I don't know, but my guess is the H or A pawns.
Which file gets the most pawn promotions?

Is there any statistics or study that shows which file is more likely to get a pawn promotion?? A lot of the endgames I watched from GMs get promotions from the B and G files. What do you guys think?
That would definitely be an interesting study. But to clarify - are you talking about the pawn's starting rank, or the rank on which the pawn gets promoted? I think it's the latter, but I want to make sure (and the former is also pretty interesting to consider).

Is there any statistics or study that shows which file is more likely to get a pawn promotion?? A lot of the endgames I watched from GMs get promotions from the B and G files. What do you guys think?
That would definitely be an interesting study. But to clarify - are you talking about the pawn's starting rank, or the rank on which the pawn gets promoted? I think it's the latter, but I want to make sure (and the former is also pretty interesting to consider).
Both actually. Where it started and which file it got its promotion. It would definitely be an interesting study.

Something else to consider is that often times players will resign when they see that they cannot stop their opponent from promoting a pawn. Would that count as a promotion for the purposes of the hypothetical study? And what about situations where one side has two passed pawns, and the opponent can stop either one, but not both. Say black resigns in this position:
(black to move)
How would you count that? 1/2 of a promotion for both pawns?

Yes, this is a statistical study, thus this sort of situations will sum up and give something sensible.

I have thought on this topic before as I have observed games, and I have noted that the c- and h-files have tended to lead most often to Pawn promotions, although I have no hard data to support this.
I searched my database with Chess Assistant, finding all games where promotion occurred on each file. The database has almost 5.37 million games so you can get a sense of how often promotions happen even with the issues of multiple promotions and prior resignation mentioned above.
a | b | c | d | e | f | g | h | Total | ||
White | 30777 | 27375 | 26676 | 26544 | 21049 | 21422 | 17422 | 20041 | 191306 | |
% Total | 16.1% | 14.3% | 13.9% | 13.9% | 11.0% | 11.2% | 9.1% | 10.5% | 100.0% | |
Black | 27184 | 24717 | 23673 | 21776 | 19458 | 18965 | 15656 | 17793 | 169222 | |
% Total | 16.1% | 14.6% | 14.0% | 12.9% | 11.5% | 11.2% | 9.3% | 10.5% | 100.0% |
Is there any statistics or study that shows which file is more likely to get a pawn promotion?? A lot of the endgames I watched from GMs get promotions from the B and G files. What do you guys think?