APBRmetrics

I'm new to analytics and have a question that I need a little direction on. The question is:

How much does an increase in team 2FG% from 43% to 44% affect the number of games they should expect to win on the year?

I've seen the Pythagorean expectation method and eFG% but I don't know how to get it for ONLY 2FG and not 2FG and 3FG combined. Any help would be appreciated.

First, refer to this series: http://www.d3coder.com/thecity/2010/12/ ... rs-part-2/

Then, convert the 2Pt% to EFG%. Since 2 Point FGs are about 75% of shots, increasing 1% in 2FG% would be about 1%*.75 = .75% of an increase in EFG%.

Per the post referenced above, 1 point of EFG% is worth about 1.45 points of point differential.

And one point of point differential is worth 2.54 wins.

So your increase of 1% in 2FG% would be worth .75*1.45*2.54 = 2.75 wins.

how did you arrive at eGF% giving the 1.45 pd? Thank you

Given this year's averages roughly 3.4 wins more, assuming all else is held equal.

This year teams took an average of 5084 two point attempts over the course of an 82 game season. Making 43% of shots will net you roughly 4372 points (0.43*2*5084), whereas making 44% will net you roughly 4474 points (0.44*2*5084). Taking the difference of these we see that we gain about 102 points scored over the season.

Now, the marginal value of points in terms of wins is known to be roughly 0.033 per point. This leads us to a gain of: 102*0.033 = 3.366 wins

This ignores that in fact since fewer are missed there will likely be fewer rebounded and hence fewer attempts in all. However, the difference caused will be small relative to the gain in points from the increased percentage.