In our last game, Stalin played very perversely: he
sabotaged the war effort whenever he could (seizing
the declaration of war on Japan whenever possible, and putting
all production he could get his hands on into Offensive Support
markers on his Far Eastern front). He played hard for Pol/Mil.
At the end of the game, neither Germany nor Japan had
surrendered, and his lead in points survived the randomizing
die-rolls. We, FDR and Churchill, in our inexperience and
inability to imagine someone not "doing their bit" as one of
the allies, were thrown for a loop.
In our post-game conversation, we decided that, in historical
terms, Stalin had collaborated with the Axis as much as possible,
and had gained a mutually profitable arrangement with them after
the war, to our detriment. Stalin added to this, "My Far East troops
were fraternizing with the Japanese troops instead of fighting
them, playing poker. And throwing all those weapons represented
by the Offensive Support markers into the Sea of Okhotsk."
In retrospect, I suppose we (FDR and Churchill) should have
coordinated our card-play, debating Stalin and never each other,
and tried to seize control of more DO's and production, to force
Stalin to fight Germany.
. . or each of us could have imitated him, ignoring the war
as much as possible, and concentrating on Pol/Mil.
It's important to keep in my the three victory conditions and the various VP permutations that will implicate each. Or, more precisely, which VC each player will want to pursue based the incentives of each victory condition. Under such an analysis, I think you'll find that everyone pursuing a pol-mil heavy strategy is not viable.
Broadly speaking, if everyone pursues a pol-mil heavy strategy, you're essentially guaranteeing a condition-3 victory. As such, you will be at the mercy of the die rolls at the end that modify the scores. If everyone stays roughly equal in VP, the winner will essentially be randomly selected. This is not - or at least should not - be appealing to anyone. You simply do not play a game the length and depth of Churchill for the winner to be determined randomly. To make the strategy viable, you need a relatively comfortable spread between you and both 2nd and 3rd place. (Another point to keep in mind is that the global issue is a potential 10 point swing between any pair of players; this makes the VP situation going into the last conference inherently unstable, especially in the context of a condition-3 end-game). In essence, to make a pol-mil strategy viable for you, you have to build a sizeable VP lead through pol-mil over the other players throughout the game
But what does such a VP lead do to the incentives of the other 2 players? In short, it incentivizes them to pursue a condition-2 victory. If you're gobbling up VP through Pol-mil, then the 1st-to-3rd spread becomes more likely to implicate condition-2. Moreover, if you just keep blindly gobbling up VP, it will take the die roll out of play and make 2nd place the automatic winner. The caveat here is that condition-2 is predicated on axis surrender. But two players focusing on axis surrender makes such surrender virtually assured. (With some experience moving fronts becomes relatively easy and the game becomes one of determining which front to move and when).
It is also worth noting that it is typically not viable for 2 players to pursue a pol-mil/ignore-the-war strategy. Again, without Axis surrender, a close game between 1st and 2nd (with a distant 3rd) means the game will come down to an arbitrary die roll (plus the inherent instability of Global issues shifting VP in the last conference). The alternative is for one player to build enough of a lead over 2nd place throughout the game to be in more control of the situation. But at that point, 2nd place has a huge incentive to force condition-2 through Axis surrender. At that point, his incentive aligns with the 3rd place player.
Obviously, the instability created in the last conference by the swings the Global issue can cause will be inherent in any of the victory conditions. The question is how to mitigate the potential swing. One way to do this is to win the global issue on the last conference. The ideal situation to do this is when its the only issue you need to win at a conference. The more "other things" you need to do at the conference, the less likely you'll be able to win it.
It's also important to keep in mind that if you're pursuing a pol-mil, you likely need the Global issues not just for their VP, but for the benefits they bestow. As such, you've likely won both of yours; this makes your situation especially precarious. The better situation is to have won neither so you are not subject to the VP swing. But this is in tension with the necessity of winning the issues to gain their benefits.
Act Three! Everybody Dies!
(really smart stuff)
And truthfully, as I found, sticking to one strategy throughout the game can't work as long as one other player notices what you're going for. If it becomes obvious that you are going for, say, a Condition 2 and pumping up the 1st player's points, then the 1st player merely gets the 3rd player to help him prevent the 2nd player from achieving this, and give 2nd place to the 3rd place player. The 3rd place player will gladly take the deal, hoping that Condition 2 will be taken later in the game.
In my PBeM, we are in Conference 8, and I've watched the US go from Condition 1 to a Condition 3 and now back to a Condition 1 run, the UK is going for Condition 2, I think (the wiley bastard), and I'm hoping to delay things a Conference or two just so I can figure out which Condition I should gun for.