Again, please note that throughout this series we have defined "Clutch Moments" as the last five minutes of the fourth quarter or overtime, where neither team was ahead by more than five points, although obviously you could use tighter or looser requirements. The problem with going with "Super Clutch" type settings, say last two minutes, within three points, is that sample sizes become truly tiny.

Also Note:
- All stats reflect games played through March 15th of the 2004-2005 season
- Generally we require a player to have some minimum of 50 minutes of clutch playing time on the year
- You can see detailed clutch stats for every player in the league by going to the team page, clicking on the player's name and then clicking on the "Clutch Play" link (for instance, see Dwyane Wade in the clutch).

1) How the game changes in Clutch moments
One of the first things to consider is how performance changes in the last minutes of close games on a league wide basis. To that end we've listed a number of key stats below with the "All Season" and "clutch only" distinctions.

Stat Full Season Clutch  Points per 100 Possessions 105.4  110.4   Field Goal % .446  .408   Effective FG % (eFG) .481  .450   Assisted FGM 59.0%  54.1%   Blocked FGA 6.1%  6.5%   Close Shot 33.7%  30.4%   Quick Shot (within 10 secs) 38.3%  35.2%   Late Shot (21+ secs) 13.5%  16.9%   Shooting Foul/Shot 11.3%  11.9%   Free Throw % .757  .765   Offensive Rebound% 30.8%  31.9%   Defensive Rebound% 69.2%  67.8%   Turnover/Possession 15.9%  13.0%

On the surface this can seem contradictory -- FG% goes down considerably in the tight moments of a game, but Points per Possession goes up? Ah, well let me tell you about one oddity of late game antics -- the intentional foul! Not only do teams get free throws at a higher rate at the end of games, it's often also a team's best free shooter getting fouled. If we took out the intentional freebies, the points per possession would drop back down in the clutch. It is important to note the higher offensive rebound rate (desperate teams go all out to get the second chance), and the lower turnover number. The lower rate of getting close shots -- you would figure perhaps that teams would try harder to go inside at the end of games -- is no doubt due to missing out on easy transition opportunities in those "timeout after every possession" stages...

2) Player PER ratings in the Clutch
John Hollinger (writer of the excellent "Pro Basketball Forecast" annuals and now a part of ESPN Insider) designed a complicated 'overall' rating of player performance on a per minute basis that is well regarded around the NBA stats community. Indeed we have used it on 82games for the Production by Position breakouts (eg see Tim Duncan or Miami Heat pages). The idea is to take into account every stat measured in conventional box scores, so rebounds, assists, points scored, blocks, etc are all in the formula. Now there are certainly different views on some of his weightings (eg Steals are thought to be more valuable then the weight John assigns to them by several notable people) but it's a very good starting point for analysis. The league wide average is set to 15.0, and we've calculated the ratings using clutch specific data as the baselines (eg the average points per possession in the clutch instead of overall).

#	Player	Team	FGA	eFG%	FTA	iFG	Reb	Ast	T/O	Blk	PF	Pts	PER*
1	Nash	PHO	21.1	.578	17.8	21%	4.6	17.8	5.3	0.7	4.0	39.5	55.1
2	Stoudemire	PHO	15.4	.800	20.4	60%	14.2	1.2	3.1	2.5	4.3	38.9	53.7
3	Ginobili	SAS	18.6	.450	30.4	60%	8.7	5.6	5.0	1.2	8.7	41.6	49.2
4	Nowitzki	DAL	27.3	.383	24.8	18%	17.1	3.0	2.1	2.6	6.4	42.7	44.7
5	Marion	PHO	18.0	.657	8.7	37%	12.3	0.0	1.5	2.1	5.1	31.3	43.0
6	Duncan	SAS	22.2	.395	15.8	47%	19.3	3.5	1.8	1.2	5.3	29.8	40.9
7	James	CLE	25.3	.415	14.3	30%	7.2	9.1	2.9	1.9	1.9	32.9	38.9
8	Hughes	WAS	20.1	.543	19.3	21%	7.9	8.3	1.3	0.0	4.4	38.1	36.6
9	Francis	ORL	29.9	.411	22.3	36%	9.8	4.9	2.6	0.8	9.4	43.5	36.2
10	Stackhouse	DAL	18.6	.522	20.3	30%	8.1	2.4	1.6	0.0	3.2	34.8	36.1
11	Kidd	NJN	21.7	.565	11.3	15%	7.1	9.0	2.8	0.0	2.4	33.5	35.8
12	Lewis	SEA	20.0	.620	8.2	19%	10.0	1.7	1.3	1.3	4.3	30.0	35.6
13	Griffin	MIN	16.3	.722	4.5	33%	15.3	3.6	0.9	5.4	5.4	27.1	35.6
14	O'Neal	IND	30.3	.500	17.5	34%	17.5	3.5	2.9	3.5	6.4	43.1	35.1
15	Gordon	CHI	27.7	.607	14.3	21%	5.9	2.0	2.0	0.0	5.4	45.5	34.1
16	Terry	DAL	15.0	.692	9.3	30%	1.7	4.6	1.2	0.6	3.5	28.9	33.0
17	Allen	SEA	23.3	.451	12.8	25%	6.4	4.1	1.4	0.0	1.8	32.0	31.8
18	Camby	DEN	9.8	.625	6.7	37%	13.5	1.8	0.6	6.1	4.9	18.4	31.8
19	Boykins	DEN	15.5	.643	15.0	21%	3.3	5.0	2.2	1.7	7.8	32.7	31.7
20	Marshall	TOR	16.5	.681	6.0	19%	16.1	0.0	0.9	1.8	6.4	27.1	30.7
21	Daniels	SEA	14.3	.560	8.6	8%	5.1	9.7	2.3	0.0	4.0	24.6	30.5
22	Gooden	CLE	10.8	.688	6.1	62%	17.5	1.3	2.0	2.0	4.7	18.9	29.9
23	Ilgauskas	CLE	18.9	.427	12.9	43%	15.2	0.9	3.7	1.4	6.0	26.7	29.9
24	Wade	MIA	26.7	.458	20.0	30%	8.3	7.9	6.4	2.3	5.3	41.4	29.2
25	Radmanovic	SEA	16.1	.656	6.0	18%	5.0	3.0	1.5	0.5	6.0	26.6	28.7
26	Garnett	MIN	23.8	.500	9.0	29%	15.4	7.1	3.9	1.3	7.7	31.2	28.6
27	Randolph	POR	23.6	.386	18.2	38%	16.6	2.1	3.2	0.5	4.3	33.2	28.3
28	Ridnour	SEA	8.4	.318	12.9	27%	5.3	7.6	0.8	0.0	4.6	17.5	28.0
29	Marbury	NYK	28.3	.370	17.5	34%	3.1	8.0	2.1	0.0	5.2	34.9	27.2
30	Cassell	MIN	28.9	.451	9.9	17%	4.9	9.2	2.1	0.0	7.1	35.3	27.0
31	O'Neal	MIA	16.4	.581	16.8	69%	16.4	5.7	2.3	3.8	8.4	26.4	26.0
32	Lue	ATL	12.4	.647	9.4	23%	2.2	10.2	4.4	0.0	8.0	24.7	25.9
33	Jefferson	NJN	25.8	.407	15.0	30%	15.0	6.6	4.8	2.4	6.6	31.8	25.8
34	Richardson	GSW	34.7	.500	10.2	25%	13.2	3.3	3.7	0.7	4.0	41.3	25.4
35	Bryant	LAL	35.2	.371	25.3	26%	6.7	4.7	5.9	1.6	4.0	46.3	25.3
36	Billups	DET	19.7	.447	17.3	10%	7.3	4.8	2.4	0.7	5.2	33.2	25.3
37	Redd	MIL	22.5	.521	14.2	14%	7.7	3.2	3.2	0.0	3.2	36.1	25.1
38	Jamison	WAS	24.3	.493	9.8	19%	12.2	1.7	1.4	0.3	3.4	31.4	23.8
39	Wright	MEM	17.1	.478	6.0	52%	15.6	2.2	1.5	0.7	6.0	19.3	23.4
40	Williams	MEM	17.4	.667	5.8	14%	2.6	5.1	3.9	0.0	6.4	27.7	23.4
41	Arenas	WAS	23.3	.478	20.3	36%	6.4	4.4	4.1	0.0	6.8	37.5	23.2
42	Hill	ORL	21.0	.489	11.2	13%	7.9	0.5	1.9	0.9	4.2	28.0	22.6
43	Tinsley	IND	19.5	.609	7.9	40%	7.9	8.5	6.1	0.0	6.1	28.6	22.4
44	Rose	TOR	25.7	.459	13.9	22%	4.2	1.7	3.4	0.0	5.9	36.2	22.3
45	Alston	TOR	18.3	.510	8.6	28%	4.5	7.1	1.5	0.7	6.7	25.5	22.0
46	Knight	CHA	16.8	.407	9.9	25%	4.3	13.0	1.9	0.0	4.3	21.7	21.9
47	Croshere	IND	18.1	.450	12.7	23%	12.7	1.2	1.8	0.6	7.9	27.8	21.8
48	Atkins	LAL	12.8	.720	10.9	17%	5.0	4.4	1.9	0.0	4.4	25.9	21.6
49	Howard	DAL	7.6	.583	4.4	58%	8.8	1.3	1.9	2.5	3.8	13.2	21.5
50	Iverson	PHI	27.8	.370	14.9	31%	5.0	7.2	5.0	0.0	4.6	34.3	21.1

So there you have it, the top fifty PER rated players for their clutch efforts so far in 2004-05. We must emphasize that with the small amounts of playing time, these numbers would come with some pretty huge standard deviations if we calculated them, so take the hierarchy of players with a grain of salt. Additionally that "Intentional Foul" issue is a problem, boosting a guy like say Ridnour who has shot poorly and infrequently. While we could remove the freebie FT's, it seems some credit should be given to the guys who at the end of the game can be relied upon to take an inbounds pass without turning it over, and then knock down the pressure shots at the other end.

PER veterans will know it's rate to find a 25 rated player for the NBA full season, let alone a 55! Still 'Nash for MVP' types will probably run with the table above to the local media outlet. As for Kobe Bryant fans, stunned to see their man at #35, well, first off a 25.3 rating is very good, and secondly Kobe gets brought down by a few key areas -- that ugly .371 eFG (or .337 FG% if you prefer), and that high 5.9 turnovers per 48 minutes.

Ah, but we put quotes around the 'overall' measure description of the PER for a reason -- there's a lot they don't take into account, primarily defense (beyond blocks and steals), and the influence on team performance as a whole. Measuring individual defense is of course the biggest challenge currently facing NBA statisticians, and we're working on a variety of things there. The simplest to present, is the PER Counterpart numbers...

3) Player "Counterpart" Defense in the Clutch
A simple notion for tracking on some level the defensive efforts of a player is to provide the numbers for how his opposing "man" does while he is on the floor. The assumption is that a PG is guarding the opposing PG, the SG the SG, etc, which is admittedly flawed, and problematic given zone defenses, switches, difficulties in assigning a guy a position, etc and not helped any by using only the small subset of data of clutch minutes. Nevertheless, it is worthy of consideration, and here is what we find:

#	Player	Team	FGA	eFG%	FTA	iFG	Reb	Ast	T/O	PF	Pts	PER*
1	Wade	MIA	12.4	.258	5.3	33%	5.3	3.0	2.6	6.4	9.8	1.4
2	Claxton	GSW	16.6	.217	10.5	23%	7.2	3.9	2.8	9.4	13.8	2.5
3	Chandler	CHI	19.9	.282	7.7	38%	10.2	1.5	1.0	10.2	15.3	2.8
4	Hassell	MIN	18.5	.210	6.6	25%	7.1	4.2	1.2	3.0	11.9	2.8
5	Brand	LAC	14.5	.220	9.6	36%	9.0	1.7	0.9	7.3	13.6	3.5
6	Gasol	MEM	13.2	.400	8.8	26%	10.6	0.9	5.3	12.3	15.0	3.9
7	Camby	DEN	15.4	.240	5.5	48%	9.2	3.1	2.5	8.6	11.1	3.9
8	Davis	BOS	14.2	.282	4.6	20%	7.0	4.6	2.6	4.1	11.6	4.7
9	Jeffries	WAS	17.0	.350	8.5	5%	5.1	2.5	0.8	7.6	17.0	4.8
10	Deng	CHI	12.7	.395	4.0	5%	6.7	2.7	2.7	8.0	12.7	4.9
11	Jackson	HOU	9.3	.167	6.2	25%	8.5	2.3	0.8	3.1	8.5	5.1
12	Anthony	DEN	15.0	.212	4.0	34%	8.6	4.6	2.3	4.0	9.8	5.1
13	Maggette	LAC	13.1	.341	8.0	19%	6.1	3.8	3.2	7.0	15.6	5.4
14	Iverson	PHI	14.9	.282	8.0	10%	5.3	5.3	6.1	7.2	16.0	5.6
15	Van Horn	MIL	17.5	.375	2.6	35%	7.9	0.0	0.9	4.4	14.9	6.6
16	Carter	NJN	15.3	.310	8.7	33%	8.7	1.5	1.5	7.3	15.3	6.7
17	Wesley	NOH	15.9	.333	6.0	12%	5.3	6.0	2.7	4.0	13.9	6.7
18	Hinrich	CHI	14.2	.284	6.2	16%	7.3	6.9	1.9	7.7	11.9	6.9
19	D.Jones	MIA	16.2	.380	3.6	24%	4.2	4.2	2.9	5.2	15.6	6.9
20	Miller	IND	12.5	.385	5.3	15%	5.8	1.9	3.8	4.8	13.5	7.1
21	Turkoglu	ORL	8.0	.423	3.1	23%	6.8	1.2	0.6	6.8	9.2	7.4
22	Tinsley	IND	12.2	.400	8.5	5%	1.8	6.1	4.3	5.5	14.6	7.9
23	Payton	BOS	14.6	.462	4.8	17%	5.6	3.9	3.6	6.4	17.4	8.1
24	Peterson	TOR	20.9	.281	7.8	15%	9.8	2.0	1.3	3.9	17.6	8.2
25	Prince	DET	13.3	.378	5.5	19%	7.1	1.9	1.0	3.2	13.6	8.2
26	Mihm	LAL	12.4	.200	14.8	53%	19.0	1.6	3.3	9.9	15.7	8.2
27	Smith	NOH	12.4	.469	10.9	31%	7.0	3.1	3.9	3.1	16.3	8.3
28	Hudson	MIN	10.2	.250	3.4	25%	9.3	12.7	3.4	5.9	5.9	8.5
29	Stojakovic	SAC	14.9	.436	6.5	25%	6.9	2.7	2.3	5.0	17.9	8.7
30	Fisher	GSW	16.9	.347	10.8	30%	8.0	5.2	2.8	5.6	17.8	9.0
31	Bryant	LAL	18.2	.402	4.0	32%	7.1	5.5	2.0	9.5	16.6	9.1
32	Stackhouse	DAL	16.2	.325	4.9	15%	6.5	4.1	1.6	6.5	13.8	9.2
33	Richardson	GSW	13.9	.408	4.8	31%	6.2	2.2	0.7	4.8	15.3	9.7
34	O'Neal	MIA	12.6	.242	13.4	36%	13.8	0.8	1.5	9.9	17.2	9.7
35	Boozer	UTA	15.7	.387	5.6	45%	10.7	2.0	2.0	7.6	15.7	9.8
36	Billups	DET	16.6	.344	5.9	20%	5.5	8.0	1.7	7.6	15.9	9.9
37	Foyle	GSW	10.2	.273	16.7	54%	12.0	0.9	0.9	8.3	16.7	9.9
38	Allen	SEA	12.3	.407	1.8	37%	7.8	2.3	1.8	3.7	11.4	10.0
39	Gordon	CHI	21.3	.326	9.9	25%	7.4	4.5	2.5	5.5	21.3	10.1
40	Simmons	LAC	10.3	.564	5.5	20%	4.5	3.4	1.8	4.5	16.6	10.1
41	B.Wallace	DET	16.4	.409	6.0	40%	6.7	1.1	1.5	4.5	17.9	10.3
42	McGrady	HOU	10.5	.422	6.5	31%	7.8	3.3	2.3	6.5	14.0	10.5
43	Blount	BOS	14.5	.353	11.1	44%	11.1	2.1	0.9	4.3	17.4	10.6
44	Williamson	PHI	15.5	.417	13.5	25%	16.1	1.3	3.2	9.7	20.0	10.8
45	Redd	MIL	18.4	.482	4.2	21%	7.1	2.6	1.6	7.4	20.3	11.3
46	Jamison	WAS	12.2	.431	6.8	47%	11.8	3.0	2.0	8.1	15.2	11.3
47	Ginobili	SAS	11.8	.342	9.3	57%	7.5	1.9	3.1	8.7	16.2	11.3
48	Sura	HOU	21.2	.364	7.1	27%	5.8	7.7	1.3	4.5	19.9	11.4
49	Odom	LAL	17.3	.407	7.7	29%	10.3	1.9	1.3	7.4	19.2	11.5
50	Knight	CHA	14.9	.354	1.9	12%	8.1	9.9	3.1	5.0	12.4	11.6

You get some of the players with "defensive" reputations showing up well -- Chandler, Hassell, Camby types -- along with a number of superb (and superbly active) offensive players...Wade, Maggette, Iverson, etc. One theory would be that the go-to guys on offense leave their man focused on defending them and not doing much at the other end (or rebounding say) that their counterparts consequently post middling to no stats of note themselves. This however could be a very important consideration -- a great offensive player can be a positive on defense as well, since the poor sap having to guard him needs to catch his breath when the team with the ball changes. Again, with the league average set to 15, a lot of players posting better than average counterpart stats aren't cracking the top fifty.

Okay, now it's time to take a stab at putting it all together

4) A semblance of an overall Player Clutch Rating
So what should go into the "ultimate" player rating system? Certainly we want to include a player's easily trackable elements, like shooting, rebounding, and the like. Clearly we would love to have better measures for what the player gives up in those same stats on the other end. Finally, we would want to include something that gets at how well the team plays with the player on the court -- after all, there are lots of ways to contribute in basketball that go unmeasured currently, and whether your team wins or loses is supposed to be the true goal, not whether you are a highlight film regular.

We've been working on a substantial project for some time to do just this, bring these different pieces together, AND incorporate the extra statistical information we have here at 82games (eg did a passer contribute more on an assisted dunk than on an assisted 2-point jump shot? or, did the free throws come via a drawn shooting foul, or a backcourt intentional foul). This Uber-Rating is however going to have to wait until the regular season concludes, and we wrap up a few more reseach projects relating to this.

In the meantime we'll go with something quick and easy -- PER minus Counterpart PER (also referred to as "PER Difference") adjusted by 1/5 of the player's clutch plus/minus per 48 minutes. We have published ratings that use PER Diff + ON/OFF, but as mentioned elsewhere, the on/off splits are useless for clutch breakdowns since some players are never off the floor in those moments. Adding in a full +/- number is too strong, and too dependent on the teammates, but surely putting a player on the hook for 1/5 of the team performance doesn't seem too much, indeed, it's in most cases a minor adjustment to the PER Diff component.

# Player Team PER dPER Diff +/- Rating 1  Ginobili SAS 49.2  11.3  37.9  +9  39.0  2  Stoudemire PHO 53.7  19.5  34.2  +16  36.2  3  Nash PHO 55.1  23.8  31.3  +31  35.4  4  Wade MIA 29.2  1.4  27.8  +79  33.7  5  Nowitzki DAL 44.7  16.0  28.7  +43  32.3  6  Camby DEN 31.8  3.9  27.9  +22  30.6  7  Griffin MIN 35.6  15.8  19.8  +58  30.2  8  James CLE 38.9  12.4  26.5  +13  27.8  9  Stackhouse DAL 36.1  9.2  26.8  +5  27.6  10  Allen SEA 31.8  10.0  21.7  +58  27.0  11  Gordon CHI 34.1  10.1  24.0  +26  26.5  12  Terry DAL 33.0  12.8  20.2  +52  26.2  13  Hughes WAS 36.6  14.6  22.0  +42  25.7  14  Lewis SEA 35.6  15.4  20.2  +59  25.4  15  O'Neal IND 35.1  14.6  20.5  +18  22.6  16  O'Neal MIA 26.0  9.7  16.3  +67  21.4  17  Daniels SEA 30.5  16.6  13.9  +65  21.4  18  Kidd NJN 35.8  18.9  17.0  +38  20.5  19  Tinsley IND 22.4  7.9  14.5  +44  19.9  20  Ridnour SEA 28.0  12.2  15.9  +22  19.2  21  Bryant LAL 25.3  9.1  16.3  +34  19.0  22  Garnett MIN 28.6  13.6  15.0  +55  18.6  23  Gooden CLE 29.9  13.2  16.7  +12  18.3  24  Richardson GSW 25.4  9.7  15.7  +31  18.0  25  Cassell MIN 27.0  15.9  11.2  +42  17.1  26  Billups DET 25.3  9.9  15.4  +23  17.0  27  D.Jones MIA 18.4  6.9  11.5  +82  16.8  28  Boykins DEN 31.7  18.2  13.5  +28  16.6  29  Francis ORL 36.2  19.6  16.5  -10  15.8  30  Iverson PHI 21.1  5.6  15.5  +2  15.7  31  Marion PHO 43.0  28.9  14.1  +15  15.7  32  Jamison WAS 23.8  11.3  12.4  +45  15.5  33  Marshall TOR 30.7  16.6  14.1  +6  14.7  34  Hudson MIN 19.9  8.5  11.4  +17  14.3  35  Carter NJN 13.9  6.7  7.3  +43  13.5  36  Randolph POR 28.3  13.1  15.2  -16  13.5  37  Smith NOH 19.3  8.3  11.1  +15  13.4  38  Redd MIL 25.1  11.3  13.8  -12  13.0  39  Jefferson NJN 25.8  13.2  12.6  +3  13.0  40  Hassell MIN 15.9  2.8  13.1  -2  12.8  41  Davis BOS 17.0  4.7  12.3  +4  12.5  42  Chandler CHI 13.1  2.8  10.3  +17  12.0  43  Ilgauskas CLE 29.9  19.2  10.7  +13  11.9  44  Butler LAL 20.8  11.9  8.9  +27  11.5  45  Claxton GSW 12.5  2.5  9.9  +11  11.2  46  Mihm LAL 15.1  8.2  6.9  +26  11.1  47  Van Horn MIL 16.5  6.6  9.9  +4  10.6  48  Radmanovic SEA 28.7  25.4  3.4  +69  10.3  49  Anthony DEN 13.5  5.1  8.4  +16  10.2  50  Turkoglu ORL 16.2  7.4  8.8  +11  10.1

All right, a lot of people would balk at the notion of Manu Ginobili as the NBA's most clutch player. The facts though are he is very productive on offense (#6 in the league in points per minute during clutch stages), draws an incredible number of shooting fouls (#8 and the top perimeter player), knocks down the free throws (81%), creates his own shot (only 8% of his clutch FGM are assisted), holds his counterpart to sub-par performance (.342 eFG and only 1.9 Assists to 3.1 Turnovers), and while his +9 plus/minus doesn't seem like much, the Spurs have actually been -21 when he's off the court during crunch time, good for the #12 on/off or Clutch Roland Rating in the league. Did we mention that 60% of his crunch time shots are taken inside? Manu takes it to the hole! This will not sway Kobe Bryant fans any, who are still reveling in his incredible recent performance in the 4th quarter against Charlotte. That however is the problem in a nutshell -- people remember the game where he's brilliant, and forget the games (yes, plural) where he falls short. Let's state it one more time, Kobe is a good clutch player, but not currently the best. For further evidence that Kobe is not infallible, you might want to check out his 03-04 playoff clutch efforts, when he had a Field Goal Percentage of, ahem, 28%... No, the better arguments against Manu come from the likes of Steve Nash, Dwyane Wade, and Amare Stoudemire, all of whom have some fantastic numbers to show off in the "money minutes" this season. Truthfully though, any conclusions we can draw from these numbers are limited by the reality of so few minutes of clutch time performance for players to assess. The solution though is simple -- the next time we talk about Clutch stats we will use a Three-Year survey of data at which point we should have on hand something substantial. For number fiends, or those whose favorite player didn't make the top fifty at left, we've published Team-by-Team lists of Player Clutch Performance Ratings.