And the winner is...(or will be)

[Crow]

Who has best chance at 2015 title?

I'd subjectively start with best 5 in east and 8 in west. Then I'd probably apply offensive and defensive efficiency and balance criteria. I'll do that later today or tomorrow. I think I have it figured out already though. Golden State. Many could win, but I think they are most likely. Not a radical view, but I am saying it. With the intention of following up more systematically later to be sure and to name the other 2-4 top contenders.

Have your own method and / or pick? Feel free to lay it out, if you want to.

[nbacouchside]

I think the Hawks might have the best chance, given the thinness of the East & how dominant they've become. Overall, the Warriors are easily the best team, but have a much tougher road in the playoffs.

[Crow]

Statistically that might be true given east - west imbalance. Last night in thinking about this post I had Atl and Sas as top contenders to GSW. GSW has yet to lose a single game to an eastern team, so that influenced me.

[Mike G]

Right now at b-r.com, forecast sez

Code: Select all

GSW   .418
Atl   .154
Tor   .115
Por   .077
LAC   .067
Dal   .046
Hou   .027
Mem   .026
Chi   .022
SAS   .017
Was   .014
Mil   .009
Cle   .006
Phx   .001
OKC   .001   

If I could get 1000:1 odds on the Thunders, I'd put a dollar there.

[Crow]

Applying a 108 Offensive efficiency threshold could push 3 teams out of top contention. A 104 defensive efficiency standard could knock out 4. But if you allow very strong performance on one to compensate for missing by a bit on the other 4 get saved. This approach only knocks out Wiz and Bulls. I saved Spurs from a mild TKO because of the injuries. In years past this method knocked more teams out. More legit contenders this time.

So I end up looking at SRS and performance against top 10. If you reset the criteria as over plus 4 on SRS and a winning record against top 10, the three qualifiers are GSW, Atl, and Tor. But subjectively I just can't ride with Toronto's defense and I won't count on the reigning champs from the top 3 so I am sitting on GSW at top, Atl and Sas next. Toronto might be 4th but I am not splitting those hairs at the moment.

Of course this is crude, zig-zag and could change. Mainly just wanted to open the topic up.

[Mike G]

Teams > .500 vs. Top 10 according to Sagarin:
Atl 8-3
GS 6-3
Tor 5-3
Mem 7-5
Chi 7-6
SA 8-7

http://www.usatoday.com/sports/nba/sagarin/
According to him (and SRS), the Clipps are the (distant) 2nd best team; they're 5-8 vs top 10 teams.

[Crow]

Mem and Chi being top 6 here is a plus. I wouldn't project a team less than top 6 against top 10 as winner. I think my past research showed the title winner us more often top 2 on this than not (or it might be almost always. I forget.)

[AcrossTheCourt]

If I could get 1000:1 odds on the Thunders, I'd put a dollar there.

Vegas isn't that stupid. I see one site that has them with the fifth best odds. The b-ref projections systems, instead, are pretty stupid how they're set-up -- not adjusted for who's in the rotation and who they have for the rest of the season.

Applying a 108 Offensive efficiency threshold could push 3 teams out of top contention. A 104 defensive efficiency standard could knock out 4. But if you allow very strong performance on one to compensate for missing by a bit on the other 4 get saved. This approach only knocks out Wiz and Bulls. I saved Spurs from a mild TKO because of the injuries. In years past this method knocked more teams out. More legit contenders this time.

So I end up looking at SRS and performance against top 10. If you reset the criteria as over plus 4 on SRS and a winning record against top 10, the three qualifiers are GSW, Atl, and Tor. But subjectively I just can't ride with Toronto's defense and I won't count on the reigning champs from the top 3 so I am sitting on GSW at top, Atl and Sas next. Toronto might be 4th but I am not splitting those hairs at the moment.

Of course this is crude, zig-zag and could change. Mainly just wanted to open the topic up.

Prove that SRS/win record against the top team matters more than the record versus everyone else. Otherwise it's just unsubstantiated conjecture. I haven't seen any evidence.

Last year the Spurs were a lot worse versus top teams, but then they dismantled everyone in the playoffs.

[Crow]

SRS as conjecture probably has more support than anything.

Spurs were weak against top 10 last season then came back to do well by end of season, for a 20-13 rs (by final sagarin top 10) which was best in league.

I'll look for the evidence suggesting against top 10 performance matters that I presented before. Haven't had time yet.

If you are not particularly onboard to using SRS plus against top 10, what would you use instead? Minutes weighted RPM is worthy of a check, with more time and focus. But I would have doubts that it is more reliable.

[BigLeagueInsights]

People always cite offense and defense rating when it comes to championships.

http://hangtime.blogs.nba.com/2014/06/1 ... pionships/
"Out of the top 10 is not where you want to be. Over the last 37 years (since the NBA started tracking turnovers in 1977-78), only three teams have won a championship after ranking outside the top 10 in defensive efficiency in the regular season. Twice as many champs have ranked outside the top 10 in offensive efficiency."

Here's a good breakdown:
http://www.sportingcharts.com/articles/ ... pions.aspx

[Crow]

Here is one set of posts, pretty far down the page. viewtopic.php?f=2&t=120&p=898&hilit=sagarin#p898

Top 2-3 against top 10 predicted pretty well, even after correcting for the inclusion of playoff data in Sagarin's ratings.

There may be other posts. I recall some other back and forth on this but can't find it.

I may re-do study for longer period and to make sure that it is pure regular season data when time permits.

[Crow]

Found this from a similar thread last season:


Of the 8 finalists over last 4 years, the average playoff efficiencies were 109.5 on offense and 104 on defense.

Hence my look at teams doing similar in regular season, even though I didn't end up finding that particularly useful this season.

[Crow]

In last 15 years, the team with the best regular season record against top 10 teams (according to final sagarin ratings) won at least 9 titles. There are 6 cases I'd have to look at further to determine the top record (difficult on my phone right now) but the title winner appears to have been at least in top 4 and may have been best in a couple more.

[Crow]

By far best against top 10 teams on Sagarin GSW, in the finals. Cavs third by a hair, in the finals.

http://www.usatoday.com/sports/nba/sagarin/

[Mike G]

According to that link, 5 teams have winning records vs top-10 teams:

Code: Select all

top10   W     L     W%
GSW    28    10    .737
Cle    20    12    .625
Atl    15    10    .600
Mem    22    18    .550
SAS    19    17    .528

Before the Cavs swept Atl, they were 16-12 (.571), and the Hawks were 15-6 (.714)