I had a problem: my hoplite lines weren't breaking.
One of the core things I wanted to do in With It Or On It was to represent units - those individual square pieces of cardboard - as parts of a whole. I wanted players to think of the ten or twelve counters that made up a wing as a single "piece". When an individual counter was flipped to its exhausted side as the result of combat, it was like a little stress fracture. As these pile up, it became more and more likely that the line would break.
When it did break, the game's rout check system increased the chances of that break being catastrophic. For the uninitiated, the way it works is that once a unit is Eliminated, you take a quick look at all the little cardboard squares. If it's adjacent to two other units (or adjacent to a unit that satisfies that condition), it passes the check. Otherwise, it, too, is Eliminated - so if your opponent pokes enough holes in the line, or pokes them in the right place, the whole thing will collapse.
It was this decisive moment that I was chasing after. It was this moment that captured the essence of hoplite warfare as I saw it: the stubborn push of shields and spears, the small cracks slowly spreading through the line, the mounting pressure and tension, and then everything unravels suddenly and dramatically. It was this moment that never quite happened.
Oh, eventually the lines would collapse, and the units would rout, but it took so long to get there that it felt less like a dramatic reversal of fortune - like something that might happen at any moment - and more like something that was inevitable - something that would happen, eventually. It's like the difference between a car going to the junkyard because it was in a terrible crash, and going because it had slowly and inexorably run itself down.
The root of the problem was simple enough to diagnose. Whenever a unit took a hit, the owner had a choice: either flip that unit to its exhausted (and useless) side, or flip a fresh unit that was adjacent to it. An exhausted unit that suffered another exhausting result was eliminated instead, prompting the rout checks, but this could only occur when there were no fresh units to resolve the hits. This made sense thematically but mechanically, in practice, it meant that the rout checks - the decisive moment, the whole point of the game - only really happened after a long slog, feeling deterministic. The only time it happened any earlier is when someone rolled a direct Elimination result.
Those were rare, and, of necessity, they had to be - otherwise the model wouldn't work. And I didn't really want to change the ability of adjacent units to absorb hits for the target, because that too was crucial to the model - it showed the importance of the line as a whole, and how each part of the formation works to support it. I felt very strongly that I was on the right track, but that it tended too much toward stability. The game needed a destabilizing element (most games do).
While I'm letting my brain pick apart this problem, I'm also working on the counters for the game. To represent the exhausted side of each unit, I use the handy-dandy universal symbol for the reverse side of a wargame counter, ye olde horizontal stripe. But I don't know where I want to position that stripe, and so I experiment with two possible locations: one directly across the center of the counter, and one across the lower third. At one point in this process, I happened to make both layers visible, and suddenly I'm looking at a counter with two stripes.
And in that moment was born the idea of brittle units, of guys who break a little more easily. When their counter is flipped, and the double stripes revealed, they are eliminated immediately, triggering those rout checks. A player knows how many such units they have in the line - they have a general sense of the discipline and morale of the troops under their command - but they don't know where they're positioned - they don't know when the line might break or why. This was the destabilizing element that the game needed, this is what made the model work, this is what created those moments the whole thing had been chasing after. And all because I happened to make both layers visible while tooling around with the counters.
Sometimes the answer is the result of hard work - of puzzling something out. Sometimes, it's about time - about letting your subconscious do the work for you in the background. And sometimes, the answer just falls into your lap completely by accident.