First off you shouldn't be using shot stick in this test. Its increases the amount of variables and you should eliminate all variables you can. Second, consistency would be a factor in whether Melo at 85 or Novak 65 should be on the court at the end of the game and hitting these shots. Also, i think its clear it would be easier to a good look with Melo as he is faster, quicker, better ball handler, etc.
I think your theory should also be two parts- covered/contested jumpshots vs open jumpshots. This is important as we expect Melo to dominate in covered jumpshots and could easily be a push in open jumpshots based on the 87 (Melo) vs 81 (Novak) med rating. Open jumpshots would be easier to test as you could do that in practice mode. Covered jumpshots is a bit harder as you need to do that versus an opponent, you would need to have the same player guard them, and you would need to do similar shots and test all the various hot vs neutral vs cold zones. A sample size of 100 in each zone would likely be somewhat telling. That would be 10 areas of jumpshots each (a 1000 shots total). You should label the test- covered vs open (who against if covered), area on the floor, the hotzone color, and the player used for each.
You may or may not care enough to do this but its premature to say this is why your missing open jumpshots or ask if your better off with Korver and Novak without doing more than a 5 minute video of showing some shots one way or the other. I did a quick 25 open shot (neutral hotspot) with Melo at the Right side med area and he was 4-13 to start, but then he finished 13 for 25. Bottom-line, I have no idea if your theory is correct or not but you really don't know until you test it correctly.

lol

Comment