OldFan wrote:

EwingsGlass wrote:

SupremeCommander wrote:I like the idea of this equation. I think you need to tinker with it more. A possession can end with no output regarding a PG's stats. And the ball doesn't necessarily have to be controlled by a point during every possession. Ideally you'd like to reward a point for initiating the offense and the outcome netting points.

Good idea though

I was thinking the same thing. The PG can pass the ball and result in a missed shot by the recipient. On the otherhand, the PG pass it off to a player standing 27 ft from the basket who throws up a prayer (which goes in) and it counts as an assist. My thought is that the missed shots can be normalized over time if you give a FG% adjustment... the only issue there is whether the PG has a positive effect on the team's FG% or whether the team's FG% has a positive effect on the PG's output in this equation.

I think we can all agree that the quality of players surrounding Nash and Rondo are much better than the quality of players surrounding Chris Paul>

Also, there is an obvious bias against shoot-first PGs. Except for guys who drive more often than shoot (i.e. Rondo, Tony Parker), their FG% are going to be less than 50%. However, the Assist to Turnover Ratio is going to be a positive number for most PGs and should be around 3 o 3.5 for the best PGs. What happens is that the Assist to Turnover Ratio is skewed very negatively for guys who take a lot of shots (like Isiah Thomas did).

Yeah - I noted this bias in my original post. That's why Oscar, West and Fraziers numbers really standout. They shot a lot, didn't have the advantage of the 3 pointer and still had good ratios.

I agree it needs some work. I think trying to account for possessions where the play didn't terminate in the PG getting assist, making a turnover or taking a shot would require watching film. Don't know how you'd account for that based on box scores.

You might be able to adequately fake the stat on possessions where the possession where it ended on a non-assist or a missed shot. First, you can figure out what percentage of minutes the PG played out of the total minutes for the team's season. Second, you could figure out what percentage of assists the playmaker PG had compared to the rest of a team as a whole. Third, the number of turnovers for the PG vs the rest of the team. Fourth, figure out the percentage of shots the PG took compared to the total number for the team. Fifth, the number made shots the PG took. And Sixth, the numer of missed shots the PG took.

So instead of (Assists+Made Shots)/)TO Plus Missed Shots)

I would look at {[(assists by PG/Total Assist by Team) - (total shots by PG/ total shots by team)]/[(Total TO by PG/Total TO made by Team)-(Total Missed Shots/Total Missed Shots By Team)]} all of which is multiplied by (total minutes by PG/Total Game Minutes).

You'll still have to get deeper in there to account for foul shots and 3 pters, but this will get you closer and remove some of the team biases, I think.

EwingsGlass wrote:

OldFan wrote:

EwingsGlass wrote:

SupremeCommander wrote:I like the idea of this equation. I think you need to tinker with it more. A possession can end with no output regarding a PG's stats. And the ball doesn't necessarily have to be controlled by a point during every possession. Ideally you'd like to reward a point for initiating the offense and the outcome netting points.

Good idea though

I was thinking the same thing. The PG can pass the ball and result in a missed shot by the recipient. On the otherhand, the PG pass it off to a player standing 27 ft from the basket who throws up a prayer (which goes in) and it counts as an assist. My thought is that the missed shots can be normalized over time if you give a FG% adjustment... the only issue there is whether the PG has a positive effect on the team's FG% or whether the team's FG% has a positive effect on the PG's output in this equation.

I think we can all agree that the quality of players surrounding Nash and Rondo are much better than the quality of players surrounding Chris Paul>

Also, there is an obvious bias against shoot-first PGs. Except for guys who drive more often than shoot (i.e. Rondo, Tony Parker), their FG% are going to be less than 50%. However, the Assist to Turnover Ratio is going to be a positive number for most PGs and should be around 3 o 3.5 for the best PGs. What happens is that the Assist to Turnover Ratio is skewed very negatively for guys who take a lot of shots (like Isiah Thomas did).

Yeah - I noted this bias in my original post. That's why Oscar, West and Fraziers numbers really standout. They shot a lot, didn't have the advantage of the 3 pointer and still had good ratios.

I agree it needs some work. I think trying to account for possessions where the play didn't terminate in the PG getting assist, making a turnover or taking a shot would require watching film. Don't know how you'd account for that based on box scores.

You might be able to adequately fake the stat on possessions where the possession where it ended on a non-assist or a missed shot. First, you can figure out what percentage of minutes the PG played out of the total minutes for the team's season. Second, you could figure out what percentage of assists the playmaker PG had compared to the rest of a team as a whole. Third, the number of turnovers for the PG vs the rest of the team. Fourth, figure out the percentage of shots the PG took compared to the total number for the team. Fifth, the number made shots the PG took. And Sixth, the numer of missed shots the PG took.

So instead of (Assists+Made Shots)/)TO Plus Missed Shots)

I would look at {[(assists by PG/Total Assist by Team) - (total shots by PG/ total shots by team)]/[(Total TO by PG/Total TO made by Team)-(Total Missed Shots/Total Missed Shots By Team)]} all of which is multiplied by (total minutes by PG/Total Game Minutes).

You'll still have to get deeper in there to account for foul shots and 3 pters, but this will get you closer and remove some of the team biases, I think.

Wow - You spoofing me :)

orangeblobman wrote:There you have it, Duhon is NOT the worst point guard in the world . Nice statting, Oldfan.