Updated: Mar 14
An Elo rating is a number intended to calculate and represent the relative likelihood of winning between competitors in a pool that play in a one vs one competition against one another regularly.
Every competitor in a division or pool starts with the same rating. 1500 is a common one and the one I used. When two competitors have a match or game, the winner's rating goes up, while the loser's goes down. If the winner had a higher rating than the loser, (as expected) the amount of the adjustment is lower. If there is an upset, the adjustment is higher.
It isn't a perfect system, it tends to over-rate competitors who my common sense tells me are being fed tin cans (GUNN CLUB!) but any attempts to improve this make its predictions worse elsewhere, (increased the sum of squared error,) so I've settled on it as the best representation I have. Below are the top baker's dozen of each AEW division based on Elo through the Revolution supercard. There is a column with AEW's last official rankings before the show for comparison, as well as a column for the win record, which is just the number of losses subtracted from the number of wins. If someone has a higher Elo than someone with a better win record, it indicates that they must have had a stronger schedule. Also there is a line graph showing the change in AEW competitors' Elo over time, starting from the very first show.
Keep in mind that this only includes matches which are comparable with one another: standard-rules AEW one-on-one matches, (or two-vs-two in the case of the tag division.)
I hope that you all find this data interesting or thought provoking.