Retro POY '95-96 (Voting Complete)

[ElGee]

Let's just say, i find it a lot easier to trust my method i.e.
oppFG%, Opp TOV's, opp 3PA, opp 3P%, DRB

What? That is all covered by the points per possession, EVERYTHING of what you are using in your "method". Seriously, DRtg is really simple and the most effective way to say something about defense.

hmmmm

okay.

The biggest reason why I distrust defensive rating is that you can allow a high oppFG% and still end up with a good defensive rating, which is a no-no in my book.

But I will look deeper into this - maybe there is more to defense than just oppFG%, which I weigh really heavily.

It's because there are other things that contribute to points per possession outside of FG%. Fouling (FTA's) would be the largest. Think of it this way: Why would we want a defense who surrenders more points but always fouls (lower FG%) than a team who surrenders fewer points but never fouls (higher FG%)?

Heck, besides the raw points allowed, not fouling keeps starters out of foul trouble.

[mysticbb]

20th seems about right given their differential.

They were 10th after the 2003 season, but since then 10 teams had a better differential (like the Spurs, Celtics and even the 2007 Bulls). The Spurs 2004 have the biggest differential with 8.8 of All-Time. Interesting fact the Mavericks 2004 have the highest differential on offense of All-Time.

[Silver Bullet]

20th seems about right given their differential.

They were 10th after the 2003 season, but since then 10 teams had a better differential (like the Spurs, Celtics and even the 2007 Bulls). The Spurs 2004 have the biggest differential with 8.8 of All-Time. Interesting fact the Mavericks 2004 have the highest differential on offense of All-Time.

do you have the list ?

[mysticbb]

The Bulls for example.

44.8 fg% is high? Really?

[Silver Bullet]

The Bulls for example.

44.8 fg% is high? Really?

8th in the league for that season alone ?

[mysticbb]

do you have the list ?

Well, yes, part of that is in Dean Oliver's book page 47 (until 2003) and the rest since 2003 is in my head. ;)

[mysticbb]

8th in the league for that season alone ?

That is well below league average. Why does that count as "high"?

[Silver Bullet]

8th in the league for that season alone ?

That is well below league average. Why does that count as "high"?

because we're talking about an all time level....

For one of the greatest defensive teams of all time, it's high.

[mysticbb]

because we're talking about an all time level....

For one of the greatest defensive teams of all time, it's high.

Uh, that is not the way you can evaluate that. You need to look into the context of that season. We don't know how they would have played defensively maybe in 2004 when the average fg% was lower.
We are talking about an era which was offensively dominated more in the post with a lot of strong offensive center, thus you can expect that the fg% is higher. But on the other side they forced a lot of turnovers, the Bulls 1996 have a clearly higher opponents tov% than the 2004 Spurs for example.
Obviously they aren't the best defensive team ever, but they are 20th All-Time, that is really, really, really good. If you take a look at the playoffs, they even improved their DRtg.

[ElGee]

because we're talking about an all time level....

For one of the greatest defensive teams of all time, it's high.

Uh, that is not the way you can evaluate that. You need to look into the context of that season. We don't know how they would have played defensively maybe in 2004 when the average fg% was lower.
We are talking about an era which was offensively dominated more in the post with a lot of strong offensive center, thus you can expect that the fg% is higher. But on the other side they forced a lot of turnovers, the Bulls 1996 have a clearly higher opponents tov% than the 2004 Spurs for example.
Obviously they aren't the best defensive team ever, but they are 20th All-Time, that is really, really, really good. If you take a look at the playoffs, they even improved their DRtg.

Which gets back to my point about Pippen's value in the playoffs. League average was 0.8 lower than the regular season. The Bulls defense was 2.0 points lower. I give Pip some credit for that. -7.4 relative to league average in the playoffs is historically good.

Who's been better you might ask? (played more than 7 games)

88 Pistons (-7.7)
95 Knicks (-7.9)
95 Spurs (-7.5)
97 Heat (-8.0)
00 Heat (-9.8)
05 Pistons (-8.3)

That's it since 1980.

OK, next post will be about Robinson!

[Manuel Calavera]

A stat I really dont get is winshares. I know what it represents (estimated wins by a player), but I don't know why it's used when PER seems to be what it is only better.

Also, here's my list:

1. MJ
2. Malone
3. Drob
4. Payton
5. Pippen

[lorak]

Great post.

Hakeem and KG are the only two big men in NBA history who have ever been "swarmed". All other big men face single coverage.

[Doctor MJ]

A stat I really dont get is winshares. I know what it represents (estimated wins by a player), but I don't know why it's used when PER seems to be what it is only better.

Well here's one thing I'll say: PER measure's per possession impact normalized to 15 as the league average, WS measures total season impact unnormalized. So if a guy gets injured and misses a ton of games, I want there to be a stat that tries to measure total impact so that I know that the injured guy isn't making the league leaders unless he's off-the-charts-good.

There's actually a few different decision points here, and not only do PER & WS measure different things, but there are variants I'd like to see not covered by either.

Ah and then there's the more advanced stuff like adjusting based on actual assisted on % rather than league average (see APER on hoopdata.com and watch Nash's "production" jump from 16th in PER to 6th just by factoring that in more precisely), that are only not in there because the data wasn't recorded historically.

So yeah, it's a messy thing, and there's actually both not enough advanced stats (major things unanswered or approximated poorly) and too many (because people use them improperly).

[Silver Bullet]

do you have the list ?

Well, yes, part of that is in Dean Oliver's book page 47 (until 2003) and the rest since 2003 is in my head.

mystic if you don't mind - would you mind paraphrasing his explanation for using .475 as a coefficient for FTA.

[ElGee]

TrueLAFan made a point about David Robinson's defensive impact in the playoffs. Indeed, there seems to be some validity to that, if we believe that Robinson has a major influence on their defense as the anchor.

Here are the Spurs DRtg in his 93-96 stretch compared with regular season performance. The opponents ORtg is weighted by games played vs. team (eg a 6 game series counts twice as much as a 3 game series).

1996 111.5 (+8.0 vs. regular season) +4.7 relative to league average. Weighted opp avg. 112.1
1995 103.1 (-2.3 vs. regular season) -7.9 relative to league average. Weighted opp avg. 109.3
1994 110.3 (+5.7 vs. regular season) +5.3 relative to league average. Weighted opp avg. 108.6
1993 107.0 (+0.2 vs. regular season) -0.9 relative to league average. Weighted opp avg. 111.3

So it seems there is something to what TrueLAFan is saying. Perhaps with the exception of 1995's first 2 rounds...but we know what happened after that.

My general problem with Robinson, like everyone else, is his playoff performances. It's not that I viewed him as some sort of choker, just that he faired much better much better in a regular season style and against weaker competition. Which isn't too damning (we shouldn't overstate the unfortunate fact that he ran into Malone and Hakeem at the worst time). His regular seasons still count. I just know what type of contribution I'm getting toward a championship (he's a little too subject to certain styles and matchups, and when he runs into them there's a larger drop off then we see from other stars.)

1993 struggled against Duckworth/Williams vs. Portland. Then a better series against Phoenix (but a bad game 7).
1994 ran into Utah. Struggled.
1995 He DESTROYED LA (Dviac and Campbell) a round before playing Hakeem.
1996 great against Phoenix. Then Utah again...

That's something that I'll take into account, in general, for all of his prime seasons.

[tha_rock220]

1. Jordan
2. David Robinson
3. Hakeem Olajuwon
4. Gary Payton
5. Karl Malone

[Doctor MJ]

Last call. Please make clear any changes you make from this point.

[ElGee]

My 1996 POY Ballot:

1. Michael Jordan
2. Karl Malone
3. David Robinson
4. Anfernee Hardaway
5. Scottie Pippen

Most of my thoughts are in the thread. Jordan's alone at the top. Malone and Robinson in the next group, and Malone gets the nod based on much of what has been discussed already.

4-6 was between Penny, Pippen and Olajuwon. I'll rarely won't pick the big out of a group, but I thought Olajuwon's playoffs were basically the worst and that was the ultimate factor. I don't think there's really a wrong choice between these three. I thought Hakeem was the "best" player of the 3 at the time, but again, this is based on how they played that year.

[Doctor MJ]

'95-96 Results

Code: Select all

Player             1st 2nd 3rd 4th 5th Pts   POY Shares
1. Michael Jordan   24   0   0   0   0 240   1.000
2. David Robinson    0  11  10   3   0 136   0.567
3. Karl Malone       0  12   4   3   5 118   0.492
4. Anfernee Hardaway 0   0   5   8   5  54   0.225
5. Gary Payton       0   0   2   8   6  40   0.167
6. Hakeem Olajuwon   0   1   2   2   3  26   0.108
7. Scottie Pippen    0   0   1   0   4   9   0.019
8. Shawn Kemp        0   0   0   0   1   1   0.004


[Sedale Threatt]

Dream in sixth. Bastillon's head is going to explode.