Ranking Shooting Guards by Points Gained
Ranking Shooting Guards by Points Gained
Okay context - I have for a while now wanted to try to create a single metric that captures a player’s impact across all aspects of the game. In the past, I’ve used points per possession to judge a player's offensive effectiveness, and then kind of squinted at the defensive and rebounding grades and stats to try to make the rest out.
Well, it recently occurred to me that you could convert a player’s defensive and rebounding output to possessions gained, and then from possessions gained to an estimation of points gained.
Offensive points gained are pretty easy to calculate - working from the assumption that passing doesn’t matter at all, it should be Possessions Used * (PPP - a baseline PPP). To capture defensive points gained, I’ve followed the same formula to a degree, capturing the rebounds, steals, and blocks above some average baseline. From there, I make an assumption about what percentage of each of the above are truly possessions won (always the case for a steal, and seemed like 0.6 was the right coefficient for blocks and rebounds). Initially I did the above based on the per-game stats logged for each player, but realized that dramatically overstates the impact of starters and higher-minute types, so I then converted all of the above to per-minute played metrics, and then output both a per-minute and a per-game Points Gained score.
The other somewhat special sauce is calculating what the baselines should be. For example, at first I started by ascribing 1 point per possession for each calculated defensive possession gained. I found eventually that this was a little high, and consequently overstated defensive impact. The median team-level PPP obviously fluctuates a bit over time, but this year at present is 0.927.
In terms of the player-level baselines, I wanted to solve for the different opportunities afforded to different positions to record points / rebounds / blocks, so I’ve broken the pool of players down by the positions they actually play. This was annoying, and I wouldn’t recommend doing it by hand. From there, the other big stylistic choice was whether to use as a baseline the average across the data set, or instead, try to determine a replacement value and use that. To keep things easy here, I chose the former.
Some obvious things like positional defense are not considered in these metrics. I thought about trying to build something on top of them that would layer in a defensive rating, but that ultimately seemed messier than just ignoring non-stock and rebound defense. Also all of the choices I had to make above definitely have a dramatic impact on the final product - I tried to go with what made the most sense, but who knows what percentages of blocks are ultimately recovered by the team, ya know? When Tim Hardaway Jr gets a steal, is he really passing, or am I getting a THJ shot?
I liked listening to the PG rankings the Over/Under Pod fellas made up, which was part of the inspiration to publish this. I might get to PGs by the end of the year, but wanted to let their list breathe a bit, so I’ve started here with SGs. As I said above, I wanted to define the positional groups based on where people were playing, rather than where they were listed, so I’ve gone through all of your DCs to tease out who to call you starting SG. In the case of some teams that have cycled a few guys through, I’ve listed them all (within reason). The Suns for example have two representatives: Julian Wright and John Wall. I did the final bit of legwork for this before sim 8 was posted, so don’t yell at me if these numbers aren’t quite the same as what you see now. Finally, for those that care, across all of what I’ve deemed the SG starters in SLN, the average line is 19.8 PPG, slashing 46/40/85, with 5.5 boards, 1.5 steals, 0.3 blocks, on 1.8 turnovers a game.
Without further ado, I present a rough attempt at a Points Gained player ranking system. Hope you enjoy!
Well, it recently occurred to me that you could convert a player’s defensive and rebounding output to possessions gained, and then from possessions gained to an estimation of points gained.
Offensive points gained are pretty easy to calculate - working from the assumption that passing doesn’t matter at all, it should be Possessions Used * (PPP - a baseline PPP). To capture defensive points gained, I’ve followed the same formula to a degree, capturing the rebounds, steals, and blocks above some average baseline. From there, I make an assumption about what percentage of each of the above are truly possessions won (always the case for a steal, and seemed like 0.6 was the right coefficient for blocks and rebounds). Initially I did the above based on the per-game stats logged for each player, but realized that dramatically overstates the impact of starters and higher-minute types, so I then converted all of the above to per-minute played metrics, and then output both a per-minute and a per-game Points Gained score.
The other somewhat special sauce is calculating what the baselines should be. For example, at first I started by ascribing 1 point per possession for each calculated defensive possession gained. I found eventually that this was a little high, and consequently overstated defensive impact. The median team-level PPP obviously fluctuates a bit over time, but this year at present is 0.927.
In terms of the player-level baselines, I wanted to solve for the different opportunities afforded to different positions to record points / rebounds / blocks, so I’ve broken the pool of players down by the positions they actually play. This was annoying, and I wouldn’t recommend doing it by hand. From there, the other big stylistic choice was whether to use as a baseline the average across the data set, or instead, try to determine a replacement value and use that. To keep things easy here, I chose the former.
Some obvious things like positional defense are not considered in these metrics. I thought about trying to build something on top of them that would layer in a defensive rating, but that ultimately seemed messier than just ignoring non-stock and rebound defense. Also all of the choices I had to make above definitely have a dramatic impact on the final product - I tried to go with what made the most sense, but who knows what percentages of blocks are ultimately recovered by the team, ya know? When Tim Hardaway Jr gets a steal, is he really passing, or am I getting a THJ shot?
I liked listening to the PG rankings the Over/Under Pod fellas made up, which was part of the inspiration to publish this. I might get to PGs by the end of the year, but wanted to let their list breathe a bit, so I’ve started here with SGs. As I said above, I wanted to define the positional groups based on where people were playing, rather than where they were listed, so I’ve gone through all of your DCs to tease out who to call you starting SG. In the case of some teams that have cycled a few guys through, I’ve listed them all (within reason). The Suns for example have two representatives: Julian Wright and John Wall. I did the final bit of legwork for this before sim 8 was posted, so don’t yell at me if these numbers aren’t quite the same as what you see now. Finally, for those that care, across all of what I’ve deemed the SG starters in SLN, the average line is 19.8 PPG, slashing 46/40/85, with 5.5 boards, 1.5 steals, 0.3 blocks, on 1.8 turnovers a game.
Without further ado, I present a rough attempt at a Points Gained player ranking system. Hope you enjoy!
Name | Possessions Used | Total Points | Raw PPP | Points Gained Offense | Points Gained Defense | Total Points Gained | Points Gained per Minute | Total Points Gained per Game |
---|---|---|---|---|---|---|---|---|
Deron Williams | 968.16 | 1099.80 | 1.14 | 130.77 | 13.13 | 144.62 | 0.10 | 3.71 |
Victor Oladipo | 615.72 | 693.00 | 1.13 | 76.72 | 42.76 | 121.39 | 0.10 | 3.47 |
Monta Ellis | 875.88 | 946.20 | 1.08 | 69.53 | 48.16 | 119.52 | 0.09 | 3.15 |
Carmelo Anthony | 919.88 | 997.20 | 1.08 | 76.49 | 27.00 | 105.05 | 0.08 | 2.92 |
Tyreke Evans | 1080.80 | 1201.20 | 1.11 | 119.42 | -8.16 | 110.67 | 0.07 | 2.63 |
Derrick Rose | 769.46 | 837.00 | 1.09 | 66.85 | 5.72 | 72.89 | 0.07 | 2.35 |
Gordon Hayward | 934.68 | 975.80 | 1.04 | 40.28 | 33.13 | 74.95 | 0.06 | 2.20 |
Luol Deng | 547.90 | 547.80 | 1.00 | -0.59 | 68.14 | 70.89 | 0.05 | 2.15 |
Yaroslav Korolev | 764.98 | 766.50 | 1.00 | 0.83 | 70.36 | 74.06 | 0.05 | 2.12 |
Tim Hardaway Jr. | 914.26 | 1026.30 | 1.12 | 111.22 | -51.14 | 57.84 | 0.05 | 1.75 |
Terry Rozier | 422.18 | 469.00 | 1.11 | 46.44 | 11.38 | 58.43 | 0.06 | 1.67 |
Jimmy Butler | 969.24 | 1029.80 | 1.06 | 59.69 | -6.93 | 52.56 | 0.04 | 1.38 |
Devin Booker | 769.40 | 809.10 | 1.05 | 39.01 | -4.72 | 34.01 | 0.03 | 1.17 |
Jaylen Brown | 799.56 | 834.90 | 1.04 | 34.62 | 3.42 | 38.05 | 0.03 | 1.15 |
Klay Thompson | 863.62 | 910.20 | 1.05 | 45.80 | -27.67 | 16.70 | 0.01 | 0.45 |
Thabo Sefolosha | 511.72 | 518.00 | 1.01 | 5.82 | 5.04 | 10.75 | 0.01 | 0.29 |
Julian Wright | 271.48 | 217.80 | 0.80 | -53.92 | 50.36 | -0.96 | -0.00 | -0.03 |
Marcus Morris | 503.28 | 489.60 | 0.97 | -14.13 | 12.53 | -0.62 | -0.00 | -0.02 |
Sergio Llull | 292.36 | 288.00 | 0.99 | -4.62 | 2.09 | -2.39 | -0.00 | -0.10 |
Zach Lavine | 887.24 | 885.00 | 1.00 | -3.04 | -1.55 | -4.71 | -0.00 | -0.15 |
Ben Gordon | 272.28 | 275.40 | 1.01 | 2.87 | 2.02 | 4.89 | 0.00 | 0.11 |
Shabazz Muhammad | 552.24 | 542.70 | 0.98 | -10.04 | -15.42 | -25.46 | -0.00 | -0.80 |
Lou Williams | 773.12 | 787.50 | 1.02 | 13.68 | -4.86 | 8.83 | 0.00 | 0.27 |
Danny Granger | 567.08 | 540.00 | 0.95 | -27.59 | -17.72 | -45.31 | -0.00 | -1.32 |
Harrison Barnes | 683.60 | 680.00 | 0.99 | -4.22 | -2.56 | -6.78 | -0.00 | -0.16 |
Bradley Beal | 650.72 | 636.90 | 0.98 | -14.41 | -4.16 | -18.58 | -0.00 | -0.62 |
Jeremy Lin | 593.30 | 577.10 | 0.97 | -16.74 | -2.05 | -18.79 | -0.00 | -0.74 |
Taurean Prince | 652.20 | 589.00 | 0.90 | -63.79 | -7.18 | -70.97 | -0.01 | -2.36 |
Brandon Roy | 336.48 | 328.00 | 0.97 | -8.78 | -4.69 | -13.47 | -0.01 | -0.45 |
Chris Paul | 411.28 | 379.60 | 0.92 | -32.05 | -9.32 | -41.37 | -0.03 | -1.39 |
John Wall | 384.12 | 321.00 | 0.84 | -63.47 | -5.34 | -68.81 | -0.02 | -2.31 |
Bogdan Bogdanovic | 523.58 | 481.00 | 0.92 | -43.05 | -6.76 | -49.82 | -0.01 | -1.51 |
OJ Mayo | 337.8 | 292.9 | 0.87 | -45.21 | -8.59 | -54.26 | -0.12 | -1.87 |
Jalen Brunson | 575.48 | 540.8 | 0.94 | -35.2 | -36.66 | -73.62 | -0.07 | -2.30 |
Last edited by quady on September 24th, 2024, 7:33 pm, edited 2 times in total.
Ranking Shooting Guards by Points Gained
This is fucking cool!


2008, 2012, 2014
Ranking Shooting Guards by Points Gained
this is awesome - love seeing new metrics!!!
AND DERON WILLIAMS IS THE BEST SG IN THE LEAGUE FOLKS!!!
AND DERON WILLIAMS IS THE BEST SG IN THE LEAGUE FOLKS!!!
Ranking Shooting Guards by Points Gained
This is cool - are these names sorted in a particular order?
Celtics ah the balls
Ranking Shooting Guards by Points Gained
Seeing that John Wall is ranked at the very bottom, I’m guessing best to worst SGsJoe wrote: September 24th, 2024, 10:05 pm This is cool - are these names sorted in a particular order?
Ranking Shooting Guards by Points Gained
Lol I guessed that, but I was looking at the total points gained column and its not quite in that order
Celtics ah the balls
Ranking Shooting Guards by Points Gained
It was supposed to be sorted by the per-minute metric, but something wonky seems to have happened at the endJoe wrote: September 24th, 2024, 11:52 pmLol I guessed that, but I was looking at the total points gained column and its not quite in that order
- GrundleSauce
- General Manager
- Posts: 740
- Joined: December 20th, 2023, 12:13 pm
- Status: BUMPIN. BOUNCE. BOUNCE.
-
- Newest Posts
- Top Active Users
- Newest Users