I've designed a lot of different games for a lot of a different player counts on a lot of different subjects. At first glance, my wargames look nothing like my train games - to the point where at least one person didn't realize that the same guy had done them, and even suspected the Tom Russell who does train games as being some kind of pseudonym. But there are certain features that all my games have in common, such as my usual obsessions with tempo, momentum, the need to weaken your own position in order to advance it, feedback loops, and fragility.
One consequence of all this, of course, is that once someone starts winning, they generally continue to do so, and it becomes harder and harder, and (sooner than you'd like) impossible to reverse it. They'll never let go of the tempo and momentum. They might be weakening themselves as they continue to advance, but their lead is strong enough to mitigate that almost completely, whereas their opponents are constantly exposing pressure points for the leader to exploit. The trick, of course, is not to let them get the lead in the first place, but often you're only cognizant of that in retrospect: three moves ago I could have, and should have, stopped him, but it's too late now.
Someone recently said that I have "good taste for abstraction but too little interest in balance", and if balance is meant to encompass catch-up mechanisms and a self-correcting equilibrium, and to ensure that games are close up until the end, then, yes, I have almost no interest in that. My games have some kind of equilibrium - sometimes players are evenly, and symmetrically, matched, and sometimes one will have an advantage (slight or large-and-looming) - and it's the goal of the players to distort that equilibrium and to push things decisively in their favor. Once they do, the game can and often does become a steamroller.
So the question for me, with regards to the losing player or players, isn't "How can I help them out?", because after all, it's their own fault that they let this happen - they deserve it, he says, with a sadistic twinkle in his eye - but rather, "How long should I punish them?" One reason why my games are generally on the short side is that while I don't mind playing from a hopeless position, I do mind playing from a hopeless position for several hours. So most of my games top out at an hour or two. And when there's a higher luck factor involved - like in Table Battles, or my fairly atypical, ultra-chaotic dice-fest High Speed Hover Tank - I aim more for twenty or thirty minutes.
But if a player's position can be rendered hopeless, why not end the game there, at that moment, the very second that someone at last pushes past the stalemate and achieves that decisive edge? It's a fair question, and one I've been turning over in my head. The trick of course would be knowing where, precisely, to draw that line. As I said above, sometimes the tipping point is only visible after it's already well in the rear view mirror; it's hard to quantify when someone's victory has become inevitable. And sometimes what seems to be a tipping point, what appears to be a decisive advantage, isn't really and will evaporate with the right counter. And then, beyond that, the ability to end the game suddenly when you achieve some kind of apparently definitive advantage, in a game that's built to create such distortion, can lead to a lot of meta-gaming, and canny folks pushing hard for a condition that will win them the game now, but in a longer game, would prove to be untenable and easily reversible.
There are definitely pitfalls to such an approach, and a designer should tread carefully. But it's a question I've become increasingly fascinated by, especially as more and more of my designs become about two sides pushing, pushing, pushing at each other until someone blinks: why not end the game when they blink? Probably the game where this concept is the focus is Table Battles. In designing the first expansion, The Wars of the Roses, I experimented with decreasing the number of morale cubes afforded to each side. Most of the battles in the base game had five or six morale cubes in play, split two-three or two-four. The expansion has a couple of larger, longer battles that follow that mold, but many of the battles only have three or four cubes in a one-two or one-three split. There's precious little wiggle room, and a player who manages to push past the deadlock needs only push a little bit further to end the game in victory. It's death at first blink.
I'd be hesitant to cut an historical hex-and-counter game off at its knees like that, because then you lose the narrative, the full grand sweep of history; there's an experiential value to wargames that other, "purer" competitive games don't have to take into consideration. And, like I said, there are some games where the advantage might become clearly visible to the players, but is too nebulous a thing to quantify. I got around this with Supply Lines by advising gamers that if the game appears hopeless, it probably is, and they should just concede. I was a little worried at the time that some doomed player might insist on playing the whole thing out anyway, even though it was clear he had lost and could never recover, and spent considerable time trying to build in some mechanism by which one side could "force" him to concede. I came up empty-handed, but eventually came around to the idea that that was a silly thing to worry about; most players aren't that masochistic.