Open debate/discussion

[Snedecor]

Hi all,

I have and answer to open a debate or discussion about some metrics. For example the eFG, x player did 10 games you get the shoots and then take it into the formula and thats the accumulated eFG for that player but could be more appropriate statistically get the 10 eFG and get the average?

Thanks
Regards

[Crow]

Find the efg% for all shots at once. Finding efg% for each of 10 games and taking the simple average of them is not appropriate for finding the true average performance at shot level because the number of shots in each are unlikely to be equal. Doing it this second way would lead to unequal weighting of shots.

Doing it by game would tell you something new & different about consistency / likely performance level but you shouldn't do the second without the first or use the second as if it were the true average performance at shot level. The two methods will probably converge with time for most players. Do the second if it interests you as an add on but there is no debate which is the right method for the simple average performance.

[Snedecor]

Thanks Crow.

So the perfect thing here would be calculate the %efg for total and then do the average and the sd to see the consistency, making both and see.