Is PDSS a good way to analyze the Raptors defense?

[bthrawn]

A link to a PDF version of his 2009-10 PDSS is available here:

http://www.mediafire.com/?p4zrtxxppxghwka

Bosh and Bargs have identical 110.0 DRats, but their net points are different (Bosh is higher).

I would be cautious of using net points.

Player A could allow twice the points than player B but take 3 times the shots thus his net points will be higher than player B. Net points really just shows the players with the highest scoring average have the best net points.

[BorisDK1]

And therein lies some healthy debates which would spice up this board in season, as opposed to every post game thread becoming, "they sucked, they stank, I can't watch these bozos, fire _____, trade_____".

I'd love to see the statheads bring their individual game stats to a comparison thread say, within 48 hours of a game being concluded. Then debate the conclusions so we can compare what they graphed/charted with what we saw, or at least thought we saw. That would be immensely valuable, and in some ways more productive than arguing about certain player virtues from purely historical view. I prefer forward looking trends that suggest improvement or lack thereof in our young players, as opposed to telling me so and so sucked as a rookie.

I hope that Boris in particular isn't turned off by some of the ego here. Ripp in particular, who cannot abide ever being wrong on anything, like his manhood would shrivel if he ever admitted he had been shown evidence that forces a change in his opinion. We have a number of obsessive compulsives who believe they are God's gift and always right and will spend hours, even days, debating the number of angels on a pinhead rather than admitting that maybe there never was such a thing as an angel.

I don't pretend to be right all the time. I was wrong on Araujo, wrong on Jose - though not vs TJ Ford, just versus the evolving field of point guards. I think Bargnani is a supremely talented offensive player, with loads of flaws, but unlike some who would trade him for an expiring contract, I believe his situation is a lot more gray than the black and white terms often used here to describe him. And that's how I feel about stats. They enlighten, but only up to a point. The NBA corrupts its own stats. We know that. There have been startling admissions about how assists and rebounds are handed out. The league doesn't distinguish between high-difficulty rebounds and assists and ridiculous gimmes. Individual statisticians pad stats, sometimes because they are ordered to do so. It's a flawed world, and there is a lot of room for the subjective and objective to grind against each other to create a new Hegelian synthesis of understanding. Which half the board won't accept anyway.

That's my hope. I hope I'm providing data which doesn't exist popularly that can be used and people here can find ways to incorporate.

At the end of the day, I'm like you, dagger: I was basically supersub at raptorspace and had to deal with all the criticism and arrogance ("I don't need stats when my eyes tell me x") that he has had to deal with (probably not in the same volume). And in that, I made some fairly regrettable mistakes and frankly I'm glad I did. Without those analytical gaffes, I wouldn't have been motivated by the need for better defensive data and gone through with PDSS. So I'm not the guy demanding that the entire world conform to my view or this method (even though I frankly think it is the most direct and transparent method) but I do think it's data that has to be interpreted and built into the "story" we want to tell at some point.

[BorisDK1]

A link to a PDF version of his 2009-10 PDSS is available here:

http://www.mediafire.com/?p4zrtxxppxghwka

Bosh and Bargs have identical 110.0 DRats, but their net points are different (Bosh is higher).

A caution: that is very much a work in progress and there are some editing mistakes in that document. If I had to do it again, I wouldn't attempt to adjust for "team", defensively: those possessions happened and frankly I don't think they can be distributed.

I also think the lack of 3FGA data is hurting me there. If I may, I'll call that a trial run down the slope as opposed to the race itself.

[Dr Positivity]

I guess I just don't see how there's been 2 threads and 40+ pages arguing about Bargnani's defense

He's one of the worst defensive Cs in the league, if not the worst. There is no arguing this.

With that said, no player can single handidly submarine a defense. Especially if he ended up playing PF, or "C" beside Amir who'd be more of the defensive C anyways. If Jeff Green and Michael Beasley can start for top 10 defenses, so can Bargnani providing the rest of the team is there.

Last year's defense sucked not because we had Bargnani, but because we had Calderon, a rookie, Hedo, a 'no fouls for me!' Bosh, and Bargnani all in one starting lineup. That's arguably the worst PG, SF, and C defenders, a rookie, and a PF taking that side of the floor off. And it seems like Triano is a great offensive coach who's completley clueless defensively

Ideally by his late 20s Bargnani will be the same defensive player Dirk is, that is perfectly average... but we'll see. Even if he's still a squeaky wheel, teams can get by with one in their lineup. San Antonio can play great d with Matt **** Bonner starting

So really there's not much to say. Bargnani sucks at defense without question. He is not the defensive apocalypse and single handidly responsible for our defensive problems, or so bad defensively that every team he's on will be a DRTG bottomfeeder.

I guess if you want to argue which crapass defender had the least value between Jose, Hedo, and Bargnani, that's fine, but what's the point. We already know we're trading Jose first oppurtunity, and keeping Bargnani until further notice

[BorisDK1]

hey Boris,

how do Bosh/Bargs compare using PDSS???

Thanks Courtside for including the link!

Bargnani produced stops more regularly, but Bosh faced more possessions individually while he was on the floor. The team was slightly worse when Bosh didn't play. All in all, about the same.

[BorisDK1]

Just wanted to chime in one more time here, because I am being directly accused of things:
- Never said you were not competent enough. I just questioned the veracity of your computations when they don't jive with about 3 or 4 other statistical measures (Berri, APM, ON/OFF DRTG, Synergy), and 3 and 4 other anecdotal proofs, like the coach and GM saying that Bargnani has to improve his defensive awareness. Heck, even Doug Smith chimed in this morning:

I guess the biggest thing he needed to do this summer, in my opinion, was the play games and continue to develop the on-court intuition – especially defensively – that had been lacking.

Everything is pointing to the opposite of what you presented, hence my wariness.

There's no question that Bargnani needs to improve his awareness defensively in some regards. (Those regards are, namely, stopping dribble-drives and screen sequences and probably needs to learn the value of an all-out sprint his first three steps in transition, where on occasion he gives up early deep position due to running and not sprinting.) I've said so numerous times myself. That said, there are things his intuition does just fine with.

Personally, I believe Bargnani should have spent his summer away from that waking nightmare of an Italian program and in the weight room, getting stronger, quicker and a more powerful leaper. That's part of the problem. The mental gap that occurs due to that lack of confidence you have due to being at a physical disadvantage results in less aggressiveness. In basketball, the aggressor wins. And we know Bargnani ain't aggressive.

- The hyperbole is just too much, to tell you the truth. I never said "hopelessly". Again, I questioned your subjectivity when you admitted that you'd have no problem running Jose over with your car if he were to stop in front of you (jokingly of course).

I think I have been courteous to you, the person, while attacking the arguments and the stat.

For the most part? Sure you have. But questioning my objectivity really isn't going to get anybody anywhere: even if it were true, I'm not going to shoot myself in the foot after watching x amount of hours of Raptors' broadcasts, including the same 5 or 6 commercials in a seemingly endless loop.

[supersub15]

For the most part? Sure you have. But questioning my objectivity really isn't going to get anybody anywhere: even if it were true, I'm not going to shoot myself in the foot after watching x amount of hours of Raptors' broadcasts, including the same 5 or 6 commercials in a seemingly endless loop.

I apologize for this. Wasn't intended to question you the person, but a possible reason for the contradictory result of PDSS.

Are you up to a predictive thread? The idea is you, me, Ripp, and a couple of other cats put down our predictions for each team in the Eastern Conference (and Western, if we feel like it), and see who gets the closest results at the end of the season. Just for fun, of course.

[Reignman]

Just wanted to chime in one more time here, because I am being directly accused of things:
- Never said you were not competent enough. I just questioned the veracity of your computations when they don't jive with about 3 or 4 other statistical measures (Berri, APM, ON/OFF DRTG, Synergy), and 3 and 4 other anecdotal proofs, like the coach and GM saying that Bargnani has to improve his defensive awareness. Heck, even Doug Smith chimed in this morning:

I guess the biggest thing he needed to do this summer, in my opinion, was the play games and continue to develop the on-court intuition – especially defensively – that had been lacking.

Everything is pointing to the opposite of what you presented, hence my wariness.

There's no question that Bargnani needs to improve his awareness defensively in some regards. (Those regards are, namely, stopping dribble-drives and screen sequences and probably needs to learn the value of an all-out sprint his first three steps in transition, where on occasion he gives up early deep position due to running and not sprinting.) I've said so numerous times myself. That said, there are things his intuition does just fine with.

Personally, I believe Bargnani should have spent his summer away from that waking nightmare of an Italian program and in the weight room, getting stronger, quicker and a more powerful leaper. That's part of the problem. The mental gap that occurs due to that lack of confidence you have due to being at a physical disadvantage results in less aggressiveness. In basketball, the aggressor wins. And we know Bargnani ain't aggressive.

- The hyperbole is just too much, to tell you the truth. I never said "hopelessly". Again, I questioned your subjectivity when you admitted that you'd have no problem running Jose over with your car if he were to stop in front of you (jokingly of course).

I think I have been courteous to you, the person, while attacking the arguments and the stat.

For the most part? Sure you have. But questioning my objectivity really isn't going to get anybody anywhere: even if it were true, I'm not going to shoot myself in the foot after watching x amount of hours of Raptors' broadcasts, including the same 5 or 6 commercials in a seemingly endless loop.

Hey Boris, just out of curiousity, what were your thoughts when you heard Triano say Bargs doesn't understand the concept of help D in his end-of-season presser?

It seemed like Triano was saying he just doesn't have a fundamental understanding of what to do outside of man D situations.

Based on what you've seen when tagging games is that your assessment as well?

Personally I agree with Jay based on what I've seen but I'm asking for your take on it because you've tracked data first hand (minus 8%).

[OvertimeNO]

For the most part? Sure you have. But questioning my objectivity really isn't going to get anybody anywhere: even if it were true, I'm not going to shoot myself in the foot after watching x amount of hours of Raptors' broadcasts, including the same 5 or 6 commercials in a seemingly endless loop.

I apologize for this. Wasn't intended to question you the person, but a possible reason for the contradictory result of PDSS.

Are you up to a predictive thread? The idea is you, me, Ripp, and a couple of other cats put down our predictions for each team in the Eastern Conference (and Western, if we feel like it), and see who gets the closest results at the end of the season. Just for fun, of course.

I was actually going to propose this for you guys way back at the beginning of the thread, but didn't because I'm assuming that Boris only has PDSS stats for the guys who were on our roster last year, and only for last year.

A predictive thread would have to be more limited in scope if we're going to incorporate the PDSS data. My two cents.

[BorisDK1]

I apologize for this. Wasn't intended to question you the person, but a possible reason for the contradictory result of PDSS.

Yeah, I know. But maybe the result is what it is?

Are you up to a predictive thread? The idea is you, me, Ripp, and a couple of other cats put down our predictions for each team in the Eastern Conference (and Western, if we feel like it), and see who gets the closest results at the end of the season. Just for fun, of course.

I've never done a lot of prediction before league-wide. Sure, I'm in.

[BorisDK1]

Hey Boris, just out of curiousity, what were your thoughts when you heard Triano say Bargs doesn't understand the concept of help D in his end-of-season presser?

It seemed like Triano was saying he just doesn't have a fundamental understanding of what to do outside of man D situations.

Based on what you've seen when tagging games is that your assessment as well?

Personally I agree with Jay based on what I've seen but I'm asking for your take on it because you've tracked data first hand (minus 8%).

I'd agree with Jay. It needs to be hashed out a little more, as to exactly what he does poorly in help situations because there are a couple of things he does well. But there's no question that he's a waking nightmare with screen sequences and helping on dribble-drives.

[timdunkit]

I guess I just don't see how there's been 2 threads and 40+ pages arguing about Bargnani's defense

He's one of the worst defensive Cs in the league, if not the worst. There is no arguing this.

With that said, no player can single handidly submarine a defense. Especially if he ended up playing PF, or "C" beside Amir who'd be more of the defensive C anyways. If Jeff Green and Michael Beasley can start for top 10 defenses, so can Bargnani providing the rest of the team is there.

Last year's defense sucked not because we had Bargnani, but because we had Calderon, a rookie, Hedo, a 'no fouls for me!' Bosh, and Bargnani all in one starting lineup. That's arguably the worst PG, SF, and C defenders, a rookie, and a PF taking that side of the floor off. And it seems like Triano is a great offensive coach who's completley clueless defensively

Ideally by his late 20s Bargnani will be the same defensive player Dirk is, that is perfectly average... but we'll see. Even if he's still a squeaky wheel, teams can get by with one in their lineup. San Antonio can play great d with Matt **** Bonner starting

So really there's not much to say. Bargnani sucks at defense without question. He is not the defensive apocalypse and single handidly responsible for our defensive problems, or so bad defensively that every team he's on will be a DRTG bottomfeeder.

I guess if you want to argue which crapass defender had the least value between Jose, Hedo, and Bargnani, that's fine, but what's the point. We already know we're trading Jose first oppurtunity, and keeping Bargnani until further notice

Shhhh ... the stats geeks are talking !!!

[RonaldArtest]

Nice to see you on board here Boris, I was Jack Oakley on RC. Good luck preparing the data this season, look forward to some interesting debate.

[supersub15]

I've never done a lot of prediction before league-wide. Sure, I'm in.

Alright, get your stuff ready. It can be statistically-based, intuitively-based, or a combination of both. I know Ripp wants to do it too. I'll run some numbers tomorrow starting for the team with the letter A (Atlanta), then go down alphabetically.

[Reignman]

I've never done a lot of prediction before league-wide. Sure, I'm in.

Alright, get your stuff ready. It can be statistically-based, intuitively-based, or a combination of both. I know Ripp wants to do it too. I'll run some numbers tomorrow starting for the team with the letter A (Atlanta), then go down alphabetically.

I'd love to represent the empirical evidence contingent to give you nerds some balance.

[Reignman]

Hey Boris, just out of curiousity, what were your thoughts when you heard Triano say Bargs doesn't understand the concept of help D in his end-of-season presser?

It seemed like Triano was saying he just doesn't have a fundamental understanding of what to do outside of man D situations.

Based on what you've seen when tagging games is that your assessment as well?

Personally I agree with Jay based on what I've seen but I'm asking for your take on it because you've tracked data first hand (minus 8%).

I'd agree with Jay. It needs to be hashed out a little more, as to exactly what he does poorly in help situations because there are a couple of things he does well. But there's no question that he's a waking nightmare with screen sequences and helping on dribble-drives.

Do you believe the on/off DRTG captures those weaknesses better than PDSS? If you believe PDSS captures that aspect better than DRTG then is it safe to assume PDSS doesn't put a heavy weight on those aspects based on his overall rating?

[Guy Smiley]

LOL, PDSS vs DRTG reminds me of Jose vs TJ for some reason.

Is there anything that doesn't remind you of Jose vs TJ?

[Reignman]

LOL, PDSS vs DRTG reminds me of Jose vs TJ for some reason.

Is there anything that doesn't remind you of Jose vs TJ?

haha, I always expect the TJ wise cracks to come from Dagger. Nice to see you back brother, long time no see. You've gone from guerilla warfare to rooftop sniper, impressive.

[BorisDK1]

Alright, get your stuff ready. It can be statistically-based, intuitively-based, or a combination of both. I know Ripp wants to do it too. I'll run some numbers tomorrow starting for the team with the letter A (Atlanta), then go down alphabetically.

I work big hours Friday, Saturday and Sunday so don't expect anything non-pithy until Monday at the earliest (and usually I'm zonked then, too).

[Ripp]

I'd be down to do prediction stuff...but my algorithm isn't very computationally friendly (still trying to make it run faster), and hasn't yet been tested on the 09-10 dataset (which is what I'd want to do if I want to predict what will happen in 10-11.) I could use someone else's methods, but I'd much rather just use my own. Why don't we give it a week or two, let's say?