Is PDSS a good way to analyze the Raptors defense?

[Truthrising]

I agree as well, BorisDK1 is it ok if you do the game analysis and also the SoTD from now on?

I don't know about that...I'll certainly make the results for every game available here, if people want that.

I'd appreciate it and i'm sure others will too even though it may be a lot of work on your part but at least you are objective with your results as opposed to being biased towards certain players.

[BorisDK1]

Boris how do you account for offensive rebounds given up? Lets say a PF/C gives up an oreb which leads to a kickout and then a fg or 3 by a guard. Is that reflected in PDSS?

The offensive rebound itself? No. The event after the offensive rebound (the shot coming on the kickout pass)? Yes.

[ranger001]

Boris how do you account for offensive rebounds given up? Lets say a PF/C gives up an oreb which leads to a kickout and then a fg or 3 by a guard. Is that reflected in PDSS?

The offensive rebound itself? No. The event after the offensive rebound (the shot coming on the kickout pass)? Yes.

So do you assign .5 to the pf and .5 to the man who was guarding the guard?

[BorisDK1]

I'd appreciate it and i'm sure others will too even though it may be a lot of work on your part but at least you are objective with your results as opposed to being biased towards certain players.

Oh, I have my biases. I'll admit that if Jose Calderon walked out in front of my car as I go to my local country market here soon, I'd have a very difficult time convincing myself to really brake hard. j/k And if I could put a contraption onto Andrea's eyes like in A Clockwork Orange and somehow make him more aggressive, hey - I'd do that, too.

It's just that I invest so much of my time and my effort to do these that I really don't want to negate the entire process with blatant dishonesty.

[BorisDK1]

So do you assign .5 to the pf and .5 to the man who was guarding the guard?

No, because I'm just measuring what happens on the shot, not what created it.

[andreafan]

I'd appreciate it and i'm sure others will too even though it may be a lot of work on your part but at least you are objective with your results as opposed to being biased towards certain players.

Oh, I have my biases. I'll admit that if Jose Calderon walked out in front of my car as I go to my local country market here soon, I'd have a very difficult time convincing myself to really brake hard. j/k And if I could put a contraption onto Andrea's eyes like in A Clockwork Orange and somehow make him more aggressive, hey - I'd do that, too.

It's just that I invest so much of my time and my effort to do these that I really don't want to negate the entire process with blatant dishonesty.

Oh boris so you think we should use amir to man up on shaq huh? Oh ofc i forgot his shot blocking will make up for the 150 lb difference in bulk. Edit S#it i might have assign the wrong quote to the wrong person.

[bassmastert]

Boris how do you account for offensive rebounds given up? Lets say a PF/C gives up an oreb which leads to a kickout and then a fg or 3 by a guard. Is that reflected in PDSS?

The offensive rebound itself? No. The event after the offensive rebound (the shot coming on the kickout pass)? Yes.

So do you assign .5 to the pf and .5 to the man who was guarding the guard?

It seems to me after the response I received that the whole weight would go towards the guard. That can be a problem if you incorporate both qualitative and quantitative analysis, since the PF missing the rebound MAY have missed it due to poor defensive positioning/ability. I guess the thing with these stats is knowing how to interpret them and what you can and can't extrapolate from the data.

[ranger001]

So do you assign .5 to the pf and .5 to the man who was guarding the guard?

No, because I'm just measuring what happens on the shot, not what created it.

Shouldn't some blame be assigned to the pf/c though who gives up offensive rebounds? An offensive rebound leads to a fg at a higher percentage than a regular fg. If a guy is giving these up it hurts the team defensively and should count negatively towards the guy giving them up.

[Scott Carefoot]

You could also say why is the Raptors defense better when their leading shot blocker isn't playing. He lead the team by a ton over Bosh/Amir.

Amir is a superior shot-blocker to Bargnani. He averages 2.4 blocks per 36 minutes over his career while Bargnani averages 1.2 blocks per 36 over his career. It remains to be seen if Amir can reduce his foul rate while still being a good shot-blocker.

Scott you mean amir is a better weak side shot blocker please clarify thank you. Or to clarify a football analogy, you could say help side shotblocker. From a man to man straight up defender, andrea"s lenght and bulk make him much better shot blocker. Unfortunately amir is much too weak as
he can get out muscled when matched against centers or power forwards who outweigh him by 50lbs min. I hate when people use stats to generalize a particular individual.

You're really something. Should we start distinguishing between steals from stripped dribbles and steals on passes?

Among guys who played 1,000-plus minutes last season, Bargnani finished 36th in the league in Block Percentage. (Amir finished 27th.) Bargnani is not a prolific shot-blocker, no matter how his fans try to portray that as a strong part of his game. I seriously don't know whether to laugh or feel embarrassed at the reaches some people are making in these Bargnani threads.

[BorisDK1]

Shouldn't some blame be assigned to the pf/c though who gives up offensive rebounds? An offensive rebound leads to a fg at a higher percentage than a regular fg. If a guy is giving these up it hurts the team defensively and should count negatively towards the guy giving them up.

Well, there's a two-pronged answer for this. Should I assign blame directly? No, because no points were directly scored on the offensive rebound. And that's what we're assigning blame for: points, not other things. (The same argument might be made for the point guard who allows a ballhandler to blow right past him and drop a pass off for a layup to the player who was guarded by one of our bigs who came over to stop the drive. That score generally gets assigned to "team", not the point guard who allowed the drive - but the drive itself didn't directly result in the score, in a sense.)

Indirectly, though, there is a result. There is a "forced miss weight" aspect to calculating stops from the raw data, that tells us exactly how much a forced miss (either an FM or a blocked shot) is worth relative to defensive rebounds. There have been games where the Raptors have been abused on the boards where forced misses were worth < .2 stops. I remember one game where Andrea forced like 12 misses and blocked another shot or two and ended up with fewer than 3.5 stops. It does hurt bigs in particular to not get those defensive rebounds.

[andreafan]

I never said andrea was a strong shot blocker, i just said he was a better shot blocker when manned up against a guy very simple.And yes amir is much better help side shot blocker because instintively he understands help side defense better then andrea that's all.

[BorisDK1]

You're really something. Should we start distinguishing between steals from stripped dribbles and steals on passes?

Among guys who played 1,000-plus minutes last season, Bargnani finished 36th in the league in Block Percentage. (Amir finished 27th.) Bargnani is not a prolific shot-blocker, no matter how his fans try to portray that as a strong part of his game. I seriously don't know whether to laugh or feel embarrassed at the reaches some people are making in these Bargnani threads.

1) This isn't a Bargnani thread, although some have gone to extremes to make it such.
2) I'm sorry, I didn't realize there was a moratorium on this board for data that provides depth and texture to the whole "Bargnani is a poor defender" story people here want to tell. I didn't realize we should globally just insist he's terrible and not qualify that, where necessary. My apologies. I promise in the future to never discuss at all that there are a couple of things that Bargnani does well, without running it by you first.
3) Yes, I understand that people who downplay the extent of the issues that Bargnani causes / doesn't solve are frustrating, but that doesn't give you, me or anybody else carte blanche to twist or distort truth to serve your beliefs.

[Courtside]

The flaw I see with on/off stats is that they don't take into account what the opposition is doing. It's not measuring the defense, it's measuring the opponent's offense. It's not a mirror, or a black and white thing.

For example - while Bargs is in the game - do we know that the opponents score more inside or outside? Which is he more responsible for? Then, when he's off, for the on/off stats to be taken at face value, we have to assume that the opposition does the exact same thing whether he's there or not - which we know is not the case. What you put out there defensively is attacked differently by the offense.

This is why simply removing Jose's 118 from the team cannot simply be removing the lowest/highest from a group and adjusting the averages. This can explain why Calderon and Jack appear to have a similar defensive efficiency when looking at the overall on/off numbers, but it only means that the offense is doing something different. They don't attack Jack the same way they attack Calderon, and offenses don't go at Andrea the same way they go at Amir.

Some things to look at that can help provide additional context could be do we give up more 3 pointers with Andrea in, or out? Do we give up more driving layups with Jose in, or out? We have to understand what the offense is doing when each player is on/off to understand the value of their on/off numbers. Just like those (very interesting) blogs about the value of a rebound, we have to understand the value of each player when weighted against what they force the other team to do.

By using PDSS, we can see that Bargs was the defender on the largest number of possessions. We don't get that in on/off. Using Synergy in conjunction with PDSS can tell us a lot of other details that on/off don't - such as if Andrea faced only 289 one-on-one plays out of 1040 total defensive plays, then he is providing help on as many as 751 other plays. Using PDSS, we see that Bosh and Bargs both make about 0.37 defensive plays per minute, while Amir makes 0.44 defensive plays per minute. Is he more active, or do they go at him more because he fouls so much? Probably both - and that probably affects his ON numbers along with Andrea's OFF numbers.

[Truthrising]

You're really something. Should we start distinguishing between steals from stripped dribbles and steals on passes?

Among guys who played 1,000-plus minutes last season, Bargnani finished 36th in the league in Block Percentage. (Amir finished 27th.) Bargnani is not a prolific shot-blocker, no matter how his fans try to portray that as a strong part of his game. I seriously don't know whether to laugh or feel embarrassed at the reaches some people are making in these Bargnani threads.

1) This isn't a Bargnani thread, although some have gone to extremes to make it such.
2) I'm sorry, I didn't realize there was a moratorium on this board for data that provides depth and texture to the whole "Bargnani is a poor defender" story people here want to tell. I didn't realize we should globally just insist he's terrible and not qualify that, where necessary. My apologies. I promise in the future to never discuss at all that there are a couple of things that Bargnani does well, without running it by you first.
3) Yes, I understand that people who downplay the extent of the issues that Bargnani causes / doesn't solve are frustrating, but that doesn't give you, me or anybody else carte blanche to twist or distort truth to serve your beliefs.

not to derail to the topic but...

Let me clarify one thing for you, most of the people that are bashing Bargnani were HUGE Bosh supporters so you can see why talking about Bargnani or Bosh has been a touchy subject in the past. As for Bosh fans they just cannot fathom the fact that Bargnani is still around and will do anything to extrapolate his weaknesses and never bring any encouraging signs about his improvements, if they want talk about bargnani at least be objective about it instead always magnifying his weaknesses and nothing more. This is why it's refreshing to see you come aboard

[BorisDK1]

1. Absolutely. I've already admitted such. But the fact remains that the team numbers go counter to your PDSS numbers. If Calderon were such a terrible defender (and he is getting beat at the point of attack, I haven't disputed that), 4000+ minutes without Bargnani should correspond to your PDSS. But they don't, which is making me question PDSS.

2. I've already answered that. There are 9000+ minutes in my sampling. Those 9000+ minutes include Bargnani starting/bench and Calderon starting/bench. Any noise that you keep mentioning is smoothed to insignificant levels. Yet, you refuse to accept the fact that 9000+ minutes is a bigger sample than the 1800 minutes or so that Calderon played last year.

3. I am coming to logical conclusions supported by 4 years of data. I know that you are trying to dismiss it because it goes counter to your PDSS conclusions.

I have shown you that the team plays better (by 8 points per 100 possessions) with ALL AND ANY 5-player combinations, yet you refuse to acknowledge that.

As Fenris-77 has said, I agree that Bargnani isn't entirely useless defensively. He's above-average on post-ups and when isolated one-on-one, but those are a finite number of plays over a season. For instance, Synergy counts 281 post-up plays against Bargnani out of 9869. That's 3% of the total amount of plays, and he only stopped only 130 of those. That's 1.625 stops per game over his entire season.

1) Good to see that acknowledged, now. I say that because you asked me if I thought Jarrett Jack was a better defender than Jose (in those words!) and went right to on/off court data to try to get an answer to that. So it's nice to see some enrichment, here.

As far as Bargnani goes, there's more than just isolations and post-ups that Bargnani does well defending. When switched onto perimeter players, he does surprisingly well (and that's always something I appreciate, because the rule in my motion offense is, in double-mismatch situations, the smaller player should drive against the bigger player instead of throwing it into the post). There are things he does rotating on the perimeter in ball reversal and closing out to the perimeter quite well. There are things he isn't effective at all at (helping on dribble drives, defending screen sequences) and certainly his rebounding is a huge and going concern. I think there's a lot of stuff about his game defensively you're missing for whatever reason(s), but that through this method I am forced to see.

2. I don't see the noise diminishing, I see just a larger noisy sample being cobbled together. I'm not saying it's worthless - it does communicate something and we need to acknowledge that. But what does it communicate? That Bargnani himself is the ultimate cause of all badness, that there's no solution for Andrea? I don't think you can justify that stance, and certainly not from this kind of analysis.

3. You are coming to conclusions provided by a certain kind of data. And frankly, you're just taking the various metrics you want to look at at face value, there's no texture, there's no depth beyond that. And I am going to keep hammering home on the fact that I, personally, prefer metrics to be used created by a coach who is happens to be an engineer for coaching purposes than by engineers for their purposes.

4. It's not that I refuse to acknowledge the results of those metrics, supersub. It's that I refuse to accept your interpretation of them at face value and pretend as though they speak for themselves, with final authority. They provide analysis by periphery. They're helpful to look at, but they don't tell you: "who's responsible? for how much? what can be done?". You're certainly not helping that discussion along, either. And I think it's funny that while you do admit that you cannot use these metrics to ultimately determine the quality of play of an individual defender, you insist an awful lot that that's the only story that could possibly be told viz. Bargnani. And I'm not buying that story without major, major qualification.

I'm not trying to defend PDSS to you. You are trying to force me to accept that all things metrical on the defensive end need to conform themselves to your interpretation of on/off court data, and I'm not going to do that. Not out of stubbornness, but because there's direct data that you want to dismiss in toto. We've already seen some really, really bad accusations against PDSS (like, the initial post in this entire thread, for one). I do think that direct data taken from the basketball floor is far more valuable than just counting outcome possessions where a player happened to be on the floor, or off of it.

This entire thrust of argumentation once again starts to pit basketball analysis against statistical evidence. I truly think we need to consider both, not just dismiss one out-of-hand because of unfamiliarity/novelty/preference. I'm more than willing - and have, in my own interpretation, considered the impact of on/off court data, but I truly have seen almost no evidence that you have considered this.

[dagger]

Wow, Carlos and Ripp crash and burn.

[Calderon]

Wow, Carlos and Ripp crash and burn.

Piggy-backer.

[andreafan]

And on a side note, every time andrea is late on a rotation , what do you want him to do, pound the offensive player thus fouling, this would accumulate more fouls then the foul rate of amir. It's either that or have andrea try to finesse block like he's trying to swap at flies when he's late on a rotation. Don't you want andrea on the court ?

[Scott Carefoot]

not to derail to the topic but...

Let me clarify one thing for you, most of the people that are bashing Bargnani were HUGE Bosh supporters so you can see why talking about Bargnani or Bosh has been a touchy subject in the past. As for Bosh fans they just cannot fathom the fact that Bargnani is still around and will do anything to extrapolate his weaknesses and never bring any encouraging signs about his improvements, if they want talk about bargnani at least be objective about it instead always magnifying his weaknesses and nothing more. This is why it's refreshing to see you come aboard

I don't recall seeing too many "huge Bosh fans" around here except when he was playing at an All-Star level in the first half of the season. I've pretty much always identified him as a Pau Gasol type who is better suited to playing the role he's now playing in Miami, and I think most Raptors fans with a clear head saw it the same way. Personally, I'm thrilled Bosh left because this means the Raptors can break out of their treadmill of mediocrity and have a legitimately awful record that should earn them a top-five draft pick. Of course, as long as Bargnani averages 20 PPG next season regardless of the team's performance, then that will please a certain type of fan.

[Schad]

Wow, Carlos and Ripp crash and burn.

So they're having an interesting back-and-forth, and your input is to deem one side the conclusive winner and declare the others to be crashing and burning? Feel free to sit this thread out.