I am a little confused on how Falco isn't first on the Jump Height list. He has the highest initial jump in the game, and even if it isn't(it is) It'd still be higher than MK's.
Edit: Your results are probably a bit skewed because(as long as I'm reading it right) there were actually two factors in your test: jump height and horizontal air speed. You might want to find some other way of testing.
Edit2: Fall speed is also a factor.
I do not use horizontal air speed. I do not fall to test anything. I think you did not understand my method. Please read carefully. ^^ I explained it a lot.
Supersun. You may be right. I was thinking about that. But if you look at it. In terms of KO potential, a move that cannot kill is truly no option for a character, so it deserves no points. A move that can kill at some percentage may be an option but it is true that will be rarely used (who gets to 400%?).
I thought that having few powerful moves, rather than a lot of weak moves that can kill at very high percentages, is probably better. So, instead of considering the score for each move linearly, we could consider it like a curve. Our exponent could be 2 (we must decide this number).
Example:
A move kills at 150%. We have 999 - 150 = 849. Then 849^2 = 720801 points
A move kills at 100% . We have 999 - 100 = 899. Then 899^2 = 808201 points
A move kills at 50% . We have 999 - 50 = 949. Then 949^2 = 900601 points
The point is that score does not increase linearly with (999% - KO%) but exponentially (a curve if you make a graphic). As we are getting giant numbers, I think we should divide these scores by 10000.
This way, attacks that kill early will get a much more points, and those who kill late will get fewer points than before (relative to more powerful moves). Moves that cannot kill will still get zero.
Did I explained myself properly? What do you think of this method?
If you want to get a KO% average is ok with me. But this new scoring method does not match with that average I am afraid.
I was also thinking we could use 300% (sudden death value) for the substraction and not 999%. Killing someone with an attack that kills beyond 300% is unllikely to happen anyway. Maybe we could just change this on the method and do it linearly. (Also, this way, the average ranking will match with the score ranking).
Note: For those who are testing this, I suggest we get the KO% for each move until we decide the best scoring method.