clock menu more-arrow no yes

Filed under:

Projecting the Knicks' Standing using Point Differential

New, comments

John Hollinger predicted that the Knicks will finish the year with a mediocre record in a recent Insider article (read: $). Mediocre is relative, of course. For elite teams, seeding 5th-8th in the Eastern Conference should be considered mediocre. But for a team going through the type of turnaround the Knicks are experiencing, the mediocre feels pretty damn good.

Now Hollinger's clairvoyance didn't exactly cause the biggest stink as we've all heard the arguments about the weak schedule the Knicks enjoyed before this next stretch of games. But what you should care about is his citing of point differential as a predictor of success.

Additionally, I would submit the Knicks aren't as good as their current record for another reason -- they've outscored the opposition by only 0.9 points per game, which normally would leave a team just a game or so above .500 at this point in the season.

The Knicks have certainly been in their share of close games, both against the elite and the bottom feeders, but .500 feels about as complimentary as a "Don't Buy!" rating. With that, let's take a look at how accurate point differential is at predicting standings.

I took the offensive and defensive efficiency differentials of each team the last three full seasons and averaged out their winning percentages and raw standings.
DIFF RANK AVG. WIN % AVG. STAND.
1 .777 1.3
2 .740 2.0
3 .691 4.0
4 .679 4.0
5 .659 6.3
6 .687 7.3
7 .646 8.0
8 .634 8.0
9 .626 8.7
10 .626 9.0
11 .590 10.0
12 .602 11.0
13 .577 12.3
14 .537 14.3
15 .512 15.0
16 .516 15.3
17 .480 16.7
18 .459 18.3
19 .468 18.7
20 .439 19.7
21 .402 21.3
22 .382 22.7
23 .374 23.3
24 .305 25.0
25 .317 25.3
26 .305 25.7
27 .297 26.0
28 .276 27.0
29 .187 29.3
30 .211 29.3

A lot of numbers, I know, but the leftmost column is where teams rank in point differential, and you see its naturally high correlation to seeding.

What I did was take these average seeds, apply them to the current leaders in point differential, and take the difference between current seed and average seed to get a projected seed. Basically, where a team ranks in point differential should change their actual standing when all's said and one.

TEAM EFF. DIFF. CURR. STAND. AVG. STAND. DIFF. PROJ. STAND.
MIA
11.4
4 -2.7 1
SAS
9.6
1 1.0 2
BOS
9.2
2 2.0 4
CHI
5.6
5 -1.0 4
LAL
6.3
6 0.3 6
DAL
7.0
3 4.3 7
ORL
6.1
8 0.0 8
ATL
2.7
10 -2.0 8
UTH
2.4
7 1.7 9
OKC
3.0
9 0.0 9
DEN
2.1
11 -1.0 10
NOR
2.3
12 -1.0 11
NYK
2.1
13 0.7 12
POR
-0.3
14 0.3 14
PHI
-0.5
21 -6.0 15
HOU
0.6
15 0.3 15
PHO
-1.7
18 -1.3 17
MIL
-3.1
19 -0.7 18
MEM
-1.1
16 2.7 19
IND
-1.8
17 2.7 20
GSW
-4.6
20 1.3 21
TOR
-4.7
23 -0.3 23
CHA
-4.8
22 1.3 23
WAS
-6.9
28 -3.0 25
LAC
-5.0
25 0.3 25
DET
-5.1
24 1.7 26
NJN
-5.7
26 0.0 26
MIN
-7.2
27 0.0 27
SAC
-7.2
30 -0.7 29
CLE
-10.2
29 0.3 29

The "Average Standing Difference" (fourth column) gets added to the team's actual standing, so negative numbers are good here. The Heat, having the best differential, get moved from fourth to first in standing, while the third seeded team, Dallas, gets bumped all the way down to seventh.

Now for the reason you're all here. The Knicks are currently 13th in point differential. According to the averages, they'll also finish with the 13th best record in the league which places them as the sixth seed in the playoffs behind the Hawks and sets them up for a first-round matchup against the Bulls. And that's exactly how the standings look today.

So as a predictor of success, there's strong evidence suggesting that the better your point differential is per game, the better your standing. The Knicks are projecting to finish right where they are but other teams are much more affected. The Pacers get bumped out of the playoffs while home court advantage changes hands in the West. But while point differential is a good measure of regular season success, the eventual champions finished first, fourth and seventh the past three years. Perhaps this downward trend works in the Knicks' favor.