The battle system for my block game Siege of Mantua wasn't quite where I wanted it. Don't get me wrong. The basic construction was very sound. The thing worked, and worked well. But it sometimes felt just a little loose – in the way that a shirt can be just a little too big.
The short version is that on your go, you spend Commands to advance and attack with your troops. Said troops are randomly pulled from a pool of counters to reflect the uncertain morale situation in a given battle; that pool shifts over time, morale degrading when you take casualties and morale improving when you win battles. A side can end the engagement early to stave off mass casualties, but only after a certain number of turns – determined by a 2d6 roll – has elapsed.
The 2d6 bell curve being what it is, most battles last a minimum of six to eight turns. Some battles could be much shorter of course, and some much longer. That's a feature, not a bug, but also contributed to that feeling of looseness I mentioned up top. The shorter rolls meant battles were over before they started. The longer ones tended to run a turn or two too long, which sapped the tension. I realized that I needed some way to tighten the thing, but I wasn't sure how to go about it.
Meanwhile, I started to have doubts about the unit pools. Not about the central concept, or even the way those pools progressed. But because the campaign saw Bonaparte's French army fighting against both an Austrian and a Tirolean army, I had three pools that needed to be drawn from and managed and that felt a little fiddly. If this was a three player game, that'd be fine, but it's not; the Austrian and Tirolean armies are controlled by a single player. It was also the only reason why the game had a full sheet of counters instead of a half-sheet, and that grated against my sense of thrift.
The obvious solution there was to sacrifice a little bit of simulation value, having the Austrians and Tiroleans draw and make adjustments to a common pool. This was a lot smoother and now the number of counters fit on a half-sheet.
Well, almost. I was now exactly three counters over the confines of a half-sheet. But being that each player had a Command counter, and that there was a shared counter for tracking the number of battle turns, I could very easily swap these out for fifteen millimeter wooden discs. After all, we're already putting ten red and ten blue blocks in a bag, might as well slip in three discs while we're at it.
But this of course meant changes to the Command and Turn Tracks. A track built for a half-inch rectangular counter has each box smack-dab up against the other. Twelve boxes on the turn track corresponds to six inches of real estate on the battle display. As this will be printed on a US letter-sized (8.5" x 11") sheet of canvas, that fit just fine. But a fifteen millimeter disc is nearly a tenth of an inch larger. And from an aesthetic point of view, it doesn't do to stack circle spaces against each other the way you would with rectangular spaces. You really need to put a teensy bit of distance between them, give them space to breathe.
But twelve circular spaces, with gaps between them, would stretch out nearly ten inches – an inch and a half too long along the shorter 8.5" edge reserved for it. I couldn't run it along the taller 11" edge; that would unduly constrict the battlefield where, after all, most of the action was going to be taking place.
"Well, I guess I have to make the turn track smaller," and that in turn provided the solution to my original problem. My new track would run nine spaces only:
As you can see, the "2" space is followed by the "4" is followed by the "6". When you make that initial "minimum number of battle turns" roll, you place the marker on the highest space you can without going over. For example, a roll of a "3" plops the marker down on the "2" space. This is actually the third space of the track, so even a roll of snake eyes will guarantee a minimum of three turns. An average roll will see a minimum of five or six turns. Larger rolls result in a greater number of minimum turns, but since this caps out at nine with a statistically unlikely roll of boxcars, it's not overstaying its welcome, and those "longer" nine turn battles aren't happening as frequently.
This version of the track tightens the possibilities while leaving room for the desired variation. And, you know, I only did further testing to the degree necessary to confirm that the nine space track did what I wanted it to. I'm sure some other designer would then proceed to do tests with an eight space track, and a ten space track, and do versions of each where the numbers have different values, and then collate the data to determine which was the best approach and, you know, more power to 'em.
But this was a problem that was solved more-or-less accidentally: the new track has nine spaces because that's all I had room for if I was gonna use a fifteen millimeter disc, and I switched to a disc because I was three counters over the limit of a half-sheet. And I've learned over the years to trust these sorts of things rather than second-guessing them or picking them apart. My design process has always been powered more by instinct than intellect, and one of the advantages of working in a medium that results in physical objects with physical limitations is that sometimes the medium makes the decisions for you, if you're willing to listen to it.