Tracking Duke's Defense - 2018-19

[Kedsy]

A couple data points to help confirm the completely obvious theory that to play great defense you need all your best players to play:

Against Syracuse, Tre played 12 possessions. A small sample, to be sure, but:

WITH TRE:
0.50 dRating (0.46 adjusted); 41.7% TO%

WITHOUT TRE:
1.13 dRating (1.08 adjusted); 12.7% TO%


Against Florida State, Zion played only in the first half:

FIRST HALF, WITH ZION:
0.98 dRating (0.89 adjusted)

SECOND HALF, WITHOUT ZION:
1.16 dRating (1.07 adjusted)


For comparison's sake, 0.89 adjusted dRating is the level of the 5th best defense in the country; 1.07 dRating is the level of the 255th best defense (and 1.08 is the level of the 274th best defense).

Just sayin'.

[uh_no]

dRtg: 1.05 (adjusted that's 1.00, not great but considering we played without our two best defenders, it might be as good as we could have hoped for)
eFG%: 50.6% (pretty bad but, again, we played without our two best defenders)
3pt%: 44.0% (really bad but, ditto)
2pt%: 43.6% (not bad, considering)
%threes: 31.3%
FT rate: 23.8%
DR%: 72.3% (really good for us, against a pretty good (though not great) offensive rebounding team)
TO%: 16.4% (broke our nine game streak of having a 20%+ TO%; clearly Tre and/or Cam are the key to Duke turning our opponents over)
a/to: 0.87:1 (tenth straight game our opponents have had more turnovers than assists)
%assisted: 37.1%
fast break pts: for some reason, we only have first half data for this, and in the first half Syracuse had 4 fb pts for 8.3% of their points
block%: 15.0%; 21.8% of 2-point shots (very strong for the second-best blocking team in the country)
steal%: 8.8% (third straight only so-so game, compared to our season steals performance; but without Cam and Tre, that was to be expected)


I hate to sound like a broken record, but playing without our two best defenders our overall defensive performance wasn't that bad.

can't be mad with the ridiculous threes they were hitting. If there was a concept of "effeciency luck" which accounted for good shots which miss and bad shots which go in, we were unlucky on both sides of the ball last night. Fortunately their 0-6 from deep down the stretch helped bring their overall number down a bit.

[COYS]

can't be mad with the ridiculous threes they were hitting. If there was a concept of "effeciency luck" which accounted for good shots which miss and bad shots which go in, we were unlucky on both sides of the ball last night. Fortunately their 0-6 from deep down the stretch helped bring their overall number down a bit.

I know some of the motion-capture cameras track that data for NBA games. I don't recall the specific articles, but I've read a number over the years that have used such data to show that Steph Curry is ridiculously good at making closely-guarded threes that would be awful shots for almost anyone else. I wonder if there is a similar database for that for college. I guess it would depend on the individual school arenas. And it might not be publicly available. But it would be interesting to see. I really do think we got a bit unlucky last night. Our defense wasn't quite as bad as it seemed after Tre went down. 'Cuse hit a lot of closely guarded shots with players that don't usually hit threes. Even with Tre, that game would have been closer than any of us would have liked (that, or Tre would've just continued to steal the ball and we would have run away with it).

[uh_no]

I know some of the motion-capture cameras track that data for NBA games. I don't recall the specific articles, but I've read a number over the years that have used such data to show that Steph Curry is ridiculously good at making closely-guarded threes that would be awful shots for almost anyone else. I wonder if there is a similar database for that for college. I guess it would depend on the individual school arenas. And it might not be publicly available. But it would be interesting to see. I really do think we got a bit unlucky last night. Our defense wasn't quite as bad as it seemed after Tre went down. 'Cuse hit a lot of closely guarded shots with players that don't usually hit threes. Even with Tre, that game would have been closer than any of us would have liked (that, or Tre would've just continued to steal the ball and we would have run away with it).

duke uses the tech. you'll note that they put trackers just below the neckline on the inside of the jerseys during warm-up. You'd need dataa from both teams to be able to get a complete picture.

[Kedsy]

dRtg: 1.11 (adjusted that's 0.94, decent but probably better than decent, considering our best defender watched from the bench)
eFG%: 55.7% (bad)
3pt%: 17.6% (amazing, but not sure how much was skill and how much was luck)
2pt%: 69.4% (horrible)
%threes: 32.1%
FT rate: 32.1% (this is probably too high, but not alarmingly so)
DR%: 75.9% (really good for Duke, though Virginia isn't the best offensive rebounding team, either)
TO%: 12.7% (way too low, but actually this is about average for Virginia's offense)
a/to: 1:1 (after ten straight games holding our opponents to fewer assists than turnovers, Virginia broke that streak by having the same number of assists and TOs)
%assisted: 28.6%
fast break pts: 8 (11.4% of their points)
block%: 1.9%; 2.8% of 2-point shots (we only had one block, which for this team is almost hard to believe)
steal%: 7.9% (fourth straight only so-so game, compared to our season steals performance; but against Virginia and without Tre, it's actually pretty decent)


Not the best defensive effort, but I think we'll all take it.

[MChambers]

dRtg: 1.11 (adjusted that's 0.94, decent but probably better than decent, considering our best defender watched from the bench)
eFG%: 55.7% (bad)
3pt%: 17.6% (amazing, but not sure how much was skill and how much was luck)
2pt%: 69.4% (horrible)
%threes: 32.1%
FT rate: 32.1% (this is probably too high, but not alarmingly so)
DR%: 75.9% (really good for Duke, though Virginia isn't the best offensive rebounding team, either)
TO%: 12.7% (way too low, but actually this is about average for Virginia's offense)
a/to: 1:1 (after ten straight games holding our opponents to fewer assists than turnovers, Virginia broke that streak by having the same number of assists and TOs)
%assisted: 28.6%
fast break pts: 8 (11.4% of their points)
block%: 1.9%; 2.8% of 2-point shots (we only had one block, which for this team is almost hard to believe)
steal%: 7.9% (fourth straight only so-so game, compared to our season steals performance; but against Virginia and without Tre, it's actually pretty decent)


Not the best defensive effort, but I think we'll all take it.

Without Tre, will definitely take it. Will be interesting to see how his presence changes the rematch, assuming he’s back in three weeks. I expect Duke will be able to speed up the tempo and force more turnovers with Tre back.

[devildeac]

Not Duke's defense, but our 3 point shooting is now 31.2%, good for a logjam/tie from #290-295 in the nation.

[Kedsy]

dRtg: 0.93 (adjusted that's still 0.93, decent but considering our best defender watched from the bench...)
eFG%: 44.3% (acceptable, not great)
3pt%: 20.0% (second straight super strong performance, but hard to say whether Duke was good or Pitt was bad)
2pt%: 50.0% (not great)
%threes: 28.3% (good, but Pitt's 258th in the country in this, so probably to be expected)
FT rate: 37.7% (this is probably too high, but considering Pitt is 9th in the country in this, with an average of 45.7%, it's not bad)
DR%: 67.6%
TO%: 20.4% (anything over 20% is cool with me)
a/to: 0.43:1 (12th straight game with our opponents having equal or more turnovers than assists)
%assisted: 27.3% (3rd straight very low number in this category)
fast break pts: 13 (20.3% of their points; only 2nd time our opponents have topped 20% in this all season (Auburn 25.0%))
block%: 13.2%; 18.4% of 2-point shots (back to looking like the 2nd best blocking team in the nation)
steal%: 13.1% (back to looking like the best stealing team in the nation, and without Tre, that's saying something)


Fourth straight adjusted dRating of 93 or worse. Not the happiest news but seeing as we're doing this without our best defender, I'm willing to look past it.

[Truth&Justise]

dRtg: 0.93 (adjusted that's still 0.93, decent but considering our best defender watched from the bench...)
eFG%: 44.3% (acceptable, not great)
3pt%: 20.0% (second straight super strong performance, but hard to say whether Duke was good or Pitt was bad)
2pt%: 50.0% (not great)
%threes: 28.3% (good, but Pitt's 258th in the country in this, so probably to be expected)
FT rate: 37.7% (this is probably too high, but considering Pitt is 9th in the country in this, with an average of 45.7%, it's not bad)
DR%: 67.6%
TO%: 20.4% (anything over 20% is cool with me)
a/to: 0.43:1 (12th straight game with our opponents having equal or more turnovers than assists)
%assisted: 27.3% (3rd straight very low number in this category)
fast break pts: 13 (20.3% of their points; only 2nd time our opponents have topped 20% in this all season (Auburn 25.0%))
block%: 13.2%; 18.4% of 2-point shots (back to looking like the 2nd best blocking team in the nation)
steal%: 13.1% (back to looking like the best stealing team in the nation, and without Tre, that's saying something)


Fourth straight adjusted dRating of 93 or worse. Not the happiest news but seeing as we're doing this without our best defender, I'm willing to look past it.

Kedsy, I doubt you have numbers for this, but I'm curious what the breakdown would be for a zone defense vs. man-to-man. Based on the eye test, our zone defense looks pretty bad. We're not yet making the right rotations against simple actions, and as a result we give up a lot of open corner threes. Against a good team, that's deadly.

That said, I think K was brilliant for deploying zone for most of the second half. We had a comfortable lead, and this was an opportunity to get in game reps playing zone against a not-horrible team. I'd gladly trade a dip in our defensive numbers if it means getting competent in a different scheme. K seems determined to have this team be able to play a passable zone for a few minutes in big games, and this was a good step to getting there. I would be shocked if we don't see it some more against Georgia Tech and Notre Dame if we're able to build a decent lead in those games too.

[English]

dRtg: 0.93 (adjusted that's still 0.93, decent but considering our best defender watched from the bench...)
eFG%: 44.3% (acceptable, not great)
3pt%: 20.0% (second straight super strong performance, but hard to say whether Duke was good or Pitt was bad)
2pt%: 50.0% (not great)
%threes: 28.3% (good, but Pitt's 258th in the country in this, so probably to be expected)
FT rate: 37.7% (this is probably too high, but considering Pitt is 9th in the country in this, with an average of 45.7%, it's not bad)
DR%: 67.6%
TO%: 20.4% (anything over 20% is cool with me)
a/to: 0.43:1 (12th straight game with our opponents having equal or more turnovers than assists)
%assisted: 27.3% (3rd straight very low number in this category)
fast break pts: 13 (20.3% of their points; only 2nd time our opponents have topped 20% in this all season (Auburn 25.0%))
block%: 13.2%; 18.4% of 2-point shots (back to looking like the 2nd best blocking team in the nation)
steal%: 13.1% (back to looking like the best stealing team in the nation, and without Tre, that's saying something)


Fourth straight adjusted dRating of 93 or worse. Not the happiest news but seeing as we're doing this without our best defender, I'm willing to look past it.

Perhaps you missed it, but according to the Naismith DPOY mid-season watch list, Zion is our best defender. /s

[CDu]

Kedsy, I doubt you have numbers for this, but I'm curious what the breakdown would be for a zone defense vs. man-to-man. Based on the eye test, our zone defense looks pretty bad. We're not yet making the right rotations against simple actions, and as a result we give up a lot of open corner threes. Against a good team, that's deadly.

That said, I think K was brilliant for deploying zone for most of the second half. We had a comfortable lead, and this was an opportunity to get in game reps playing zone against a not-horrible team. I'd gladly trade a dip in our defensive numbers if it means getting competent in a different scheme. K seems determined to have this team be able to play a passable zone for a few minutes in big games, and this was a good step to getting there. I would be shocked if we don't see it some more against Georgia Tech and Notre Dame if we're able to build a decent lead in those games too.

I don't think the zone is going to be used for anything other than spot support for whenever we can't stop dribble penetration. I think we'll remain a man-first defense and mix in the zone only as needed. And I suspect that once Tre Jones returns we'll see less and less need for the zone.

[Kedsy]

Perhaps you missed it, but according to the Naismith DPOY mid-season watch list, Zion is our best defender. /s

They probably place too much emphasis on stats like dRating, in which Zion ranks #6 in the nation (82.8). It's a flawed stat that over-rewards blocks and steals and gives individuals credit for team performance while they're on the floor. Texas Tech has three of the top 10 in this stat. Virginia has two of the top 10 (Braxton Key is #1 in the country with a dRtg of 77.5 and Ty Jerome is #10).

Here's how Duke's players stack up in defensive rating:

Zion: 82.8
Javin: 86.3
Cam: 88.8
Jack: 89.3
Marques: 89.9
Alex: 91.4
Tre: 92.2
RJ: 92.3
Jordan: 93.2

So if you looked at stats like that, and you wanted to pick one Duke player, Zion would be the guy and Tre wouldn't even be considered.

But we know better.

[English]

They probably place too much emphasis on stats like dRating, in which Zion ranks #6 in the nation (82.8). It's a flawed stat that over-rewards blocks and steals and gives individuals credit for team performance while they're on the floor. Texas Tech has three of the top 10 in this stat. Virginia has two of the top 10 (Braxton Key is #1 in the country with a dRtg of 77.5 and Ty Jerome is #10).

Here's how Duke's players stack up in defensive rating:

Zion: 82.8
Javin: 86.3
Cam: 88.8
Jack: 89.3
Marques: 89.9
Alex: 91.4
Tre: 92.2
RJ: 92.3
Jordan: 93.2

So if you looked at stats like that, and you wanted to pick one Duke player, Zion would be the guy and Tre wouldn't even be considered.

But we know better.

That's right, we know that freshmen OADs can't play defense. (Boom, nailed it!)

[Rich]

I don't think the zone is going to be used for anything other than spot support for whenever we can't stop dribble penetration. I think we'll remain a man-first defense and mix in the zone only as needed. And I suspect that once Tre Jones returns we'll see less and less need for the zone.

I can see it as a valuable tool coming out of a timeout where the opposing team has drawn up a play against our M2M. Element of surprise.

[Truth&Justise]

I don't think the zone is going to be used for anything other than spot support for whenever we can't stop dribble penetration. I think we'll remain a man-first defense and mix in the zone only as needed. And I suspect that once Tre Jones returns we'll see less and less need for the zone.

I can see it as a valuable tool coming out of a timeout where the opposing team has drawn up a play against our M2M. Element of surprise.

However it is deployed, it will probably be present in some fashion, and my hope is that it will eventually be more stout than it is now. Even if it's just for a few possessions, having a porous zone can be disastrous -- I'm reminded of the early second half against Kansas in the Elite Eight last year. The gambit to catch them off guard with a 1-3-1 really backfired, Kansas went on a run, and it was one of several inflection points in that game.

I think the only way to get better is real game reps until our rotations are crisp and automatic. We're still too easily undone by simple actions to generate open perimeter looks.

[English]

However it is deployed, it will probably be present in some fashion, and my hope is that it will eventually be more stout than it is now. Even if it's just for a few possessions, having a porous zone can be disastrous -- I'm reminded of the early second half against Kansas in the Elite Eight last year. The gambit to catch them off guard with a 1-3-1 really backfired, Kansas went on a run, and it was one of several inflection points in that game.

I think the only way to get better is real game reps until our rotations are crisp and automatic. We're still too easily undone by simple actions to generate open perimeter looks.

Really, the only way to get better? I agree with much of what you say--I'm positive that everyone on this board hopes the zone improves, as well as the M2M, the DR%, the 3/4 court press, and on and on. But seriously, these are freshmen and a collection of guys who haven't played much together. Practice time over the course of a season hones their teamwork and comfort as a unit. Real game reps are nice, and important to improvement, but practicing these principles and rotations over and over again in practice is huge. Perhaps I'm just nitpicking.

[cato]

I don't think the zone is going to be used for anything other than spot support for whenever we can't stop dribble penetration. I think we'll remain a man-first defense and mix in the zone only as needed. And I suspect that once Tre Jones returns we'll see less and less need for the zone.

Indeed. 2015 was the beginning of the new era, where K embraced zone during the mid-season swoon to right the ship. But by the tournaments, the team was back to playing man-to-man (thankfully, at a very high level).

When Duke plays man-to-man well, K is going to favor that. I am glad, however, he has moved to zone to fill the gaps when the team cannot stop penetration.

[Kedsy]

dRtg: 0.76 (adjusted that's 0.80, back to a great number with Tre back)
eFG%: 44.0% (not terrible)
3pt%: 30.0%
2pt%: 43.8% (good)
%threes: 17.2% (very good)
FT rate: 10.3% (super low, against a team that's pretty good at getting to the line)
DR%: 73.0% (very good, for us)
TO%: 27.4% (very strong; no coincidence this happened with Tre back)
a/to: 0.58:1 (13th straight game with our opponents having equal or more turnovers than assists)
%assisted: 45.8.3%
fast break pts: 8 (15.1% of their points)
block%: 12.1%; 14.6% of 2-point shots (another strong blocking game)
steal%: 18.7% (super strong)


No coincidence our D suffered a bit in the three games Tre was out and returned to greatness today. Welcome back Tre

[uh_no]

dRtg: 0.76 (adjusted that's 0.80, back to a great number with Tre back)
eFG%: 44.0% (not terrible)
3pt%: 30.0%
2pt%: 43.8% (good)
%threes: 17.2% (very good)
FT rate: 10.3% (super low, against a team that's pretty good at getting to the line)
DR%: 73.0% (very good, for us)
TO%: 27.4% (very strong; no coincidence this happened with Tre back)
a/to: 0.58:1 (13th straight game with our opponents having equal or more turnovers than assists)
%assisted: 45.8.3%
fast break pts: 8 (15.1% of their points)
block%: 12.1%; 14.6% of 2-point shots (another strong blocking game)
steal%: 18.7% (super strong)


No coincidence our D suffered a bit in the three games Tre was out and returned to greatness today. Welcome back Tre

I think the D was probably even better than advertised today. in their spurt around either side of halftime, they generated a good bit of offense off our offensive offensive execution. When we cleaned it up on that side of the flloor, they're offense went kaput.

[Kedsy]

I think the D was probably even better than advertised today. in their spurt around either side of halftime, they generated a good bit of offense off our offensive offensive execution. When we cleaned it up on that side of the flloor, they're offense went kaput.

In the last 3:57 of the first half plus the first 4:17 of the 2nd half, Ga Tech had 16 points in 14 possessions (1.14 ppp). The rest of the game (approx 32 mins) they had 37 points in 55.4 possessions (0.67 ppp).