Monday, December 23, 2013

Why Chris Paul may not want to Play with Melo

It seems there are two conflicting narratives of Carmelo Anthony. One says that he is just a star on the stat line that can't help an organization win. The other contends he simply needs an elite PG--a trade to the Clippers, which would unite him with Chris Paul is often rumored--to complement his offensive talents and consistent winning will follow.

The chart below questions the logic of the 'Melo needing an elite PG' narrative. I pulled data from NBA Stats for Carmelo through last night's games on the % of field goals (FG) that he makes that are assisted and the % that he makes that are unassisted. The chart shows that starting in the 2008-09 season something fundamentally changed in Carmelo's game--he went from being an above 50% shooter on assisted FGs to below 50% and the opposite trend occurred on unassisted FGs.

In what has become especially stark over the last three years, Melo is becoming a much less efficient shooter unless he is in an unassisted scenario. In other words, these numbers suggest that Carmelo needs the ball in his hand and only his hand to be an effective scorer. And, adding a high assist, distributor, like Chris Paul, may not actually benefit his game as his FG% off of assists continues to decline.


Sunday, December 8, 2013

Team FG Attempts and Makes

I collected team data from NBA Stats (as of December 7th's games) on field goal (FG) attempts and makes in five foot increments from the basket out to 29 feet (the NBA three point line is 23 feet 9 inches so outside of end of half, regulation and OT shots, there are likely few, if any, shots beyond 29 feet). The first chart shows the percent of FG attempts by five foot increment.


The second chart shows the percent of FG makes by five foot increment. There isn't a clear relationship between winning and percent of FGs attempted or made by distance from the basket as there are poor teams that attempt a high (Knicks) and low (Jazz) percentage of longer shots and there are better teams that shoot a high (Rockets) and low (Trailblazers) percentage of closer shots. Much of the relationship is likely explained by team personal (i.e., the Rockets shoot a higher percentage of FGs from within 5 feet because they have Dwight Howard).

Wednesday, November 27, 2013

More Nerding Out to Advanced NBA Team Stats

Both charts below show team winning % as of last night and all categories are per game (data pulled from stats.nba.com.) In both cases bigger circles indicate a greater winning percentage this season. The top chart compares turnovers per 100 possessions with assists per 100 possessions. We would expect all of the larger circles to be at the bottom right of the chart indicating that winning teams have a high average of assists per 100 possessions and a lower average of turnovers per 100 possessions. Heat, Clippers and Spurs are in that general direction and the Jazz have a high turnover ratio and low assist ratio and, thus, are in the top left, which all make sense. Interestingly, lower-tier teams like the Knicks and Jazz have an about average assist ratio and very low turnover ratio. Therefore, limiting turnovers doesn't seem to always correlate with a high winning percentage.


The second chart compares the percentage of possible offensive rebounds that a team gets with the team's possessions per 48 minutes of play. It seems reasonable that bigger circles would be toward the top right of the chart as a higher offensive rebounding percentage and more possessions per 48 minutes would correlate with winning. There doesn't seem to be any clear relationship. The anomalous Spurs, Heat and Hawks all have especially low offensive rebounding percentage but are also above average teams, which counters the idea that better offensive rebounding teams win more games. Perhaps this reflects that better teams (in terms of winning percentage) do not crash the offensive boards to ensure that enough players are back on defense to prevent fast breaks and easy baskets.


Tuesday, November 26, 2013

Anthony Davis, Andre Drummond then the rest

Comparing players within draft classes has obvious merits (i.e., determining rookie of the year, the overall strength of a class, etc.) but with players entering the draft after various years of college experience and many others reclassifying in middle and high school to gain an extra year of eligibility in AAU, actual age sometimes gets overlooked. (For example, even in his third season Bismack Biyombo is almost a full year younger than rookie Michael Carter-Williams. In this instance it might not be fair to compare a rookie to a third year veteran and expect the former to be performing at a superior level.)

I collected advanced data from basketball-reference.com as of last night's games for all players that are 22 years or younger that have played at least 50 minutes this season. It resulted in a list of 56 players. I ranked each player based on evenly weighting their PER, effective FG% and win shares per 48 minutes (albeit, an arbitrary selection of advanced stats, but all three are recognized measures of a well-rounded player). Anthony Davis is by far the bets younger player this season. Surprisingly, Andre Drummond, Terrence Jones and Jared Sullinger are at or above the likes of Kawhi Leonard and Kyrie Irving.

Player Age MP PER eFG% WS/48 Wght
Anthony Davis 20 458 28.6 0.49 0.26 9.78
Andre Drummond 20 473 21.5 0.66 0.18 7.45
Terrence Jones 22 267 19.5 0.58 0.18 6.75
Michael Carter-Williams 22 398 19.5 0.45 0.10 6.69
Kawhi Leonard 22 380 19.1 0.55 0.21 6.62
Jared Sullinger 21 333 18.9 0.50 0.13 6.51
Kyrie Irving 21 483 19 0.46 0.07 6.51
Derrick Favors 22 529 16.5 0.48 0.08 5.68
Perry Jones 22 104 14.8 0.67 0.14 5.20
Jeremy Lamb 21 233 14.7 0.51 0.13 5.11
Enes Kanter 21 478 14.3 0.50 0.05 4.95
Steven Adams 20 226 13.7 0.46 0.15 4.77
Bradley Beal 20 523 13.7 0.48 0.05 4.74
Tristan Thompson 22 469 13.7 0.43 0.07 4.73
Terrence Ross 22 246 13.3 0.56 0.13 4.66
Khris Middleton 22 317 13.2 0.53 0.07 4.60
Jonas Valanciunas 21 359 13.2 0.47 0.10 4.59
Harrison Barnes 21 299 13.1 0.55 0.11 4.59
Tony Wroten 20 359 13.1 0.44 0.01 4.52
Tim Hardaway 21 184 12.6 0.49 0.06 4.38
Austin Rivers 21 97 12.6 0.46 0.07 4.38
Cory Joseph 22 55 12.6 0.41 0.11 4.37
Michael Kidd-Gilchrist 20 404 12.5 0.51 0.08 4.36
Nate Wolters 22 351 12.5 0.42 0.04 4.32
Dion Waiters 22 332 12.5 0.44 (0.01) 4.31
Thomas Robinson 22 165 11.3 0.45 0.04 3.93
Hollis Thompson 22 181 11 0.51 0.06 3.86
Bismack Biyombo 21 297 10.8 0.65 0.12 3.86
 
The bottom 28 consists largely of rookies a few previous lottery picks that have been especially underwhelming, like Derrick Williams and Brandon Knight.

Player Age MP PER eFG% WS/48 Wght
Giannis Antetokounmpo 19 126 11 0.51 0.03 3.85
Victor Oladipo 21 357 11 0.43 (0.04) 3.80
Ben McLemore 20 284 10.7 0.47 0.03 3.73
Evan Fournier 21 181 10.5 0.49 0.06 3.68
Maurice Harkless 20 367 10.5 0.48 0.04 3.67
Alec Burks 22 418 10.3 0.39 (0.02) 3.56
Kentavious Caldwell-Pope 20 267 10.1 0.40 0.05 3.52
Xavier Henry 22 281 9.9 0.46 0.01 3.45
Rudy Gobert 21 154 9.9 0.46 (0.01) 3.45
Shane Larkin 21 59 9.4 0.43 (0.01) 3.27
Kelly Olynyk 22 317 9.4 0.41 0.01 3.27
Phil Pressey 22 169 9 0.31 0.00 3.10
Derrick Williams 22 162 8 0.37 0.03 2.80
Cody Zeller 21 275 7.7 0.34 (0.00) 2.68
Archie Goodwin 19 169 7.6 0.40 (0.05) 2.65
Sergey Karasev 20 69 7.4 0.41 0.01 2.61
Trey Burke 21 86 7.5 0.37 (0.10) 2.59
Tony Snell 22 68 7.1 0.41 0.07 2.53
Brandon Davies 22 133 5.6 0.33 (0.01) 1.97
Ian Clark 22 70 4.4 0.40 (0.08) 1.58
John Jenkins 22 67 4.3 0.41 (0.04) 1.56
Reggie Bullock 22 93 4.3 0.39 (0.04) 1.55
Solomon Hill 22 101 4.3 0.30 0.03 1.54
Dennis Schroeder 20 121 2.2 0.35 (0.13) 0.80
Anthony Bennett 20 122 2.3 0.24 (0.14) 0.80
Brandon Knight 22 87 0.5 0.27 (0.22) 0.18
Nemanja Nedovic 22 60 -1.9 0.27 (0.14) (0.59)
Marquis Teague 20 69 -7 0.19 (0.24) (2.35)