Posted by savoybg on 4/17/2025 9:45:00 PM (view original):
Posted by Midge on 4/17/2025 9:27:00 PM (view original):
A big issue I have with it is that you treat win shares from 1972 in the ABA the same way you treat win shares from 2022 in the NBA.
No I don't. I've mentioned at least 3 times already in this thread that I reduce the ABA numbers some in the formula. Every player gets their ABA seasons reduced by between 8% and 15% depending upon their playing time and their win shares total for that season. So a guy with 10.0 win shares might only get credit for 9.0 win shares when I calculate his rating.
By the way, this sim gives ABA seasons the same value as NBA seasons.
We stopped talking about the sim 10 pages ago, follow along. We are talking about your stupid math. Honestly, you should discount all players in the ABA era. For data purity, you should reduce the ABA 2.4% over the NBA in the same era and then discount them both.
BUT
Why reduce between 8-15% the ABA alone based on playing time and win shares? If playing time is this much of a factor, why do you not assign a +/-to all players outside of 2500-2800 minutes? Do you see where your logic has failed you? You can't say playing time should be adjusted here, but not adjust it for the rest of the dataset.
What am I saying? Of course, you can adjust it. You are a scientist with an 148 IQ, but I adjusted your IQ for posting time and losing at life shares to 14.8.
Now...go eat poop. It is #17 on the list.