Alex: Proof that FTA is huge and needs addressing now, and how to fix it

[awpertunity]

(09-24-2012 06:40 PM)Ravernoth Wrote:  

(09-24-2012 08:59 AM)awpertunity Wrote:  As Eijolend says, your 2.7 : 1 ratio is actually being conservative. This is skewing the outcome closer to 1 : 1 by including the full wins. (as an extreme example: if a super titan plays against a fluffy player, he will be able to win as P2 just as easily as P1)

So as Eijolend says, we should really be looking at the draw matches:

So far there have been 19 draws in the tournament (each player winning one game). Of these 19 draws, 17 matches are both P1 wins, and only two matches are P2 wins, or an 8.5 : 1 ratio.

This is huge. If there was no FTA (or even a SMALL FTA, this indicates not only that there is an FTA but it is quite large...), this is the number that should be close to 1 : 1.

Looking for results of pairs of matches i.e. 2 P2 Wins vs 2 P1 Wins will multiply the FTA by squaring it.

It's true that the number will be 1 : 1 if there was no FTA, but that's because it would be 1 squared.

If FTA was 2 : 1, the ratio of 2 P2 wins : 2 P1 wins would be 4 : 1

In this case, FTA was 2.7 : 1 , the ratio of draws would be 7.3 : 1 which is close to what you found 8.5 : 1

I take your point about mismatched skill levels lowering the appearance of FTA, so we could remove the results from some of the lower ranked players and see how it affects the numbers, but overall the tournament is pretty competitive.

And anyway, FTA is big

Edit: We could also take the draw matches and work backwards, in which case 8.5 : 1 draws would mean an FTA of 2.9 : 1

lol what? I don't understand this squaring business.

8.5 : 1 came from a total of 38 games.

34 P1 wins : 4 P2 wins = 8.5 : 1

Regardless, the important thing is we both agree FTA exists and is large