And the debate continues.
First, a quick primer for the unfamiliar. Synergy is a video logging and tagging service that indexes and records all NBA plays. It spits out a stat called PPP (points per play), a valuation of how successful (or not) a specific player was in a specific situation (offensively and defensively). Like any metric, there is noise and can't be taken as an all-encompassing truth, but the specificity of the data provides a unique perspective, and is the closest thing we have to evaluating a player's defensive ability.
One more thing: the "play" in PPP is a possession that ends in a FGA, TO, or FTA (excluding technical FTs). It doesn't account for assists (within the specific situation), either made or allowed, so it's purely a measure of a player's ability to score and prevent his opponent from scoring (some of the assist data can be teased out from watching the likely recipient's data, though. For example, watching Turiaf's finishes off the pick and roll will give you a good idea of Curry's passing effectiveness from that situation. Note: this is entirely different from the raw data that Synergy provides, which has all of Curry's assists for one's viewing pleasure, unsorted. The nature of the tagged data, organized by the shot taker/turnover maker prevents organizing assist data thusly).
Also (I lied about one more thing), the defensive tagging is based on the primary defender at the beginning of the situation. So if Steph Curry is guarding Chris Duhon, but switches onto David Lee after a screen and Lee posts him up and attempts a shot, that play gets logged as post-up defense for Curry.
Without further ado: Numbers!
Stephen Curry - Offensive PPP
| Play | %Time |
Number |
PPP |
Rank |
FGM | FGA |
FG% |
3FGM | 3FGA |
3FG% | %SF | And1 |
%TO | %Score |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Overall | 100% | 1444 | 0.95 | 161 | 521 | 1122 | 46.4 | 164 | 374 | 43.9 | 3.7 | 16 | 16.4 | 41.8 |
| Isolation | 13.3% | 192 | 0.87 | 95 | 63 | 153 | 41.2 | 12 | 28 | 42.9 | 4.2 | 6 | 13.5 | 39.6 |
| P&R BH |
26.9% | 389 | 0.81 | 84 | 131 | 297 | 44.1 | 20 | 50 | 40.0 |
3.3 | 3 | 18.8 | 38.6 |
| Spot-Up | 16.9% | 244 | 1.39 | 2 | 118 | 226 |
52.2 | 82 | 162 | 50.6 | 3.7 | 1 | 3.3 | 52.5 |
| Off Screen |
3.3% | 47 | 0.91 | 58 | 17 | 42 | 40.5 | 9 | 18 | 50.0 | 0.0 | 0 | 10.6 | 36.2 |
| Hand Off |
5.5% | 79 | 0.95 | 30 | 31 | 73 | 42.5 | 8 | 22 | 36.4 | 2.5 | 0 | 3.8 | 43.0 |
| Cut | 2.8% | 40 |
1.38 | 46 | 25 | 36 | 69.4 | 0 | 0 |
0.0 | 5.0 | 2 | 5.0 | 67.5 |
| Transition |
22% | 318 | 0.94 | 261 | 116 | 248 | 46.8 | 29 | 79 | 36.7 | 5 | 4 | 16.0 | 42.5 |
Monta Ellis - Offensive PPP
| Play | %Time |
Number |
PPP |
Rank |
FGM | FGA |
FG% |
3FGM | 3FGA |
3FG% | %SF | And1 |
%TO | %Score |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Overall | 100% | 1798 | 0.89 | 277 | 626 | 1402 | 44.7 | 76 | 223 | 34.1 | 7.7 | 36 | 12.8 | 43.3 |
| Isolation | 27.9% | 502 | 0.8 | 153 | 149 | 386 | 38.6 | 11 | 39 | 28.2 | 9.2 | 10 | 12.2 | 39.6 |
| P&R BH |
26.1% | 469 | 0.82 | 81 | 154 | 365 | 42.2 | 15 | 49 | 30.6 |
7.2 | 11 | 14.1 | 56.3 |
| Post-Up | 3.9% | 71 | 1.08 | 14 | 31 | 58 |
53.4 | 0 | 0 | 0 | 9.9 | 0 | 5.6 | 52.5 |
| Spot-Up |
7.6% | 136 | 1.06 | 83 | 53 | 121 | 43.8 | 28 | 65 | 43.1 | 0.0 | 4 | 8.1 | 41.9 |
| Off Screen |
2.4% | 44 | 0.98 | 30 | 18 | 40 | 45.0 | 4 | 8 | 50.0 | 2.2 | 0 | 4.5 | 45.5 |
| Hand Off |
6.0% | 108 |
0.88 | 45 | 40 | 90 | 44.4 | 3 | 11 |
27.3 | 4.6 | 4 | 11.1 | 41.7 |
| Cut |
4.1% | 73 | 1.00 | 216 | 32 | 61 | 52.5 | 0 | 0 | 0.0 | 6.8 | 2 | 8.2 | 50.7 |
| Transition |
15.9% | 285 | 1.19 | 120 |
130 | 225 | 57.8 | 14 | 34 | 41.2 | 11.6 | 5 | 8.8 | 57.5 |
Analysis: Words in the middle of numbers! So this is a lot of data to digest, obviously, but there are some interesting things to pull out. One, Stephen Curry was the 2nd best Spot-Up shooter in the NBA last season (if you look at the notes, where I expressly state "don't care about NBA rank," you'll see why this is all kinds of impressive. It's really hard, in this system, to get a single digit ranking like that as a full-time player). I knew he was a great shooter, but I had no idea he was that good. Shooting over 50% from three as a rookie in that situation is mind-bottling. Unfortunately, he wasn't as impressive as a scorer in the main NBA PG roles (P&R and Isolation), but that's to be expected somewhat as a rookie. He wasn't awful (actually similar to Rondo's numbers), but as offensively gifted as he is, those numbers will need to increase as he goes along (and I'd expect they would). What's encouraging about his profile (Spot-Up, Cut and Off-Screen numbers mostly) is that he doesn't need to dominate the ball to score efficiently, a good sign that he can co-exist with another ball handler like Monta. He does struggle in transition and in getting to the line, so those will also be things to work on.
Monta, on the other hand...yeah. The good news is GSW found their power forward! Just post him up and let him work, apparently. Looking at his profile it's easy to see why he had success as a second/third/fourth option. His speed working off screens and on hand offs is hard to control, as is his quickness in the open floor. His isolation game is abysmal, though, and is the reason his efficiency isn't anywhere near an acceptable level. As currently constructed, his offense just doesn't work. Something has to change.
Stephen Curry - Defensive PPP
| Play | %Time |
Number |
PPP |
Rank |
FGM | FGA |
FG% |
3FGM | 3FGA |
3FG% | %SF | And1 |
%TO | %Score |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Overall | 100% | 1181 | 0.89 | 170 |
388 |
949 |
40.9 | 113 |
302 |
37.4 | 6.4 | 21 |
11.7 | 40.3 |
| Isolation | 16.9% | 199 |
0.69 | 31 |
47 |
150 |
31.3 |
6 |
23 |
26.1 | 8.0 | 3 | 13.6 | 34.2 |
| P&R BH |
25.7% | 303 |
0.77 | 66 |
97 |
227 |
42.7 |
12 |
33 |
36.4 |
4.0 | 4 | 19.5 |
37.0 |
| Post-Up | 7.5% | 89 |
0.72 | 33 |
18 |
59 |
30.5 |
0 | 0 | 0.0 | 18.0 | 2 | 13.5 | 39.3 |
| Spot-Up |
21.3% | 251 |
1.05 | 269 |
91 |
227 |
40.1 |
62 |
157 | 39.5 |
3.6 | 4 | 4.8 | 40.2 |
| Off Screen |
4.4% | 52 |
0.90 | 100 |
18 |
47 |
38.3 |
5 |
16 | 31.3 |
7.7 | 0 | 1.9 | 42.3 |
| Hand Off | 3.1% | 37 |
1.08 | 85 |
14 |
26 |
53.8 |
5 |
7 |
71.4 | 10.8 | 2 |
18.9 | 48.6 |
Monta Ellis - Defensive PPP
| Play | %Time |
Number |
PPP |
Rank |
FGM | FGA |
FG% |
3FGM | 3FGA |
3FG% | %SF | And1 |
%TO | %Score |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Overall | 100% | 1120 | 0.98 | 367 | 403 |
876 | 46.0 | 99 |
253 | 39.1 | 6.7 | 23 |
12.3 | 44.9 |
| Isolation | 17.9% | 201 | 0.97 | 273 | 65 | 144 | 45.1 | 10 | 31 | 32.3 | 10.4 | 5 | 14.9 | 45.3 |
| P&R BH |
20.4% | 229 | 0.84 | 128 | 74 | 155 | 46.5 | 6 | 13 | 46.2 |
5.2 | 5 | 20.5 |
41.0 |
| Post-Up | 10.4% | 116 | 0.95 | 215 | 41 | 83 |
49.4 | 0 | 0 | 0.0 | 9.5 | 3 | 16.4 | 47.4 |
| Spot-Up |
21.2% | 237 | 1.05 | 269 | 85 | 211 | 40.3 | 54 | 126 | 39.7 | 4.6 | 3 | 5.5 | 41.4 |
| Off Screen |
8.8% | 98 | 0.94 | 118 | 37 | 85 | 43.5 | 3 | 13 | 23.1 | 4.1 | 1 | 6.1 | 44.9 |
| Hand Off | 3.8% | 42 |
1.43 | 102 | 21 | 30 | 70.0 | 6 | 8 |
75.0 | 11.9 | 1 |
14.3 | 66.7 |
Analysis: Quite promising numbers for the Golden Babyfaced Assassin Child. Intuitively, I felt like he had good defensive fundamentals, and understood how to stay in front of his man well enough, and Synergy backs that up somewhat. Being strong against the main attack modes he'll see his entire career as a rookie is an encouraging sign. Bigger guards will always try to post him up, but aside from the high shooting foul percentage, he's able to take care of himself, and can defend the Isolation/P&R better than a lot of veteran PGs already. Amazing. Big time struggles in the rotation department, and that shows up in Spot-Up defense. Some of that will be team defense, but a lot of that is Curry needing to learn NBA rotations. Rookies are expected to struggle in that area, and Curry is a rookie. Even if he stealth kills babies with his golden face.
Again, Monta comes out not looking so good. He's below average to downright terrible in every single category. Even excusing the Spot-Up numbers due to team defense, there are too many deficiencies to overlook. Combined with his average-at-best offense and Synergy paints a pretty damning picture of Monta. So, he's on a moped, obviously.
Notes:
-%SF means percentage of shooting fouls for that given scenario.
-I omitted exceptionally small percentages (under 2%), so you mental math heroes wondering why things aren't adding up to 100%, that's your reason.
-Take the NBA ranks with grains of salt. A player only needs >2% of their possessions in a given situation to qualify and certain stats are dominated by bigs (transition for example) and/or specialists (spot-up shooting for example). They're mostly for fun, but do carry some information (if only rudimentary). Higher still is better, and really low is still really bad.
-Some ranges for Overall Offensive PPP (based on my observations so far, spent a few days with the data, by no means definite, invert for defensive ranges): <.80 is bad; .81-.89 is decent; .90-.95 is good; .96-.99 is great; 1.00+ is elite. This varies within the data somewhat. For example, average Transition and Cut PPPs are going to be much higher than the other categories, and Spot-Up is higher on average than every non-Transition/Cut situation. P&R Ball Handler and Isolation seem to be a bit lower on average, so adjust expectations accordingly. This is something the NBA Ranks will reflect well.
