN is calculating and DRPM

Also, the problem with these +/- scores is that a lot of it depends on the players you tend to be on the court with. It may be that other players on the raptors have a poor defensive game and that's dragging his down.

 

But, I contend it is *not* a "summary of a player's performance over a period of time". I contend individual defensive numbers aren't useful "stats" just because it is the result of a convoluted formula run on questionable data -- and we say, "Well, this is kinda the best we can do."

No coach or scout would use boxscore data to measure individual defensive performance.

The stat only 5yrs of rankings, so you can't really say this. There is actual subjectivity built into the math to do what you say.

We don't need a stat to tell us who's best, and if you are using the rather large margin of error in the stat to make your final vote of 1st/2nd Team Defense, then that's a problem. -- ie, errors in the formula can become self-perpetuating, and get larger.

Any correlation you see -- when we have very Gross errors (bad players way too high, good players way too low, complete nonsense for PG/SGs) -- is because key human-created coefficients in the formula are tweaked to match Eye Test, but then a lot of random stuff gets jumbled to create useless rankings. If left unchanged over time, if not continuously tweaked to match Eye Test (which is subjectivity), the stat can get completely detached from usefulness like PER did -- which I contend, it already is, and always has been.

Thus, at best, it is a formula tweaked to match past subjective votes to rank current. And since we only have 5yrs to view its accuracy, well ...

 

It's not from Basketball Reference. They just get their PbP from BRef. It's a different stat. DBPM is not the same formula as DRPM.

Yes it is.

It's a different formula. And Basketball Reference's DBPM is probably even a worse ranking.

 

The professionals in this business who matter, coaches, scouts and GMs don't use these stats because they are useless rankings.

But, have at it. I mean, I could argue NBA2K ratings are better. Why never talk about those?

consider this: Sophisticated Football fans have far better approach to generating metrics for linemen and defenders. It's all EYE TEST per play...but tallied up. That's the best way. But NBA is not popular enough to quench our thirst for good scout-based defensive data, like the NFL has. So we run PbP data, and geeks tweak their coefficients to match past subjective history ... and we just shrug our shoulders to say, "Well, best we can do."​

NFL fans have a much better way, and so do NBA scouts.

 

Austin Rivers is one of the worst in DPRM, I think he put the nail in the coffin on that stat for me personally as he's been very impressive on defense.

 

hey, @Bobbythegreat, don't you follow some NFL stat for linemen and defenders where some amateur (if not professional) scouts rate each play on how players perform their assignments, then they apply metrics using these per play scores over time. Can you address how that is done?

 

Probably what you are looking for is the Pro Football Focus stats, they are by far the most respected when it comes to these types of evaluations....even if I don't always agree with them 100% of the time.

From here, I wrote a long reply that wasn't satisfactory when I re-read it before posting and then I remembered that I read an article on it so I'd rather let you read their own words about it than me doing a poor imitation of it.

https://www.profootballfocus.com/news/pro-how-we-grade-offensive-and-defensive-linemen

 

Thanks for article. Yeah, that's what I've seen you Texans fans talk about. The article doesn't really elaborate on who is scoring player performance on each play. Can you address that better?

Sounds like many "scouts" are contributing for each play, to balance subjective differences in identifying the assignments. Who are these "scouts?" And do the GMs use this data? I read where someone is trying to re-package such per-play metrics to help GMs make trade/salary evaluations, too? Is that from PFF?