PDA

View Full Version : Halfway through: a look at the stats, reprise



Seventyniner
01-18-2017, 12:39 PM
Damn, it's already been 4 years since I did this last? I had meant to make it a yearly thing, guess I'm just that lazy 75% (or more) of the time.
Reference: http://www.spurstalk.com/forums/showthread.php?t=208236

Anyway, we once again are at the halfway point of the regular season on January 18. With a record of 32-9 the Spurs are second in the West to the 35-6 Warriors, and leading the Southwest division over the 32-12 Rockets. With a scoring margin that ranks second in the league the Spurs are a force to be reckoned with, cliffjumpers and pessimists be damned.

This time I will use stats for the Warriors, Cavs, Raptors, Clippers, Rockets, Thunder, and Spurs for this year and last year. I am comparing half-season stats this year to full-season stats last year; I couldn't find all these stats at last season's halfway point.

See the linked post for an explanation of the four factors and how they're weighted. This next part includes

Margin of victory (MOV) - average scoring margin, is negative for teams that have been outscored on the season
Strength of schedule (SOS) - (I'm not exactly sure how this is calculated, bkref's link to the methodology is outdated)
Simple Rating System (SRS) - sum of MOV and SOS, a simple way to rate teams by adjusting scoring margin for schedule strength
Offensive rating (ORtg) - points scored per 100 possessions (i.e. pace-adjusted points scored per game)
Defensive rating (DRtg) - points allowed per 100 possessions (i.e. pace-adjusted points allowed per game)
Pace - how many possessions per game a team averages
Three point attempt rate (3PAr) - the proportion of a team's field goal attempts that are three-point attempts

OVERALL STATS
(league rank in parentheses)


TEAM
Year
MOV
SOS
SRS
ORtg
DRtg
Rtg diff
Pace
3PAr


SA
2017
8.71 (2)
-0.18 (18)
8.52 (2)
113.8 (4)
104.5 (4)
9.3 (2)
93.9 (27)
0.273 (24)


GS
2017
12.61 (1)
-0.52 (25)
12.09 (1)
116.1 (1)
103.6 (1)
12.5 (1)
100.7 (2)
0.353 (7)


CLE
2017
4.85 (6)
-0.04 (15)
4.81 (6)
112.4 (5)
107.4 (14)
5.0 (6)
96.6 (15)
0.388 (2)


LAC
2017
6.16 (5)
-0.75 (29)
5.41 (5)
111.8 (6)
105.4 (6)
6.4 (5)
96.4 (17)
0.325 (10)


HOU
2017
6.95 (3)
-0.27 (20)
6.69 (4)
115.2 (3)
108.2 (16)
7.0 (4)
98.8 (4)
0.460 (1)


OC
2017
1.26 (10)
-0.02 (14)
1.24 (9)
108.2 (14)
106.9 (9)
1.3 (10)
97.5 (9)
0.302 (16)


TOR
2017
6.85 (4)
0.02 (13)
6.87 (3)
116.1 (2)
108.9 (21)
7.2 (3)
95.6 (20)
0.284 (21)


Average

0
0
0
108.2
108.2
0
96.3
0.313


SA
2016
10.63 (2)
-0.36 (25)
10.28 (2)
110.3 (4)
99.0 (1)
11.3 (1)
93.8 (24)
0.223 (25)


GS
2016
10.76 (1)
-0.38 (28)
10.38 (1)
114.5 (1)
103.8 (6)
10.7 (2)
99.3 (2)
0.362 (2)


CLE
2016
6.00 (4)
-0.55 (30)
5.45 (4)
110.9 (3)
104.5 (10)
6.4 (4)
93.3 (27)
0.352 (3)


LAC
2016
4.28 (6)
-0.15 (22)
4.13 (5)
108.3 (7)
103.8 (5)
4.5 (6)
95.8 (14)
0.324 (9)


HOU
2016
0.20 (15)
0.14 (12)
0.34 (15)
108.3 (8)
108.1 (21)
0.2 (15)
97.6 (7)
0.370 (1)


OC
2016
7.28 (3)
-0.19 (24)
7.09 (3)
113.1 (2)
105.6 (13)
7.5 (3)
96.7 (10)
0.275 (19)


TOR
2016
4.50 (5)
-0.42 (29)
4.08 (6)
110.0 (5)
105.2 (11)
4.8 (5)
92.9 (29)
0.287 (15)


Average

0
0
0
106.4
106.4
0
95.8
0.285



Observations



The Warriors are clearly the best team in the league. Not that we needed these stats for confirmation.
The Spurs' profile is that of a major contender; MOV over 8 and rating diff over 9 are historically league-leading numbers....except for the ridiculous run the Warriors are on.
The league has drastically increased in pace recently. League average this season is 96.3; the last time the league averaged over 96 possessions per game was 1992-1993.
The other league-wide factor that cannot be ignored is the explosion in ORtg. 108.2 is the highest league average since 1994-1995. This puts last year's historically good Spurs defense in some context. It could be a long time before we see a team post a DRtg under 100 for an entire season.
This ORtg explosion is almost entirely due to an increase in three-point attempts. The league averages 26.8 per game, a big acceleration over the steady uptick seen over the last 2 decades. We are truly in a new era.
Still, the top 5 offenses have 3PAr of 7th, 21st, 1st, 24th, 2nd. San Antonio and Toronto show that you don't have to take a ton of threes to have a great offense.
I included the Thunder and Rockets for a study in contrasts: the Thunder have (understandably) gone from contender to first-round fodder after losing Durant, while the Rockets have done the opposite on the strength of an elite offense and a slight uptick in defense.
The Rockets play a dangerous game, though; a regular season is long enough for variance in three-point makes to average out (it is a powerful high-risk, high-reward strategy), but they can shoot themselves into or out of any seven-game series. A dangerous team to be sure, but hard to stay hot enough to win a title with a league-average defense.
Disclaimer: I have watched very little of teams other than the Spurs. Saying that, I'm surprised at how bad Cleveland's defense is. Statistically they actually look a lot like the Rockets, and we've seen their Jekyll/Hyde performance in two games against the Warriors. In the first they dug a hole and then got hot to steal a win. In the second they were cold throughout and got blasted.
Toronto is even more of a glass cannon than Houston or Cleveland; it's shocking to me to see a team rank 3rd in SRS with the 21st-ranked defense. It makes the Spurs complete shutdown of them look even more impressive.
The Clippers keep chugging along, a top 5 team even with Blake having missed the last month. With Chris Paul gone for 6-8 weeks, I think they'll wind up 4th or 5th in the West. That would make for a tough second-round matchup for the Warriors....except that when Donald Sterling was removed as Clippers owner the title seems to have gone to the Warriors.


FOUR FACTORS
(league rank in parentheses)

TEAM
Year
O eFG%
D eFG%
eFG% diff
O TOV%
D TOV%
TOV% diff
ORB%
DRB%
Total RB%
O FT/FGA
D FT/FGA
FT/FGA diff


SA
2017
.538 (3)
.496 (4)
.042 (3)
12.4 (12)
13.4 (9)
1.0 (7)
23.3 (15)
77.8 (5)
101.1 (10)
.226 (6)
.202 (10)
.024 (7)


GS
2017
.567 (1)
.484 (2)
.083 (1)
13.4 (21)
13.0 (13)
-0.4 (21)
22.1 (21)
75.9 (19)
98.0 (23)
.214 (18)
.202 (10)
.012 (13)


CLE
2017
.526 (8)
.511 (17)
.015 (8)
12.6 (14)
12.7 (16)
0.1 (15)
23.9 (13)
75.4 (26)
99.3 (18)
.227 (5)
.169 (3)
.058 (3)


LAC
2017
.530 (4)
.492 (3)
.038 (4)
12.3 (11)
12.6 (17)
0.3 (13)
22.1 (20)
77.0 (12)
99.1 (19)
.231 (4)
.217 (21)
.014 (9)


HOU
2017
.553 (2)
.515 (22)
.038 (4)
13.7 (25)
13.2 (12)
-0.5 (22)
25.2 (9)
76.4 (15)
101.6 (8)
.215 (16)
.201 (9)
.014 (9)


OC
2017
.507 (16)
.505 (10)
.002 (11)
13.3 (19)
12.6 (18)
-0.7 (24)
26.3 (6)
77.5 (7)
103.8 (4)
.221 (9)
.208 (14)
.013 (11)


TOR
2017
.528 (7)
.514 (20)
.014 (9)
11.0 (1)
14.2 (3)
3.2 (1)
25.6 (7)
74.3 (29)
99.9 (15)
.246 (1)
.216 (20)
.030 (5)


Average
2017
.510
.510
0
12.8
12.8
0
23.4
76.6
100
.208
.208
0


SA
2016
.526 (2)
.477 (1)
.049 (2)
12.4 (9)
14.1 (8)
1.7 (4)
23.0 (23)
79.1 (3)
102.1 (8)
.197 (21)
.182 (3)
.015 (7)


GS
2016
.563 (1)
.479 (2)
.084 (1)
13.5 (20)
12.6 (20)
-0.9 (22)
23.5 (19)
76.0 (15)
99.5 (17)
.191 (25)
.208 (17)
-.017 (22)


CLE
2016
.524 (3)
.496 (13)
.028 (6)
12.7 (13)
12.6 (20)
-0.1 (16)
25.1 (9)
78.5 (5)
103.6 (3)
.194 (23)
.205 (16)
-.011 (20)


LAC
2016
.524 (3)
.480 (4)
.044 (3)
12.1 (4)
13.8 (10)
1.7 (3)
20.1 (28)
73.8 (28)
93.9 (29)
.220 (8)
.222 (24)
-.002 (18)


HOU
2016
.516 (6)
.516 (23)
.000 (15)
14.2 (28)
14.7 (2)
0.5 (11)
25.7 (6)
72.8 (30)
98.5 (23)
.244 (3)
.219 (22)
.025 (5)


OC
2016
.524 (3)
.484 (5)
.040 (4)
14.0 (24)
11.7 (27)
-2.3 (30)
31.1 (1)
76.0 (15)
107.1 (1)
.228 (5)
.205 (16)
.023 (6)


TOR
2016
.504 (12)
.498 (14)
.006 (10)
12.3 (8)
12.7 (18)
0.4 (12)
24.6 (12)
77.7 (7)
102.3 (7)
.255 (2)
.201 (10)
.054 (2)


Average
2016
.502
.502
0
13.2
13.2
0
23.8
76.2
100
.209
.209
0




Observations



The Warriors continue to shoot the lights out. I called the Spurs and Heat’s eFG% of over 0.540 “eye-popping” four years ago, and now we have the Warriors flirting with 0.570. Good lord.
The Warriors’ other factors are not impressive: below average in turnovers and rebounding, and just above it in FT/FGA. And yet they lead the league in ORtg. Just goes to show how much shooting matters.
To hammer that last point home, look at last year’s Warriors’ differentials: 22nd in TOV, 17 in RB%, 22nd in FT/FGA. Not what I’d ever expect from a 73-win team with a double digit point differential. Elite shooting will take you far.
It’s hard to judge the Kevin Durant effect with only these numbers, but the biggest changes for the Warriors are that they are worse at rebounding and better at getting to the line. Overall the numbers are mostly the same.
In contrast, the Spurs are top 10 in each differential category, the only team in the league that can make that claim. In fact, the only stat in this table in which this year’s Spurs are below league average is ORB%, and even then only by a hair. This is a very well-rounded team.
The Spurs’ rise from 21st to 6th in O FT/FGA is striking. It has more than offset the drop in D FT/FGA. Last night’s Wolves FT parade skewed this stat a bit.
Cleveland is strange: they’re only really elite at taking and preventing free throws. They have also gone from a great rebounding team last year to a mediocre one this year. This could very well be the key to beating the Warriors, similar to how the Thunder beat the Spurs last year.
The Clippers continue to be elite at shooting and defending shots and getting to the line, while hovering around average in other areas. I’m surprised to see them so low in the rebounding rank, though Blake Griffin’s injury hurts them.
Houston’s rise appears to be due to much better shooting and defensive rebounding. I thought their mantra was threes and free throws only, but I didn’t expect to see them drop from 3rd to 16th in O FT/FGA. Then again, it’s hard to draw a bunch of fouls when 46% of the shots you take are threes.
Toronto is the opposite of San Antonio here: either really good or really bad in almost every category. That TO differential is insane, but their defense sucks otherwise. But their offense is damn solid, 7/1/7/1 on the differential rankings is very hard to achieve.



Once again, stats don't tell the whole story. Use them to supplement the good old "eye test." This year's Spurs team has been frustrating to watch sometimes; I attribute that to a letdown from last year's dominant regular season. But don't hang your head. This is a damn good team and won't be an easy out for anyone. To me only the Warriors are clear favorites over the Spurs right now.

Last time I said I'd do this analysis again at the end of the season. Let's just say I won't make any promises this time! I found a faster way to update the table so if I get back to this in April it won't take too much effort.

dabom
01-18-2017, 12:54 PM
Great job OP. :tu

Spur|n|Austin
01-18-2017, 01:08 PM
Awesome post Seventyniner, good stuff.

Warrior's eFG% of .567 :wow

Chinook
01-18-2017, 01:22 PM
It was worth the wait, 79er. Stats don't tell the whole story, but they can shed light on trends that eye-tests really can't. In that regard, I think your observations about how the league has evolved over the last year are the most important. However, it's nice to see that the Spurs are a team with fewer weaknesses than some posters believe. And to think they're very much coasting still, even if RealGM fans think they're "sprinting".

BillMc
01-18-2017, 01:48 PM
Damn impressive post OP. Harkens back to ST's golden age of analysis. Many thanks. :clap

Cry Havoc
01-18-2017, 02:31 PM
The raptors are amazing. They're almost certainly a paper tiger, but the fact that they are beating the vaunted Dubs in ORTG is absolutely incredible.

Robz4000
01-18-2017, 04:19 PM
Thanks for this :tu

SAGirl
01-18-2017, 04:40 PM
Great post! The regular season is to develop chemistry, integrate players and for Pop to sharpen every tool in the shed!

Anyways, much thanks.
:flag:

Dex
01-18-2017, 07:41 PM
Great post OP. Some excellent and insightful content.

Hoops Czar
01-18-2017, 07:43 PM
Great post! The regular season is to develop chemistry, integrate players and for Pop to sharpen every tool in the shed!

Anyways, much thanks.
:flag:

It's also for winning as many games as you can so you don't drop from the #2 seed to the #6 seed on the last day of the season.

Uriel
01-19-2017, 04:16 AM
Once again, stats don't tell the whole story. Use them to supplement the good old "eye test."

Why do analytics gurus constantly have to make this disclaimer and bend over backwards to the "eye test" crowd?

We are entering a new age. Statistics and big data have become the hallmark of the modern NBA. This is readily apparent in how many teams this season have shifted to more efficient shots like layups and corner three pointers.

Stat geeks shouldn't have to hedge against their data-driven analysis by making this disclaimer all the time. It's true that stats don't tell the whole story. But they certainly tell a lot more than just the "eye test" does.

cutewizard
01-19-2017, 05:25 AM
Impressive analysis, thank you!

I wish that I had taken my Statistics classes seriously, hahahaha

:bobo

daledondale
01-19-2017, 06:42 AM
Great post OP. Some excellent and insightful content.

Chinook
01-19-2017, 07:48 AM
Why do analytics gurus constantly have to make this disclaimer and bend over backwards to the "eye test" crowd?

We are entering a new age. Statistics and big data have become the hallmark of the modern NBA. This is readily apparent in how many teams this season have shifted to more efficient shots like layups and corner three pointers.

Stat geeks shouldn't have to hedge against their data-driven analysis by making this disclaimer all the time. It's true that stats don't tell the whole story. But they certainly tell a lot more than just the "eye test" does.

Because advanced stats are only as good as their fit to real life. The eye-test doesn't trump stats in one-off cases for the most part, but if the stat is constantly giving results that make no sense to folks who can actually see what's going on, then yeah, the stat gets called into question.

That said, I wouldn't call any of the stats 79er present "advanced" per se. By that I mean that they are not opinion-driven metrics like PER or win-shares.

elemento
01-19-2017, 08:05 AM
That's a damn good job man :toast

Seventyniner
01-19-2017, 09:08 AM
Why do analytics gurus constantly have to make this disclaimer and bend over backwards to the "eye test" crowd?

We are entering a new age. Statistics and big data have become the hallmark of the modern NBA. This is readily apparent in how many teams this season have shifted to more efficient shots like layups and corner three pointers.

Stat geeks shouldn't have to hedge against their data-driven analysis by making this disclaimer all the time. It's true that stats don't tell the whole story. But they certainly tell a lot more than just the "eye test" does.

I see your point, and I tend to view basketball (among many other things!) mainly in terms of numbers. But in the end I see stats as just another lens, albeit one that's less prone to human error than pure observation and memory.

Seventyniner
01-19-2017, 09:14 AM
Because advanced stats are only as good as their fit to real life. The eye-test doesn't trump stats in one-off cases for the most part, but if the stat is constantly giving results that make no sense to folks who can actually see what's going on, then yeah, the stat gets called into question.

That said, I wouldn't call any of the stats 79er present "advanced" per se. By that I mean that they are not opinion-driven metrics like PER or win-shares.

You put that first part better than I could. "Trust, then verify."

"Advanced" is as much a descriptor as "young" or "tall" in that it's rather subjective. But yes, I like using the stats in my post because they're objective. For individual players, though, things get much fuzzier.

I would love for there to be a way to adjust these numbers for injury and rest, i.e. what a team's profile would look like if they were perfectly healthy for the whole season. That's quite the undertaking and I'm not sure I have the time for it, unless RC wants to give me a job :lol

JuneJive
01-19-2017, 12:38 PM
Kudos

DPG21920
01-19-2017, 01:23 PM
Much love

TheDoctor
01-19-2017, 01:37 PM
Pure awesomeness OP :toast

Brazil
01-19-2017, 01:53 PM
Solid stuff niner...

:bobo

SAGirl
01-19-2017, 03:37 PM
Because advanced stats are only as good as their fit to real life. The eye-test doesn't trump stats in one-off cases for the most part, but if the stat is constantly giving results that make no sense to folks who can actually see what's going on, then yeah, the stat gets called into question.

That said, I wouldn't call any of the stats 79er present "advanced" per se. By that I mean that they are not opinion-driven metrics like PER or win-shares.

I love looking at stats s but that doesn't trump my personal observation of the game... they exist in context... you piece it all together when you observe the games first then look at stats to see trends and confirm or dispel some bias... but they always exist in context.

Its why you place guys in different situations, teams, or in abstract and they may not be the same. Superstars probably always translate anywhere but role players work in synergy with others and depend on system or others to recreate situations in which they can succeed... etc. Stats never tell the whole story IMO.