One of the key features of any game with cards - be it Cribbage or Bridge or Race for the Galaxy or Washington's War - is that you can't ever be sure exactly what's in the other guy's hand. In a wargames context, this provides for a quick and painless sense of fog of war; as Mark Herman noted in his article "My Philosophy Behind Card Driven Game Design", CDGs let "the player know the strength and location of enemy forces" but keeps you in the dark regarding "what his opponent can do with those forces and where he intends to act."
The "problem" with card-driven games is that they're difficult to play solo. Now, that's not really a problem from a creative perspective. Different games are designed for different player counts to explore different dynamics and provide different experiences. Faulting a two-player game for not working for one is like faulting baseball for not working for four. When I design a two-player game, I don't necessarily go into it with the goal of making it easy to solo. It often ends up being solitaire-friendly, but it's not like I consciously avoid or excise things that make it hard to solo; it just happens that a lot of the things that I think make for an interesting two-player game also work well for one.
Where it becomes a "problem" is not so much on the creative side, but in marketing the game, because for the wargames market especially solitaire suitability rates very highly (much more-so than in the broader gaming world, where playing a multiplayer game single-handed is regarded as weird). Games that are hard to solo are hard to sell. At least, they're hard to sell to the "core" wargames market that place a premium on the solo experience; from all accounts, CDGs do a very brisk business and appeal to a broader group of gamers, and those sales more than make up for any lost due to poor solitaire suitability.
That doesn't stop some of us from trying, however. It's relatively easy to find proposed methods of playing CDGs with one player - Stuka Joe's in particular is very popular - and some recent CDGs have even jumped on the bot bandwagon to try and approximate some of the tension you get in a "real" game against another player. Me? When I can't find another player, I just put both hands in front of me and play with perfect knowledge of the cards available. Of course that, like Robert De Niro's steak, defeats its own purpose.
At that point, I'm not really playing the game at all, but rather a closely-related game. And for all the reasons that I might solo a game - to learn how to play it, to explore its model, to study the design - that closely-related game is usually close enough. I'm not operating in the same decision space, not experiencing the intended pressures, but I am still making decisions and experiencing pressures that are rich and interesting. The decisions in this slightly-different version of the game don't hinge on "what does the other guy have in their hand?" as it does in the real thing, but rather, "if I do this, what will the other guy do? Because the other guy can do this, is it worth it to do that?"
And it occurred to me that this in and of itself is an interesting and valid decision space for two players. Potentially it might even be more interesting, particularly if the game was built with that decision space in mind, and if its structure supported and emphasized those sorts of decisions. Really, it's a dynamic I've explored in several of my games - Optimates et Populares and the two Supply Lines titles are both perfect information games about predicting how your opponent will react. They're bluffing games where all your cards are on the table, only there are no cards at all. So the question I had in mind was, could I do this same kind of thing in a card-driven context?
The result is This Guilty Land, a political CDG with completely open hands. Any time you consider any move, you have complete knowledge of how your opponent might be able to counter it. And, borrowing the Political Will (PW) see-saw mechanism from Optimates et Populares, any action you take is likely to give them additional PW which they can use to play more and better cards. As you plot your own path, sketch out your own chaining of cards and actions, potential outlines of your opponent's next turn emerge, and with them new threats that might influence the cards you play in the here and right now.
Here's the funny thing about this: I'm not sure if this really makes it all that suitable to solitaire play, at least not for a large group of gamers. Some folks slip into this "if I do X, she does Y, and then I'll do Z, oh wait, she did B instead" psych-out headspace with surprising (perhaps even alarming?) ease, and I'm one of them. Others however get locked into a sort of single-stream group-think ("I do X, now she does Y, now I do Z" without ever considering what delightful mayhem might result from the second "player" choosing "B"), or just find all those nested decisions to be frustrating instead of invigorating.
Table Battles is a good example: there are folks who play it solo just fine. There are folks who can't understand how anyone could ever begin to solo it - all that back-and-forth is exhausting. And then there are folks who only play it solo but do so badly, without any ability to surprise themselves, with no inclination toward other possibilities. This last group often arrives at the conclusion that the game or a scenario isn't balanced, despite the fact that, because they've never played it with another human being, they've never actually played the game at all.
Like I said before though, I'm not too worried about making a two-player game that "solos well". My primary goal in building an open hand CDG is to explore the decisions and player dynamics that that method allows for. If that makes it easier to solo, on a mechanical if not a psychological level, that's gravy.