Go quote me on that because I never said that. I said an X-fold change over random choice is a better metric.
But what exactly do you mean by "summing up"? All commanders in one mode? One commander in all modes? All in all?
Nevertheless, summing them up as you suggest is the most misleading way possible. This way your data represents 70-80% 4v4 with a bit of 3v3 and 2v2 and the tiniest touch of 1v1.
I did not suggest to go the route of summing up. I used the word "should" which is another probablity of showing data. Also, I asked Vipper to clarify what she did, as it is neither a ratio, nor a popularity value. But my post of asking for clarification has been removed.
And yes, you did say that a different metric, you said it yourself there again. The quote you are asking for is your post. Additionally:
Summing up game modes makes as much sense as not summing them up, it just depends on what you want to show. And while Vipper failed some important information, he at least stated that it is across all modes. The raw data does not show that at all.
Literally, we both are saying exactly the same thing. With the difference that you are adding to Vipper's posts, where as I am asking solely her to provide support to her posts without having a moderator to back up the data.