Drose is an ELITE defender~~~Synergy

[kyrv]

Thanks very much for the effort, D.S., and your English is better than most who were born here.

by a longshot

And then some.


Thanks fish and transplant and stern and all.

The defensive rating listed on basketball-reference appears for most players to seem accurate and match what most would say. However it seemed skewed based on the team defense (which we now know, it is).

Kirk with a 105. That's not that bad, if you say he was guarding the best guard every single game, is 105 still bad? Kirk's numbers go up and down, and he didin't always defend well, but, are those moves more a reflection of the team defense? That I don't know.

[coldfish]

Thanks very much for the effort, D.S., and your English is better than most who were born here.

by a longshot

And then some.


Thanks fish and transplant and stern and all.

The defensive rating listed on basketball-reference appears for most players to seem accurate and match what most would say. However it seemed skewed based on the team defense (which we now know, it is).

Kirk with a 105. That's not that bad, if you say he was guarding the best guard every single game, is 105 still bad? Kirk's numbers go up and down, and he didin't always defend well, but, are those moves more a reflection of the team defense? That I don't know.

As others have said, the number is basically overall team defensive ability +/- individual rebounds and steals. Kirk has generally played on good defenses. If you look at each year, it says that when Kirk was on the Bulls, he was one of the worst defenders on a good defensive team (much like Rose this year).

Here were Hinrich's defensive rankings on the Bulls:
03/04: 16th of 21 players
04/05: 6th of 15 players
05/06: 9th of 19 players
06/07: 11th of 14 players
07/08: 14th of 19 players
08/09: 11th of 19 players
09/10: 11th of 18 players

Derrick Rose:
10/11: 10th of 14 players (behind such luminaries as Brian Scalabrine)

[kyrv]

Great research.

is this an incredibly flawed rating or a rating that just should not be used at all?

Question to myself: Why are you quoting numbers before you really know what they represent?

[League Circles]

I don't see how anyone can take 82 games seriously, unless there is something I don't know. The reason I say this is that they have no advertising, and no real "About Us" area. They don't really explain much of anything about themselves or about how they can (supposedly) put so much (implied) effort digging up and computing these numbers. For example, I've always thought their positional minutes played for players was pure crap.

The other stat sites like basketball reference (and synergy?) seem much more legit IMO.

[lorak]

is this an incredibly flawed rating or a rating that just should not be used at all?

It's flawed ranking (and Oliver admits it), however it have some value but we have to remember that Oliver's DRTG undervalues players who don't have box score stats but are good defenders (Bowen for example) and sometimes overvalues players who plays on good defensive teams.

[coldfish]

I don't see how anyone can take 82 games seriously, unless there is something I don't know. The reason I say this is that they have no advertising, and no real "About Us" area. They don't really explain much of anything about themselves or about how they can (supposedly) put so much (implied) effort digging up and computing these numbers. For example, I've always thought their positional minutes played for players was pure crap.

The other stat sites like basketball reference (and synergy?) seem much more legit IMO.

This isn't 82games.com, basketball reference actually has the stat.

[Susan]

Oliver's stats overvalue rebounding. The best offensive rebounders will typically be very high up on the offensive ratings.

[kyrv]

Oliver's stats overvalue rebounding. The best offensive rebounders will typically be very high up on the offensive ratings.

Somebody mentioned this about Noah ( maybe you?) - his O rating is off the charts.

[Susan]

Oliver's stats overvalue rebounding. The best offensive rebounders will typically be very high up on the offensive ratings.

Somebody mentioned this about Noah ( maybe you?) - his O rating is off the charts.

It really doesn't take into account usage rates, I don't think. I'm pretty sure it just assumes that each shot created is equal. High TS% players have abnormally high ORatings even if everybody else is setting up their offense for them. For reference, Tyson Chandler is leading the NBA in ORating. Matt Bonner is 3rd and Turiaf is 5th.

[Ugly]

I go by the eye test, not stats. And from what I've seen, Derrick is a MUCH better defender. He's disrupting passes, blocking shots (which is crazy for a PG), contesting shots, staying on his man, etc. etc. It's very easy to notice should any of these national writers actually watch a game and not just base their columns on what he was before this year.

Can someone confirm that our friend here's avatar prompts a login popup to "http://home.comcast.net". It can be loaded in isolation in his user profile at..

http://forums.realgm.com/boards/memberlist.php?mode=viewprofile&u=9797

Hosting the image elsewhere, or eliminating the comcast.net redirect (whichever method is currently being used), will stop other members from getting login popups as the page loads. Public PMs appear to be disabled for this account, or I would keep it private. Sorry.

Ugly™

[BrooklynBulls]

I removed his avatar, Ugly.

[transplant]

As others have said, the number is basically overall team defensive ability +/- individual rebounds and steals. Kirk has generally played on good defenses. If you look at each year, it says that when Kirk was on the Bulls, he was one of the worst defenders on a good defensive team (much like Rose this year).

I said earlier that any stat that can't be reasonably understood is a bad stat.

Thanks to all. I now have a better understanding...and I will never use this stat.

[Susan]

As others have said, the number is basically overall team defensive ability +/- individual rebounds and steals. Kirk has generally played on good defenses. If you look at each year, it says that when Kirk was on the Bulls, he was one of the worst defenders on a good defensive team (much like Rose this year).

I said earlier that any stat that can't be reasonably understood is a bad stat.

Thanks to all. I now have a better understanding...and I will never use this stat.

I think that's horribly misguided. I'm not trying to say that this is a great stat but Dean Oliver's the man who started the whole basketball stat revolution.

[dice]

I removed his avatar, Ugly.

that's not very nice. and you're a mod?

[Susan]

http://bkref.com/tiny/Asd47

ORating and TS% are pretty closely related.

[coldfish]

As others have said, the number is basically overall team defensive ability +/- individual rebounds and steals. Kirk has generally played on good defenses. If you look at each year, it says that when Kirk was on the Bulls, he was one of the worst defenders on a good defensive team (much like Rose this year).

I said earlier that any stat that can't be reasonably understood is a bad stat.

Thanks to all. I now have a better understanding...and I will never use this stat.

I think that's horribly misguided. I'm not trying to say that this is a great stat but Dean Oliver's the man who started the whole basketball stat revolution.

In general, I look for some correlation with stats (as most people do). For example, with PER, I tend to agree with it 80% of the time or more. If PER says this guy is a 25 and this guy is a 20, in almost all cases, I would say that the 25 is the better player using the eye test.

In going through the DRTNG thing, you can't do that. There almost negative correlation. You can't compare across teams because the team defense is too much of a variable and you can't compare big men to perimeter players due to the overvaluing of rebounding. Basically, its a very limited stat and virtually no one uses it within those limitations.

Fun stat of the minute, DRTNG
Boozer 2010/11 = 97
Michael Jeffrey Jordan the year he won the DPOY = 101
Scottie Pippen's best year = 97

So, yes, Carlos Boozer is a better defender than Michael Jordan at his peak and equal to Pippen's best efforts.

[BrooklynBulls]

I removed his avatar, Ugly.

that's not very nice. and you're a mod?

I try to get out but they keep pulling me back in.

[Susan]

In general, I look for some correlation with stats (as most people do). For example, with PER, I tend to agree with it 80% of the time or more. If PER says this guy is a 25 and this guy is a 20, in almost all cases, I would say that the 25 is the better player using the eye test.

In going through the DRTNG thing, you can't do that. There almost negative correlation. You can't compare across teams because the team defense is too much of a variable and you can't compare big men to perimeter players due to the overvaluing of rebounding. Basically, its a very limited stat and virtually no one uses it within those limitations.

Fun stat of the minute, DRTNG
Boozer 2010/11 = 97
Michael Jeffrey Jordan the year he won the DPOY = 101
Scottie Pippen's best year = 97

So, yes, Carlos Boozer is a better defender than Michael Jordan at his peak and equal to Pippen's best efforts.

Oh, I agree but my issue was with this quote.

I said earlier that any stat that can't be reasonably understood is a bad stat.

[Rerisen]

Oliver's Defensive Rating is quite complicated, and involves a lot of assumptions he freely admits. It's hard to explain as well, but for those interested might want to take a look into his book, Basketball on Paper. Here is a link to some relevant information about what goes into the rating (start at page 198)

Basketball on Paper

Generally, DRTG seems to work decently when looking at it within teams (a bit like +/- in this regard). You wouldn't want to compare a bad defensive player on the Celtics say to a good defensive player on a poor defensive team using it. Because every player on a good defensive team is getting a lot of individual credit for the great team defense of his teammates.

From observing this stat over the years, it tends to be harder on perimeter players. One possible reason is that they are likely to be in the game when teams go small and are purposely giving up size and defense to get more ballhandling and scoring on the floor.

Players that get a lot of steals and blocks (hello Tyrus) may tend to be overrated by this stat.

The paragraph about Dumars and Jordan is also informative and might go a ways to explaining someone like Deng's poor team rating here:

[transplant]

As others have said, the number is basically overall team defensive ability +/- individual rebounds and steals. Kirk has generally played on good defenses. If you look at each year, it says that when Kirk was on the Bulls, he was one of the worst defenders on a good defensive team (much like Rose this year).

I said earlier that any stat that can't be reasonably understood is a bad stat.

Thanks to all. I now have a better understanding...and I will never use this stat.

I think that's horribly misguided. I'm not trying to say that this is a great stat but Dean Oliver's the man who started the whole basketball stat revolution.

Misguided? Did I say anything negative about Mr. Oliver beyond that I don't like the Drtg stat (now that I understand it better)?

I'm a big fan of the per possession stats generally, so Mr. Oliver is OK in my book.

Edit: My criticism and vow not to use the Drtg stat only applies to its use for individual players. Love it for teams.