A common complaint about a certain type of wargame is that attacks can be much more frequent (and sometimes much more decisive) than they were historically. Especially in games where factor-counting figures into it, there's a tendency to pile as many counters up near the front-line as possible, and to attack every turn with everything you've got provided the attack isn't a stupid or pointless one. As a result of this, many wargames are "bloodier" than the real thing - if the removal or flipping of a small square of cardboard can truly by said to be "bloody" - with the last few turns decided by a few survivors who gaze uneasily at the unceremonious dead pile just to the side of the map.
There are different ways for a designer to limit the amount of activity a player's forces can engage in over the course of a single turn. You can tie movement and combat to a finite number of resources - supply cubes, operation points, card plays. You can put the emphasis on command-control, such as limiting activations to the units belonging to a given formation (which I'm usually okay with) or by using a count-the-hexes-from-the-leader-or-HQ command span system (which I'm usually not).
All of these methods have their uses and can limit activity to something that's appropriate for the period or the topic. But at the same time, all it's doing is changing the limit from "as many counters as you can fit in these hexes" to "as many counters as the game's system allows you to activate or attack with". The players are still throwing everything they possibly can at each other, unless there's a good reason not to. And really, shouldn't it be the other way around?
No general worth his salt ever attacked the enemy unless it was in pursuit of a concrete objective. That objective might be as simple and as bloody as destroying the enemy army. It might be seizing a vital supply point so as to starve them into submission. It might even be as simple and as frustrating as "the politicians back home will sack me if I don't win a battle". Whatever the impetus for the attack, a good general doesn't launch it if he doesn't think he can win it; he'll also balance the gains against the risk that the attack poses to his own army.
This risk can be paralyzing in the wrong hands. George McClellan gets a bad rap for his reluctance to attack a numerically-inferior army that he imagined to be much larger than his own, and for his miserly commitment of forces, whether it be to support attacks or prop up his own defenses. And, look, I'm far from an apologist for Little Mac, whose many and egregious faults far outweigh his ability to organize and train the Army of the Potomac. He proved time and again that he was unable or unwilling to act. Of course, if he had acted, and had lost his whole army as the result, he wouldn't be any better regarded then or now.
A commander has to make decisions under stress and with limited information, and a good commander has to make the right decision more often than the wrong one, has to take the right risks at the right time, and which decision is right or wrong is only really apparent in hindsight. People rightly think Varro was an arrogant moron for falling into Hannibal's trap at Cannae, but if Hannibal hadn't pulled off that double-envelopment, we'd be taking instead about Hannibal's arrogance. It was folly for Boudicca to accept a pitched battle against the Roman Legions, but it wouldn't have been if she had triumphed. And so on and so forth.
I think a game that can replicate that kind of stress is going to be more interesting than one that doesn't. Because in such a game, it's not the designer putting constraints on the player via command spans or formation activations or command points. Rather, it's the player putting a constraint on himself, because he's seriously looking at whether or not the attack is worth the risk, and only attacking when he thinks it is. Every battle has to carry with it a serious risk, but it doesn't need to necessarily be the risk of total annihilation, which can get awfully swingy. It can be something subtler than that.
I was thinking about this recently while playing the Union side in John Theissen's Objective Shreveport! Historically, the slow-moving Union forces, hampered by cumbersome supply lines and bickering leaders, were unable to seize the titular objective. I was determined to do better than my historical counterparts. All I needed to do was preserve a force large enough to push the rebs out of the heavily-fortified city of Shreveport, while hammering the enemy elsewhere to prevent them from adding their Strength Points to Shreveport's defense.
As with the same designer's other operational ACW game More Aggressive Attitudes, battles in Objective Shreveport! are rarely decisive. I'm not going to lose my whole army in one go, and neither will the other guy. Because of the way casualties work - being a function of the size of the smaller army in the combat - probably we're going to both lose one or two steps. In a severe case, one of us might lose a Leader, which will leave an army headless and unable to press the advantage, but that's a once-in-a-blue-moon kind of thing. Also complicating things is that the defender can retreat before combat by rolling a modified three or better on a six-sider, which is relatively easy to do. You would think that this would be a recipe for me to be fighting every turn, as the risks seem to be fairly minimal.
But in point of fact, I frequently found myself holding back. Because if I attack now, I might have less SPs to bring to bear later when I need to make my big attack on Shreveport. Or maybe I get Disrupted, which leaves me supremely vulnerable to a counterattack. It's not worth the risk to attack this turn. Next turn, though, I'll move this other guy into the stack, give me 3 more SPs. Next turn, I'll attack.
Of course, when next turn came and I did get enough SPs in the stack to tilt the odds in my favor, the enemy retreated, and why wouldn't he? Why give battle unless it's absolutely necessary to do so? It certainly doesn't help his objectives to let me clobber him. And so we circle around each other, looking for an advantage, and trying to come to grips with one another at a time and a place of our choosing. As a result, the battles that do take place usually happen when the odds are about even, and the results indecisive. Neither of us have anything to show for it except a subtle decrease in SPs. And even though that loss represents a larger percentage of his forces than of mine, because I have more guys, and because I need more guys to punch through Shreveport, the losses actually hurt me more, because each time it diminished my chances of achieving my actual goal.
It is, I think, a rather subtle and lovely way to create that tension, to force a player to hesitate and to second-guess, to put off the attack for just one more turn that may or may not ever come.