Re: More Assists with better teammates.
The guys over at Bullets Forever did some really interesting analysis at the end of last year. It started with an article called "John Wall's Missed Assist Tracker"
http://www.bulletsforever.com/2012/2/27/2827243/john-wall-missed-assist-tracker-washington-wizards
And finished with, you guessed it, the "Rajon Rondo Missed Assist Tracker"
http://www.bulletsforever.com/2012/3/2/2838291/rajon-rondo-missed-assist-tracker-john-wall
The takeaway from the comparison was that:
Rondo's Boston teammates convert 55.9 percent of his assist chances into actual assists, while Wall's Wizards convert only 43.9 percent of his chances.
That was for last year. It would be interesting to see how Rondo's assist chances this year compared.
This analysis patently ignores the most important part of passing, that is whether the passer actually gets people better shots (if he does, they'll shoot at higher percentages). So I'm curious how much different the shooting %s are for the same set of players off of Rondo's passes as opposed to other people's passes.
Not sure exactly what you're asking but the info provided is this:
________________________________________________
Total makes by Boston so far this year: 1166 Total Makes: 2564 Total turnovers: 500
Total non-Rondo assisted makes: 1166-231 = 935
Total non-Rondo assisted “opportunities” (attempts + turnovers): 3063 – 413 = 2650
Celts non-Rondo assisted makes/opportunities: 935/2650 = 35.3%
Celts Rondo-assisted makes/opportunities: 231/413 = 55.9%
Difference: 20.6%.
For Wall this number is 9.2%.
This is as close as I can get to the number Kev cites in the 82games article, where the league average is 8%. I would need to take “other assisted baskets” out of the “non-Rondo assisted” baskets to do that.
___________________________________________________
So he's taking the conversion rate on chances that come from passes from Rondo (56%) and comparing it to the conversion rate for all other chances (35%). The difference is 21%, the average difference for assisted vs non-assisted is 8%. The numbers for Rondo aren't accurate but are probably fairly close.
The main problem is that he's lumping all the turnovers in as scoring opportunities when this clearly isn't the case, but the error for this is probably 5% (ish). On the flip side, the 82games study was comparing assisted scoring opportunities to unassisted scoring opportunities. He isn't comparing Rondo-assisted scoring chances to our unassisted chances, he's comparing the Rondo-assisted chances to all other chances, but about half of the chances that aren't assisted by Rondo were assisted by someone else, which are still 8% (on average) more efficient than actual unassisted chances, so the non-Rondo chances are about 4% higher than unassisted chances would be. Make sense? (it's getting late, my post may not be that digestible).
The end result would (and this might be pessimistic) have our boost in scoring efficiency from Rondo's passes to be more than double what you'd see from "average" passes.