At the Allstar break, it's clear that West is Best and East is Least

[Fafnir]

East vs Elite Teams: 51%
West vs Elite Teams: 48%

East vs West Elite Teams: 52%
West vs East Elite Teams: 48%

So who really cares if the bottom tier of the West is better than the bottom tier of the East if the East's elite post better numbers against elite teams from both conferences? When it comes time for playoff basketball no one cares if Phoenix had a higher winning % than Milwaukee or Houston was far superior to New Jersey?

The NBA is generally 5 or 6 teams and a bunch of filler. The West's filler is much better than the East's this year and in most years recently, but the elites generally play each other to a combined .500 record. 

So was this with Roy's (slanted in favor of the east) definition of elite?

What's your definition of elite?

[nba is the worst]

They do?



And who cares about the entire league.  There are 3 (maybe 4) teams that matter in the East.  (Boston, Miami, Orlando, Chicago) and 3 teams out West (SA, LA, Dallas)  The rest is just junk to wannabes.  


How do those team do against each other?

Glad you asked.

Boston is 8-4: 1-0 vs Spurs, 0-2 vs Mavs, 1-1 vs LA, 3-0 vs heatles, 1-1 vs Bulls, and 2-1 vs Magic (who don't deserve to be in this list IMO).  

Heatles are 3-6: 0-0 Spurs, 0-2 Mavs, 1-0 LA, 0-3 Celts, 0-1 Bulls, 2-1 Magic

Bulls are 7-6: 1-1 Spurs, 2-0 Mavs, 1-1 LA, 1-0 heatles, 1-2 Celts, 1-2 Magic

Magic are 6-7: 1-1 Spurs, 1-1 Mavs, 1-0 LA, 1-2 Celts, 1-2 heatles, 1-1 Bulls

Spurs are 5-4: 1-1 Mavs, 2-0 LA, 0-1 Celts, 0-0 heatles, 1-1 Bulls, 1-1 Magic

Mavs are 7-4: 1-1 Spurs, 1-0 LA, 2-0 Celts, 2-0 heatles, 0-2 Bulls, 1-1 Magic

LA is 2-7: 0-2 Spurs, 0-1 Mavs, 1-1 Celts, 0-1 heatles, 1-1 Bulls, 0-1 Magic (only reason they belong on this list is the B2B titles)

Celtics do lead the list:

Celts 8-4
Mavs 7-4
Spurs 5-4
Bulls 7-6
Magic 6-7
heatles 3-6
LA 2-7  

So I guess you come in on the side that beating other elite teams tells you something? http://www.basketball-reference.com/blog/?p=8811

es, I'm not a big believer in stats that focus on avg differential as their primary metric, like SRS (with what is it, 54% predictive accuracy) or the blog you mention above where the difference between all the variables is 7.3% (in the noise). You claim the bball-ref.com factors SOS into their power ratings, but as they use differential there as well I'm not convinced - particularly when the numbers put the heatles in 1st ahead of the Spurs and celtics.

When trying to compare conference strength - as you did with the arbitrary 4-pt giving the east a 4-2 advantage, doesn't it make sense that with over 2X the qty of below .400 teams, the east point differentials will be inflated vs the West's?

The biggest factors to consider IMO are health of the team when they played during the season vs health in the playoffs, HCA in the playoffs, and overall W-Lrecord vs ALL teams.

So I would definitely include the Thunder with their 2-games better record than the Magic as "elite" - or alternatively restrict the definition to just the Spurs, Celtics, and Mavs...

[Fafnir]

I still don't understand why you don't trust differentials, its proven better than W-L records.

SRS/Differential accurately predicted how the Knicks would crater right back to .500 just this season after all.

I don't trust teams that have built their records by winning a ton of very close games. Are the Mavs a better team than the C's because KG missed a wide open 18 foot jumper late and Kidd made a wide open three point shot?

[nba is the worst]

I still don't understand why you don't trust differentials, its proven better than W-L records.

SRS/Differential accurately predicted how the Knicks would crater right back to .500 just this season after all.

I don't trust teams that have built their records by winning a ton of very close games. Are the Mavs a better team than the C's because KG missed a wide open 18 foot jumper late and Kidd made a wide open three point shot?

Sorry, you're too quick for me - I forgot to add my point I just edited in above.

Sure, there are cases like the Knicks that support the stats guys at bball-ref.com, but to claim that a few percent "proves" something without other more important variables like health in the Finals I don't buy.

As I just edited in my previous post, when comparing conferences where there's so much difference in strength of the bottom half teams, the differential for the elite will be inflated for teams in the conference with 7 crap teams, and deflated for those in the conference with 3 crap teams.

[Fafnir]

I still don't understand why you don't trust differentials, its proven better than W-L records.

SRS/Differential accurately predicted how the Knicks would crater right back to .500 just this season after all.

I don't trust teams that have built their records by winning a ton of very close games. Are the Mavs a better team than the C's because KG missed a wide open 18 foot jumper late and Kidd made a wide open three point shot?

Sorry, you're too quick for me - I forgot to add my point I just edited in above.

Sure, there are cases like the Knicks that support the stats guys at bball-ref.com, but to claim that a few percent "proves" something without other more important variables like health in the Finals I don't buy.

As I just edited in my previous post, when comparing conferences where there's so much difference in strength of the bottom half teams, the differential for the elite will be inflated for teams in the conference with 7 crap teams, and deflated for those in the conference with 3 crap teams.

Except that the metrics I cited adjust for SOS, so that beating up on a few extra bottom feeders doesn't inflate your rating.

I'll also add that the extra (what 5?) games against crappier teams comprises overall is a small portion of an NBA team's schedule. Much more important in the NBA is how your back to backs are arranaged along with the four games in five nights.

[nba is the worst]

I still don't understand why you don't trust differentials, its proven better than W-L records.

SRS/Differential accurately predicted how the Knicks would crater right back to .500 just this season after all.

I don't trust teams that have built their records by winning a ton of very close games. Are the Mavs a better team than the C's because KG missed a wide open 18 foot jumper late and Kidd made a wide open three point shot?

Sorry, you're too quick for me - I forgot to add my point I just edited in above.

Sure, there are cases like the Knicks that support the stats guys at bball-ref.com, but to claim that a few percent "proves" something without other more important variables like health in the Finals I don't buy.

As I just edited in my previous post, when comparing conferences where there's so much difference in strength of the bottom half teams, the differential for the elite will be inflated for teams in the conference with 7 crap teams, and deflated for those in the conference with 3 crap teams.

Except that the metrics I cited adjust for SOS, so that beating up on a few extra bottom feeders doesn't inflate your rating.

I'll also add that the extra (what 5?) games against crappier teams comprises overall is a small portion of an NBA team's schedule. Much more important in the NBA is how your back to backs are arranaged along with the four games in five nights.

Am I incorrect in my belief that the SOS in those metrics includes differential as part of the SOS rating?

And how does playing either four or three times per season against 7 sub-.400 teams vs a west team playing four or three times per season against just three sub-.400 teams amount to just 5 games difference?

[Fafnir]

I still don't understand why you don't trust differentials, its proven better than W-L records.

SRS/Differential accurately predicted how the Knicks would crater right back to .500 just this season after all.

I don't trust teams that have built their records by winning a ton of very close games. Are the Mavs a better team than the C's because KG missed a wide open 18 foot jumper late and Kidd made a wide open three point shot?

Sorry, you're too quick for me - I forgot to add my point I just edited in above.

Sure, there are cases like the Knicks that support the stats guys at bball-ref.com, but to claim that a few percent "proves" something without other more important variables like health in the Finals I don't buy.

As I just edited in my previous post, when comparing conferences where there's so much difference in strength of the bottom half teams, the differential for the elite will be inflated for teams in the conference with 7 crap teams, and deflated for those in the conference with 3 crap teams.

Except that the metrics I cited adjust for SOS, so that beating up on a few extra bottom feeders doesn't inflate your rating.

I'll also add that the extra (what 5?) games against crappier teams comprises overall is a small portion of an NBA team's schedule. Much more important in the NBA is how your back to backs are arranaged along with the four games in five nights.

Am I incorrect in my belief that the SOS in those metrics includes differential as part of the SOS rating?

And how does playing either four or three times per season against 7 sub-.400 teams vs a west team playing four or three times per season against just three sub-.400 teams amount to just 5 games difference?

I don't draw the line soley at .400 records, that's your thing.

Every team plays everybody twice, having two or three more bottom feeders just means that you have 5 or 6 games against easier opponents.

Of course you also have more games against the Bulls/Heat/Magic/Celtics.

[nba is the worst]

I still don't understand why you don't trust differentials, its proven better than W-L records.

SRS/Differential accurately predicted how the Knicks would crater right back to .500 just this season after all.

I don't trust teams that have built their records by winning a ton of very close games. Are the Mavs a better team than the C's because KG missed a wide open 18 foot jumper late and Kidd made a wide open three point shot?

Sorry, you're too quick for me - I forgot to add my point I just edited in above.

Sure, there are cases like the Knicks that support the stats guys at bball-ref.com, but to claim that a few percent "proves" something without other more important variables like health in the Finals I don't buy.

As I just edited in my previous post, when comparing conferences where there's so much difference in strength of the bottom half teams, the differential for the elite will be inflated for teams in the conference with 7 crap teams, and deflated for those in the conference with 3 crap teams.

Except that the metrics I cited adjust for SOS, so that beating up on a few extra bottom feeders doesn't inflate your rating.

I'll also add that the extra (what 5?) games against crappier teams comprises overall is a small portion of an NBA team's schedule. Much more important in the NBA is how your back to backs are arranaged along with the four games in five nights.

Am I incorrect in my belief that the SOS in those metrics includes differential as part of the SOS rating?

And how does playing either four or three times per season against 7 sub-.400 teams vs a west team playing four or three times per season against just three sub-.400 teams amount to just 5 games difference?

I don't draw the line soley at .400 records, that's your thing.

Every team plays everybody twice, having two or three more bottom feeders just means that you have 5 or 6 games against easier opponents.

Of course you also have more games against the Bulls/Heat/Magic/Celtics.

It's 4 more bottom feeders, not "2 or 3"!
(At least 10 to 12 games to pad the differentials).

Even if you disregard anything but ".500+" vs "below .500", the West has 10 teams, the east has six...

[Fafnir]

It's 4 more bottom feeders, not "2 or 3"!
(At least 10 to 12 games to pad the differentials).

Even if you disregard anything but ".500+" vs "below .500", the West has 10 teams, the east has six...

*shrug* you define teams how you want, I've given my point of view. Feel free to ignore it, just as you ignore what "adjusted for strength of schedule" means.

I look forward to your repeat post in a month on this same topic. I'll give my point of view then as well.

[nba is the worst]

It's 4 more bottom feeders, not "2 or 3"!
(At least 10 to 12 games to pad the differentials).

Even if you disregard anything but ".500+" vs "below .500", the West has 10 teams, the east has six...

*shrug* you define teams how you want, I've given my point of view. Feel free to ignore it, just as you ignore what "adjusted for strength of schedule" means.

I look forward to your repeat post in a month on this same topic. I'll give my point of view then as well.

Thanks for your input.

It would be nice to know why you think metrics that put a heatles team above two teams with better records including one that is 3-0 head-to-head against them are "Proven" better than W-L records.

I also didn't see a response from you as to whether bball-reference.com's SOS ratings are derived using differential as a basis - I'd be interested in that answer.

[Fafnir]

It's 4 more bottom feeders, not "2 or 3"!
(At least 10 to 12 games to pad the differentials).

Even if you disregard anything but ".500+" vs "below .500", the West has 10 teams, the east has six...

*shrug* you define teams how you want, I've given my point of view. Feel free to ignore it, just as you ignore what "adjusted for strength of schedule" means.

I look forward to your repeat post in a month on this same topic. I'll give my point of view then as well.

Thanks for your input.

It would be nice to know why you think metrics that put a heatles team above two teams with better records including one that is 3-0 head-to-head against them are "Proven" better than W-L records.

I also didn't see a response from you as to whether bball-reference.com's SOS ratings are derived using differential as a basis - I'd be interested in that answer.

If you'd read the links I gave you, you'd have the answers.

Here's another: http://www.pro-football-reference.com/blog/?p=37.

Those all go into detail about the models and their predictive power. The fact that the Heat are ranked higher than the C's isn't surprising, despite the C's success head to head the Heat have been more successful against the rest of the league overall. No league wide robust model deals with head to head match ups, you'd have 435 pairs to worry about.

[nba is the worst]

It's 4 more bottom feeders, not "2 or 3"!
(At least 10 to 12 games to pad the differentials).

Even if you disregard anything but ".500+" vs "below .500", the West has 10 teams, the east has six...

*shrug* you define teams how you want, I've given my point of view. Feel free to ignore it, just as you ignore what "adjusted for strength of schedule" means.

I look forward to your repeat post in a month on this same topic. I'll give my point of view then as well.

Thanks for your input.

It would be nice to know why you think metrics that put a heatles team above two teams with better records including one that is 3-0 head-to-head against them are "Proven" better than W-L records.

I also didn't see a response from you as to whether bball-reference.com's SOS ratings are derived using differential as a basis - I'd be interested in that answer.

If you'd read the links I gave you, you'd have the answers.

Here's another: http://www.pro-football-reference.com/blog/?p=37.

Those all go into detail about the models and their predictive power. The fact that the Heat are ranked higher than the C's isn't surprising, despite the C's success head to head the Heat have been more successful against the rest of the league overall. No league wide robust model deals with head to head match ups, you'd have 435 pairs to worry about.

I read them, but it's not clear to me what the following has to do with an SOS calculation:
"Home ORtg = Lg Avg  ORtg + .5*HCA + Home Offense Score + Away Defense Score

Home DRtg = Lg Avg ORtg - .5*HCA + Away Offense Score + Home Defense Score"

I couldn't find where they mathematically define their method of calculating SOS.
In terms of W-L record, it's the Celtics who have been "more successful against the rest of the league overall"

[Fafnir]

It's 4 more bottom feeders, not "2 or 3"!
(At least 10 to 12 games to pad the differentials).

Even if you disregard anything but ".500+" vs "below .500", the West has 10 teams, the east has six...

*shrug* you define teams how you want, I've given my point of view. Feel free to ignore it, just as you ignore what "adjusted for strength of schedule" means.

I look forward to your repeat post in a month on this same topic. I'll give my point of view then as well.

Thanks for your input.

It would be nice to know why you think metrics that put a heatles team above two teams with better records including one that is 3-0 head-to-head against them are "Proven" better than W-L records.

I also didn't see a response from you as to whether bball-reference.com's SOS ratings are derived using differential as a basis - I'd be interested in that answer.

If you'd read the links I gave you, you'd have the answers.

Here's another: http://www.pro-football-reference.com/blog/?p=37.

Those all go into detail about the models and their predictive power. The fact that the Heat are ranked higher than the C's isn't surprising, despite the C's success head to head the Heat have been more successful against the rest of the league overall. No league wide robust model deals with head to head match ups, you'd have 435 pairs to worry about.

I read them, but it's not clear to me what the following has to do with an SOS calculation:
"Home ORtg = Lg Avg  ORtg + .5*HCA + Home Offense Score + Away Defense Score

Home DRtg = Lg Avg ORtg - .5*HCA + Away Offense Score + Home Defense Score"

I couldn't find where they mathematically define their method of calculating SOS.
In terms of W-L record, it's the Celtics who have been "more successful against the rest of the league overall"

I suggest you re-read the article in detail:

How to read this table: in week 1, the Colts beat the Ravens by 17. The Ravens were, all things considered, 1.83 points worse than average, so the Colts got a "score" of 17 - 1.83, or 15.17 for that game. In week 2, the Colts beat the Jaguars by 7. Jacksonville was 4.76 points better than average, so the Colts get an 11.76 for that game. Average their scores for each game and you've got their rating. The bottom line says:

The Colts' won their games by an average of 12 points each. Their opponents were, on average, 1.2 points worse than average. Thus the Colts were 10.8 points better than average.

That's just one section, it goes into details that answer your question.

So every team's rating is their average point margin, adjusted up or down depending on the strength of their opponents. Thus an average team would have a rating of zero. Suppose a team plays a schedule that is, overall, exactly average. Then the sum of the terms in parentheses would be zero and the team's rating would be its average point margin. If a team played a tougher-than-average schedule, the sum of the terms in parentheses would be positive and so a team's rating would be bigger than its average point margin.

It would be easy to find the Colts' rating if we knew all their opponents' ratings. But we can't figure those out until we've figured out their opponents' ratings, and we can't figure those out until. . ., you get the idea. Everyone's rating essentially depends on everyone else's rating.

So how do you actually find the set of values that solves this system of equations? In high school you probably learned how to solve 2-by-2 and maybe 3-by-3 systems of equations by putting some numbers into a matrix, doing some complicated operations on that matrix, and then reading the solutions off the new matrix. Same thing here, except you've got a 32-by-32 matrix instead of a 2-by-2 matrix. If you wanted college football rankings, it'd be 120-by-120. I recommend using a computer.

It's more instructive, though, to solve it a different way. We'll start by giving everyone an initial rating, which is just their average point margin. I'll use the Colts as an example. Their initial rating is +12.0. Now look at the average of their opponents' intial ratings:

[PosImpos]

The West has more decent teams, fewer awful teams.

The East has more great teams, more awful teams.

Whether you think the East or the West is best depends on what you think matters most.

[nba is the worst]

The West has more decent teams, fewer awful teams.

The East has more great teams, more awful teams.

Whether you think the East or the West is best depends on what you think matters most.

It also depends on your definition of "great teams"...

In that case, I'd say it's the Spurs, Celtics and Mavs - 2-1 West!

Of course, those on the "It's not wins that matter, it's point differential would say the Mavs differential isn't high enough.

(Fafnir, my post above came before I read the last link from pro-football-ref.com, sorry to waste your time. I would certainly say that this method of adjusting for SOS as a predictive model - that is only accurate 54% of the time - isn't particularly earth-shaking).