looking at the play-by-plays for all the games this season:

1. i only noted fga's and fgm's in the final 3 minutes of regulation (and all of the OT's)
2. i did not look at assists or free throw attempts.

here are the #'s:

entire season: 354-886 = 0.399 fg%
first 45 minutes: 318-795 = 0.400 fg%
final 3 minutes: 36-91 = 0.395 fg%
final 3 minutes (games decided by 5 points or less): 18-49 = 0.367 fg%

(surprisingly this took only 45 mins, and yes, i am a loser)

[Edited by - djsunyc on 03-02-2007 5:26 PM]

Posted by djsunyc:

looking at the play-by-plays for all the games this season:

1. i only noted fga's and fgm's in the final 3 minutes of regulation (and all of the OT's)
2. i did not look at assists or free throw attempts.

here are the #'s:

entire season: 354-886 = 0.399 fg%
first 45 minutes: 318-795 = 0.400 fg%
final 3 minutes: 36-91 = 0.395 fg%
final 3 minutes (games decided by 5 points or less): 18-49 = 0.367 fg%(surprisingly this took only 45 mins, and yes, i am a loser)

[Edited by - djsunyc on 03-02-2007 5:26 PM]

It's amazing this guy is consider J-Clutch

It's hard to decide what these numbers are telling us without some context. I imagine just about every player's FG% drops from their overall FG% in the closing minutes of tightly contested games. Depending on how pronounced this effect is league-wide, particularly for players whose role it is to create shots in these situations, it might actually reflect favorably on Crawford that his accuracy drops "only" by about 3%.

Plus I heard the stat on ESPN Radio (max Kellerman show) that in the final 3 minutes of a game JC shoots 64.5%...so I don't know why you are trying to prove what I heard wrong...AND YES, u r a loser...just take it for what its worth.

Another thing to consider is sample size. If you do the stats, you find that difference between the %s across the two samples (whole season FG% vs. FG% in final minutes of close games) is not close to being statistically significant. What this means is that even if Crawford's actual shot-making ability did not drop off in the final minutes of close games, it would not be uncommon to see the small 3% dip occur just due to random chance.

In order to show that such a small difference in shooting %s was actually statistically significant (i.e. that the difference could not readily be attributed to chance), we'd need much larger sample sizes. To give an idea of how much larger, suppose some player shoots .399 for the season and .367 in the closing minutes of tight games. Now suppose that this player has shot 10,000 FGAs over the whole season (if anyone could do that it would be Jamal). In order for the 3% difference between his season FG% and clutch FG% to be statistically significant, he would need to have shot about 950 crunch time FGAs. For Jamal this season, our numbers are 886 (vs 10,000) and 49 (vs 950).

So, yeah. You can't really take too much home from these numbers.

Posted by tomverve:

Another thing to consider is sample size. If you do the stats, you find that difference between the %s across the two samples (whole season FG% vs. FG% in final minutes of close games) is not close to being statistically significant. What this means is that even if Crawford's actual shot-making ability did not drop off in the final minutes of close games, it would not be uncommon to see the small 3% dip occur just due to random chance.

In order to show that such a small difference in shooting %s was actually statistically significant (i.e. that the difference could not readily be attributed to chance), we'd need much larger sample sizes. To give an idea of how much larger, suppose some player shoots .399 for the season and .367 in the closing minutes of tight games. Now suppose that this player has shot 10,000 FGAs over the whole season (if anyone could do that it would be Jamal). In order for the 3% difference between his season FG% and clutch FG% to be statistically significant, he would need to have shot about 950 crunch time FGAs. For Jamal this season, our numbers are 886 (vs 10,000) and 49 (vs 950).

So, yeah. You can't really take too much home from these numbers.

tom, I completely agree that these numbers by themselves don't tell the entire story. but what I am also in the process of doing is compiling the numbers for kobe and melo - two other clutch players.

what I presented here are just raw numbers after hearing he was a 60+ % shooter at the end of games.