Bogut Watch - now with AdonalFoyle4Prez's poll

[EvanZ]

http://stats-for-the-nba.appspot.com/teams/GSW.html

RAPM numbers are up and running as well. No idea as of what date.

Lee's number holds up.

It's a new form of RAPM using box scores as a prior. I don't really know all the details yet, but it's being discussed here:

http://www.apbr.org/metrics/viewtopic.p ... 848#p13848

This is apparently a more accurate predictor than "old" RAPM.

[Sleepy51]

http://stats-for-the-nba.appspot.com/teams/GSW.html

RAPM numbers are up and running as well. No idea as of what date.

Lee's number holds up.

It's a new form of RAPM using box scores as a prior. I don't really know all the details yet, but it's being discussed here:

http://www.apbr.org/metrics/viewtopic.p ... 848#p13848

This is apparently a more accurate predictor than "old" RAPM.

Good to know.

Evan,
This is my burning RAPM question and you are the only person I trust to answer it:

What would it take to adjust RAPM to make some kind of accounting for a player's minutes capability or "role" i.e. correct fort he Matt Bonner error? How do we get subs and role players properly adjusted for vs. starters and extended minutes rotation players to turn this particular corner. Or do we just have to continue to do that with common sense?

I don't mind that RAPM has this gap and I am capable of having a common sense conversation around those oddities, but why can't or how would some kind of minutes per game of production factor be integrated into the stat?

And should we?

If the point of this is to correlate individual players to wins, is there are statistically meaningful way to differentiate between a mediocre starter and an exceptional sub? Or a bad overused starter vs. a bad underused sub?

[EvanZ]

Evan,
This is my burning RAPM question and you are the only person I trust to answer it:

What would it take to adjust RAPM to make some kind of accounting for a player's minutes capability or "role" i.e. correct fort he Matt Bonner error? How do we get subs and role players properly adjusted for vs. starters and extended minutes rotation players to turn this particular corner. Or do we just have to continue to do that with common sense?

Obviously there's no magic formula, but I would imagine that a starting place would be to use mpg as part of the prior for doing the regression. I don't know if that is involved in xRAPM right now but it wouldn't surprise me.

I don't mind that RAPM has this gap and I am capable of having a common sense conversation around those oddities, but why can't or how would some kind of minutes per game of production factor be integrated into the stat?

And should we?

If the point of this is to correlate individual players to wins, is there are statistically meaningful way to differentiate between a mediocre starter and an exceptional sub? Or a bad overused starter vs. a bad underused sub?

If we can determine that using minutes in our prior makes the prediction of wins better in a significant way (however you define that), then yes, I believe it should absolutely be done. It's important to keep in mind, though, that there are always going to be outliers in any system. It's like whack-a-mole. Maybe we get rid of the Matt Bonners and Vince Carters and some other guys bubble up to the top that we don't "like".

Hope that answer helps.

[floppymoose]

when you mention teams missing players, shouldn't you mention 2 of our most important pieces as well? To be at .500 with already three road wins on the board, you should give a little credit.

Why should I mention them? Are they coming back soon? Or is this team the one we have for this season?

[KevinMcreynolds]

when you mention teams missing players, shouldn't you mention 2 of our most important pieces as well? To be at .500 with already three road wins on the board, you should give a little credit.

Why should I mention them? Are they coming back soon? Or is this team the one we have for this season?

Sorry, but we've played 7 of 11 games on the road missing our key player and 6th man. I'll gladly take the 6-5. My only issue is we really should be 7-4 since that Nuggets game was a choke job of epic proporations.

[Sleepy51]

Evan,
This is my burning RAPM question and you are the only person I trust to answer it:

What would it take to adjust RAPM to make some kind of accounting for a player's minutes capability or "role" i.e. correct fort he Matt Bonner error? How do we get subs and role players properly adjusted for vs. starters and extended minutes rotation players to turn this particular corner. Or do we just have to continue to do that with common sense?

Obviously there's no magic formula, but I would imagine that a starting place would be to use mpg as part of the prior for doing the regression. I don't know if that is involved in xRAPM right now but it wouldn't surprise me.

I don't mind that RAPM has this gap and I am capable of having a common sense conversation around those oddities, but why can't or how would some kind of minutes per game of production factor be integrated into the stat?

And should we?

If the point of this is to correlate individual players to wins, is there are statistically meaningful way to differentiate between a mediocre starter and an exceptional sub? Or a bad overused starter vs. a bad underused sub?

If we can determine that using minutes in our prior makes the prediction of wins better in a significant way (however you define that), then yes, I believe it should absolutely be done. It's important to keep in mind, though, that there are always going to be outliers in any system. It's like whack-a-mole. Maybe we get rid of the Matt Bonners and Vince Carters and some other guys bubble up to the top that we don't "like".

Hope that answer helps.

That helps. My gut says that just including minutes in the calculation will skew simply towards guys who play a lot of minutes without applying any qualitative screen for the quality of those minutes. How would you integrate a qualitative element so that we aren't crediting good and bad minutes equally?

[Mylie10]

when you mention teams missing players, shouldn't you mention 2 of our most important pieces as well? To be at .500 with already three road wins on the board, you should give a little credit.

Why should I mention them? Are they coming back soon? Or is this team the one we have for this season?

Because you're listing teams that are missing guys...that's why. If you don't see the inconsistency then fine.

Bogut should be back at some point. Neither you nor I know for sure though. You could also say that if the 2 supposed best shooters on the team didn't shoot like ess we may have easily won those games instead of barely beating them.

As bad as Klay is playing and shooting the ball, alot of his shots are in and out.

[Chris Porter's Hair]

when you mention teams missing players, shouldn't you mention 2 of our most important pieces as well? To be at .500 with already three road wins on the board, you should give a little credit.

Why should I mention them? Are they coming back soon? Or is this team the one we have for this season?

Because you're dismissing a win based on Horford not playing. I don't even know whether he's coming back soon or not, and it doesn't matter. The fact of the matter is that in that game, we played without Bogut, they played without Horford, and we won a game we could certainly have lost. Unless you can tell me that nobody we play going forward is going to have any injuries, then it matters; I'm not ready to say, "Sure, we've played well enough to win so far. But going forward we'll be missing Bogut, our opponents will be missing nobody, and therefore we won't be able to keep up what we've done so far". The whole point is that right now we're doing better than I would have expected given the injuries and slumps we have.

Most fanbases are goofy because they're convinced they are a championship team, and they dismiss every loss with excuses. Not sure I've ever seen a fanbase try so hard to dismiss all their wins to convince themselves their team is horrible.

[floppymoose]

You folks aren't grocking my point. I'm talking about our opponents, and how impressive our wins have been. Saying that Horford is going to come back later doesn't suddenly make our hawks win impressive.

That the hawks win wasn't impressive isn't a mark against the Warriors. It just means we have one less data point on how good the team really is.

As for players we are missing, I'm not convinced either are going to be a factor this year, so I'm rolling with what we have right now as our team.

I'm really trying to evaluate the bar set by each victory. The Clippers had a player missing too, but I didn't mention it because they were still a tough opponent.

[uteptwostep]

Assuming Bogut won't be a factor the entire season feels really pessimistic to me. Maybe I'm just in denial.

Your point about the quality of our opponents so far remains valid though. In our defense, lots of teams seem to be hurting right now, and many of those injuries are likely to persist at least as long as Bogut's.

[Mylie10]

You folks aren't grocking my point. I'm talking about our opponents, and how impressive our wins have been. Saying that Horford is going to come back later doesn't suddenly make our hawks win impressive.

That the hawks win wasn't impressive isn't a mark against the Warriors. It just means we have one less data point on how good the team really is.

As for players we are missing, I'm not convinced either are going to be a factor this year, so I'm rolling with what we have right now as our team.

I'm really trying to evaluate the bar set by each victory. The Clippers had a player missing too, but I didn't mention it because they were still a tough opponent.

I'm not trying to rail you at all Flop..me love you.

I was merely pointing it out. Because i believe that Bogut will play, and even if they just settle on a lower minute total, he's shown to be way effective.

I realize that Varajeo and Horford would have made differences in those games, but so would have more Bogut for us. Now Jefferson's out and we're screwed

My other point was that in several of our wins and losses we shot the ball terribly, and I believe that we are a much better shooting team than the numbers would say right now.

I think that the offense takes longer to gel than defensive effort does. Chemistry takes much longer, so that may contribute to the shooting woes. I mean how many little flip shots has David Lee missed that he would normally have go down? Lots. Same with Curry earlier on...Even Klay is having chip shots roll out.

[floppymoose]

Bogut could absolutely come back and make us a lot better. It won't change the quality of our existing wins, though.

[Mylie10]

At the end of the year, no-one looks at existing wins...they look at total wins.

I get your point, but it just comes off as a dig. You're right though in that it's been ugly at times. But that's what defense and rebounding give you. We could go back to the Nellie style and just shoot wide open looks in the first 7 seconds of the shot clock all day....but would we win those ugly games?

The most telling thing to me, is the 4 road wins, and the ability to win some of these close games. The young players are really showing the pension for making the right plays...it may look ugly, but it's working.

[floppymoose]

I'm not really trying to suggest the team is doing something wrong, needs a different coach, etc. I just feel like the current 6-5 record hasn't told me a lot about the team. I want to see more games against solid opponents who aren't missing key pieces, so that I have a better yardstick to measure where we are at.

[Mylie10]

I'm getting that this team will battle better than previous teams. They are playing tougher, and our 2 rookies are the perfect example of giving great effort and seemingly not worried about numbers or touches.

The road wins tell me something because it's just flat out tough to win on the road in the NBA.

We're only 11 games in though, and things can change. My hope is that with gelling and better shooting we should improve on what we're already doing, not get worse. But we will hit rough patches no doubt.

[Mylie10]

Bogut may be coming back soon...per report...will miss next three games, but will re-evaluate after.

[paul]

Bogut may be coming back soon...per report...will miss next three games, but will re-evaluate after.

Yeah after the next 3 games you have like 5 days off so that always seemed like the logical time to bring him back I guess. Hopefully it's going well.

[Franc]

If all works well , he would start on the 29th against Denver at home. Not bad start. Keep fingers crossed!!!!!

[EvanZ]

That helps. My gut says that just including minutes in the calculation will skew simply towards guys who play a lot of minutes without applying any qualitative screen for the quality of those minutes. How would you integrate a qualitative element so that we aren't crediting good and bad minutes equally?

Do you mean point differential, i.e. garbage time? That's the win probability type of model.

[EvanZ]

Aside from the Lakers and OKC, haven't we pretty much had an opportunity to win every single game? Was there another blowout I'm missing?