Olynyk = DPOY?

[Moranis]

according to basketball reference the leaders are currently.  Boston has 3 guys in the top 20, none of them are KO, who must just not be qualified (i.e. not enough minutes).  KO has a DRTG of 92 according to that site as well, so not nearly as low.

Defensive Rating
1.   Hassan Whiteside ▪ MIA   86.2
2.   Kawhi Leonard ▪ SAS   91.0
3.   Derrick Favors ▪ UTA   91.0
4.   Jared Sullinger ▪ BOS   91.1
5.   Tim Duncan ▪ SAS   92.0
6.   Jae Crowder ▪ BOS   92.3
7.   Andre Drummond ▪ DET   92.5
8.   Pau Gasol ▪ CHI   93.0
9.   Chris Bosh ▪ MIA   93.2
10.   Rudy Gobert ▪ UTA   93.6
11.   Dewayne Dedmon ▪ ORL   94.3
12.   Draymond Green ▪ GSW   95.3
13.   Festus Ezeli ▪ GSW   95.5
14.   Stephen Curry ▪ GSW   95.7
15.   Tyson Chandler ▪ PHO   95.9
16.   LaMarcus Aldridge ▪ SAS   96.0
17.   Tristan Thompson ▪ CLE   96.2
18.   Amir Johnson ▪ BOS   96.3
19.   LeBron James ▪ CLE   96.4
20.   Trevor Booker ▪ UTA   96.4

[littleteapot]

This is a great example of how you can't just look at overall team stats when you're on the floor. Jared Sullinger is 4th in D rating?

[fairweatherfan]

This is a great example of how you can't just look at overall team stats when you're on the floor. Jared Sullinger is 4th in D rating?

Sure you can, you've just gotta account for the fact that there's a lot of noise and that the amount of noise will decrease over a larger sample size.  Right now the # of Celtics highly rated tells us that our defense is really good, and that the effects are mainly seen from the 3-5 position.  Presumably the frequent pairing of IT with our best backcourt defenders is blunting the impact on their D ratings. 

It doesn't mean Sully's the #4 defender in the league, just that the team has the 4th best defensive performance when he's on the floor, relative to every other player.  It's a nice sign for the team and for Sully, you just have to strike a balance between taking it as gospel and dismissing it as useless.

[littleteapot]

Sure you can, you've just gotta account for the fact that there's a lot of noise and that the amount of noise will decrease over a larger sample size.  Right now the # of Celtics highly rated tells us that our defense is really good, and that the effects are mainly seen from the 3-5 position.  Presumably the frequent pairing of IT with our best backcourt defenders is blunting the impact on their D ratings. 

It doesn't mean Sully's the #4 defender in the league, just that the team has the 4th best defensive performance when he's on the floor, relative to every other player.  It's a nice sign for the team and for Sully, you just have to strike a balance between taking it as gospel and dismissing it as useless.

I don't think it's useless. I just want people to stop saying "well look how good Olynyk's defensive rating is - he must be a great defender." I think one big thing this shows is the benefits our of our depth and the good coaching job we're getting from Stevens - He rarely has to put a big out there when he's going to be eaten alive, and usually picks the right guys to play at the right times.

[kozlodoev]

People who think that BR's individual defensive ratings are worth anything clearly have no understanding whatsoever of how they're calculate.

A quick recap: the individual DR concept is to pretty much measure possession-ending plays (defensive rebounds, steals, blocks) and forced misses. Except because they can't measure how many missed field goals an individual player has forced, they approximate this with the total number of missed shots while the player is on the floor.

Long story short, the individual ratings are pretty strongly correlated with team defensive, and boil down to the fact that if your unit is solid defensively, and you put some defensive stats into the box score, you will score off the charts on that rating. In my book, this is actually absolutely not indicative of defensive ability. Some of the best defensive players I've watched had no contribution to the box score whatsoever, and their assignments weren't even able to take many shots.

[nickagneta]

Many of the advanced metrics for basketball are not individualized stats. Defensive and offensive rating, plus minus in all it's permutations, win shares, etc. The statisticians attempt to render out the noise and make them individual but they're not 100% successful. So putting all your weight in your opinion based on these stats isn't always the best thing to do.

For instance, Tiaggo Splitter has has some of the highest ORTG and DRTG differentials in the league for years. Excellent defensive and offensive ratings. You would think only the best of the best would have such differentials. But Splitter isn't great. He played with a great TEAM and was an excellent cog in that wheel. But that's it.

Sully and KO are not great defenders. They might only this year be approaching the threshold of decent defenders. But they play on a very good defensive team, and are surrounded by plus defenders in AB, Smart, Amir,  and Crowder. Therefore, their advanced stats that are in some way team based are going to reflect as such giving the perception the players are better than they are.

KO and Sully are better defenders than they were a couple years ago. That's what I  take out of these numbers. Nothing more. Nothing less.

[perks-a-beast]

Some of the joke responses here make it seem as if Olynyk is some huge slouch defensively, which i don't believe to be the case at all. He just needs to do much better at contesting without fouling.

[fairweatherfan]

People who think that BR's individual defensive ratings are worth anything clearly have no understanding whatsoever of how they're calculate.

I have a thorough understanding of how they're calculated, and while they're clearly prone to noise, that's not the same as being worthless.  Individual results can and will be fluky, especially after a handful of games, but on aggregate better defenders will find themselves closer to the top of the rankings than worse defenders.  The smaller the sample size and the more dissimilar their team situations are, the more variability there is, but there remains a signal under the noise.  Doesn't have to be all or nothing. 

Most people would consider players like Whiteside, Kawhi, Gobert, Draymond, Duncan, Chandler, LeBron, etc to be among the top individual defenders in the league.  Is your argument that their presence toward the top of DRatings is a total coincidence and completely unrelated to their actual defensive ability?  Because that's what the ratings being worthless would imply.

[CapnDunks]

I get that Olynyk isn't a good rebounder for a seven footer, but why does everyone insist that he's a terrible rebounder period? Isn't he average for a 4? Or better considering where he plays?

[nickagneta]

I get that Olynyk isn't a good rebounder for a seven footer, but why does everyone insist that he's a terrible rebounder period? Isn't he average for a 4? Or better considering where he plays?

No, he really isn't average for a 4 or 5. His rebounding percentage is down near 11. Average for a big is about 14 to 15. For comparison, Sully is over 20 right now which is elite.

Look at it this way. When Olynyk is on the floor there are ten players on the floor so the average percentage I rebounds any given player should get is 10%. That doesn't take into effect size of player or position or ability to rebound. Just sheer numbers. 10 players on the floor, 10% chance of landing a rebound. But KO is over 7 foot and plays PF and center and he only gets 11% of available rebounds.

That's not good. KO is average as a rebounder as a basketball player that includes all SFs, SGs and PGs. But as a big, he's not average. Not close

[clover]

This is a great example of how you can't just look at overall team stats when you're on the floor. Jared Sullinger is 4th in D rating?

Here's another look at it: http://stats.nba.com/team/#!/1610612738/players/advanced/?sort=DEF_RATING&dir=-1

This one has only TR close to KO in DR on the C's and Sully toward the bottom of the team list, with only TZ higher among players who have played any minutes to speak of.

[PhoSita]

What the DRtg numbers tell us is that the mere presence of Kelly Olynyk on the floor does not kill the team's defense, which, going by the game threads, should be a bit of a revelation to a number of frequent posters here.

[littleteapot]

This is a great example of how you can't just look at overall team stats when you're on the floor. Jared Sullinger is 4th in D rating?

Here's another look at it: http://stats.nba.com/team/#!/1610612738/players/advanced/?sort=DEF_RATING&dir=-1

This one has only TR close to KO in DR on the C's and Sully toward the bottom of the team list, with only TZ higher among players who have played any minutes to speak of.

So do you think that Evan Turner and David Lee are top defenders on the Cs? This does nothing to make me think that offensive or defensive rating of the team when you are on the floor is an indicator of individual defensive contributions.

[fairweatherfan]

This is a great example of how you can't just look at overall team stats when you're on the floor. Jared Sullinger is 4th in D rating?

Here's another look at it: http://stats.nba.com/team/#!/1610612738/players/advanced/?sort=DEF_RATING&dir=-1

This one has only TR close to KO in DR on the C's and Sully toward the bottom of the team list, with only TZ higher among players who have played any minutes to speak of.

This seems to answer the question about the numbers Scal was quoting - he's clearly looking at the NBA's formula rather than BR's.

[nickagneta]

This is a great example of how you can't just look at overall team stats when you're on the floor. Jared Sullinger is 4th in D rating?

Here's another look at it: http://stats.nba.com/team/#!/1610612738/players/advanced/?sort=DEF_RATING&dir=-1

This one has only TR close to KO in DR on the C's and Sully toward the bottom of the team list, with only TZ higher among players who have played any minutes to speak of.

This seems to answer the question about the numbers Scal was quoting - he's clearly looking at the NBA's formula rather than BR's.

That there is more than one formula and that each formula comes up with some vastly different results should tell you all you need to know on just how much absolute value one should put into this stat.