I've played about 160 of these (finished) so far, and have won around 95 of them. That's about 59%. I don't know if that makes me any kind of expert, but here are some of my observations so far.
No Spoils versus Cards: The spoils might arrive too late to affect the outcome of the game. Flat Rate is probably the most random spoils setting. Escalating rarely gets past 6 or 8 armies. On the other hand, a turn 4, 10-army set after collecting 3 cards in a flat rate game can turn the tide, especially if the other guy is holding all red cards at the time.
Consider this advice to be most applicable to No Spoils games. The other constant settings are unlimited reinforcements, and fog of war off. They're all casual, sequential games.
The Drop: Moving 1st with a good drop can decide the game. If you can get into a territory grouping (+2 or +3 bonus) or Luxembourg City (+1 bonus), with the other player's armies in disarray, the remainder of the game could be a mop-up.
If you move first with a bad drop; a drop where your armies are disconnected by the other player and neutrals, consider fortifying a single spot and waiting to see what the other player does.
Whenever the other guy takes The City or a territory grouping bonus, you have to have a stack ready to respond.
If he gets the good drop, and goes first, and you're faced with an opponent occupying a reinforced territory grouping; wherever at all feasible, make the early counterattack. Ideally, you want to end up with a bigger stack than your opponent's, right in the middle of the grouping. This way, he no longer has a chance of taking it. If possible, seize the entire bonus away from your opponent; take the entire territory grouping.
Often in the case of 'breaking up a continent,' just getting one of the territories is all you can do.
Sometimes, a player gets such a good drop and 1st turn that the player moving 2nd is basically out of the game. It's impossible for the 2nd player to break into the 1st player's bonus.
If you get a drop with say 3 territories on one side of a neutral or opponent block, and 3 territories connected on the other side, attack the connecting territory so you can assemble all of your armies through reinforcement. Sometimes more than one strongpoint is desired.
Usually though, a single strongpoint, combined with neutral hedges, and a few dislocated 3-army territories of your own can keep the territory groupings out of the other player's hands.
Be ready to break up the other player's bonus (territory grouping, or the City), wherever it happens. Try to grab bonuses of your own in the process.
Some games are won by one player simply occupying Luxembourg City, and slowly pushing the other guy back from there.
Many games are decided by a series of see-saw battles where each side builds 3 armies, places it on a 1 for a total of 4, and attempts to either a) secure a territory grouping (or the city), or b) break-up an opponent's bonus, or c) a combination of both a) and b).
The auto-rolls can produce instant victory or defeat.
Rarely is a stack equalling or greater than 20 armies seen in this type of game.
When playing according to a style sheet, I try to follow these rules:
1) Through a combination of (if necessary) link-up attacks, and reinforcement moves; mass the armies at a single point, or sometimes multiple points if the situation requires it.
2) Go for the easy bonus.
3) Always be prepared to take out your opponent's largest stack. No matter what the odds, make that attack.
4) Use the Neutrals to shield bonuses where possible.
5) Break up your opponent's reinforcement routes.
6) Always try to keep a stack of more than 1 army, somewhere on the map, and in a good location to break up bonuses. If you get down to all 1s, and your opponent has one or more stacks of multiple armies, the game is probably over; particularly in No Spoils.
Be proactive; always wipe out their biggest stack if possible. 6:8, 5:6, 11:5, 9:8; whatever the odds, make that auto-roll.
There is a lot of luck in these games, but sound tactics can help win more games over time.
Hopefully these thoughts will be of use to someone.
