OK, I just went through my first 11 1v1 games (I didn't have any team games on this map). You would of course want to take a random sample of all games, but this is at least instructive.
Your independent variables should be first turn (binary- "1" if Team 1 goes first, "0" if not), Team 1/Player 1 decaying regions, Team 2/Player 2 decaying regions; and your dependent variable would be Result (binary-"1" for Team 1 winning, "0" if not. You should also include the game number as the unique identifier for each row of data.
You will then want to run a statistical analysis to see what weight each variable has. If the decaying regions are deciding too many games, you should be able to see this from the statistical results. I think you can do this kind of stuff in excel, although it would be nice if you could get someone to code something to pull this data for you.
The coding to pull the data will be far easier in 1v1 games. Rarely is it a good strategy to attack a decaying region on your first turn, if you go first. So the code would simply look at how many times "received -1 for holding XXX" appears during that first turn for each player. You could probably get all the info you need about this map based on that BUT the results would be skewed toward your desired result (I think) because it seems like the effects of an "imbalanced" drop would be amplified in a 1v1 game.
You could use the same coding in a team game, but the larger the teams, the more likely it is that after a couple of forts, it may become advantageous to hit another team's decaying region before that player has played (especially in flat rate/escalating games, where the decaying region might be the player's only/best shot at a card). Thus the code mentioned above should still serve as a decent proxy for the information that you want, but there may be some territs that get missed.
The way around this would be a little fancier coding. You would have the names of all decaying regions in a set. The code would count player 1's decaying regions at the start of his turn. If during his or any subsequent turn, a decaying region of a player who has not yet played is successfully assaulted, then a +1 is added to the count that will be derived when that assaulted player begins his turn.
If you take the step of programming this additional element, then there is no reason not to use it for the 1v1 games above, even though it won't gather you much more data.
You will really only want to look for games that have two teams/players. Unbalanced drops in multi-player/team games shouldn't really matter as much, for obvious reasons.
Finally, if this code is written in the manner that I described for teams, then this will actually be a useful foundry tool. Inputs would be map name; a user-defined critical territ set (in our case, the decaying regions) a game range (so that a mapmaker can limit results to only the period of the most recent gameplay changes); and perhaps a game-type limiter (perhaps options would be 1v1, team v team or both). The output would tell you, within the selected territories, how much the following 3 things determine results: first turn, quantity of player1's drop w/i selected territories and quantiy of player/team 2's drop w/i selected territs.
Example: I want to know whether the amount of territs dropped in Australia (or Australia and South America or Australia, South America and any immediately adjacent region) has a decisive effect on the end result. I could run this code and find out.