RANDOMIZED UNIT POOLS (by Amabel Holland)

In a few of my battle games, you'll receive a certain number of units at set-up of varying quality, drawn from a randomized pool. The first game I did this in is Battles on the Ice. When I want to sound like I know what I'm doing, I'll tell people that the reason for this is that I wanted to model the unreliability of feudal levies, which was true. If I want to irritate the kind of grog who spends a lot of time on forums rattling on and on about accuracy, I'll tell them that the reason is because battles of the period don't really give us a reliable order of battle, so this method isn't any more or less accurate than trying to guess which unit should get this rating and which should get that one. And that, also, is true.

And if I want to be utterly candid about it, I'll tell you that it had a lot to do with the fact that a half-sheet of counters has eighty-eight squares on it, and a full-sheet has a hundred seventy-six, and having determined that between the two battles in the box I would need more than eighty-eight but less than a hundred seventy-six, I padded out the mix with extra units.

All these reasons are true, and I liked their intersection well enough that I used this trick again in The Great Heathen Army (and its expansion, Kingdom of Dyflin), and made it a central part of the Shields & Swords Ancients games (2019's With It Or On It and this year's upcoming The Grass Crown). In the medieval games, this mechanism was mostly restricted just to the lower-quality levy units, and would determine exactly how low that quality was, with units usually beginning with a "B" or "C" combat class, and being flipped to a "B", "C", or "D" after suffering a step loss.

By contrast, the ancients games apply this mechanism to every unit type. Generally, every unit within a type has the same combat class: they're all identical until they suffer a hit. When that happens, the unit is flipped, revealing either one stripe or two; if they have two stripes, the unit is brittle and routs immediately. If this opens up a hole in the line in the right place, other units might rout as well, causing your line to rapidly collapse under pressure. It's a key part of the model, and one that only works because of this randomization – one that, in fact, is built upon that randomization. If I hadn't had extra counters on the Battles on the Ice sheet, I don't know if I ever would have quite figured out how to shift the Shields & Swords chassis to the battles of antiquity.

Over the last several months, I've been toying with a rather different iteration of this idea, one that would be applied to a more operational rather than a set-piece battle framework – or, to be more accurate, an operational framework that also depends on the outcome of set-piece battles. This is the central idea between two projects that I've been working on, Oblique and Siege of Mantua.

Oblique is a game that started as Supply Lines of Frederick the Great, and would be an evolution of the logistics-focused operations in my games of the American Revolution. There are a number of differences, and one of the major ones is that while it behooved Washington and Greene to refuse a decisive battle via Fabian tactics, Freddy and friends were more likely to seek one. This meant that the simple "trade cubes for dice, roll 'em, score hits" combat model of the two Supply Lines games really wouldn't work. Really, I want something that emphasizes the required mastery of both the operational art and the art of battle, and the way the two are inextricably linked.

And while some bifurcation is necessary – i.e., you do the cat-and-mouse chase to bring the enemy to battle, and then you pause the operations so that you can fight that battle, perhaps on some kind of tactical display – I don't want the battles to be overly granular or fussy. I've heard the complaints that folks have about the left-center-flank cards in early CDGs like We the People and Hannibal; heck, I've heard the complaints of folks who felt that the set-piece battles in my Agricola and Charlemagne stopped those games in their tracks. To the point where some folks would rather not fight a battle so they don't have to go through the hassle of resolving it. Obviously, I want to avoid that!

I also want those battles to have a distinctly Frederickian character – there's a reason I'm calling it Oblique. All that is a lot to grapple with, and besides that, it's shaping up to be a rather more ambitious game than its predecessors, so I've been trying to cast about for a way to tackle some of these essential problems – the marriage of a logistically-driven operational maneuver game with a quick but meaningful battle system – in a smaller and less ambitious format.

Which brings us to Siege of Mantua. This would also serve as my latest attempt to tackle the problem of "gosh we have a lot of blue and red blocks because we had to buy them in bulk when we did Ribbit, guess I better do a block game, but one that doesn't have too many blocks because we don't have that much room in our boxes". This is a wall I've banged my pretty little head against a few times now without much result. I knew that whenever (and ifever) I did do a block game, that it would have to depend on misdirection and bluff – what else are you gonna do with a block game? – but that it would not have that thing where you tip over the unit to reveal it and then roll a number of dice equal to its steps and score hits if you roll high or low enough. That seems to be de rigueur for block games, and I'm not saying there's anything wrong with that, just that it's not ground I'm looking to retread.

At the same time, the appeal of the blocks over hidden movement counters, besides tactile pleasure, is that it lets you secretly represent up to four steps of strength. So I'd want to do something that made use of that – those steps would need to come into play somehow, and in a way that wasn't overly deterministic and chess-like.

Like I said, I've been banging my head against this problem for a while, and then I decided to bang my head against the problems posed by Oblique, and then I realized that, hey, maybe some of these problems are the same problem. Maybe I could do a small block game about that intertwining of operational maneuver and battlefield tactics, and the various failed Austrian attempts to relieve Bonaparte's siege of Mantua in Italy seemed an ideal subject.

And you're probably saying, "Gee, Amabel, this sounds great, but what on earth does all this have to do with the randomized unit mechanism from those Shields & Swords battle games? Did you accidentally splice two blog-things together?" And the answers are, in reverse order, "no", and "I'll tell you."

It occurred to me that the obvious solution to my "what do the steps signify" problem would be for the number of steps a block possessed to correspond to the number of counter-based units it could bring to bear during a battle. Plonking down three infantry divisions on a battlefield feels a lot more impressive than rolling three dice! If I roll three dice in your standard-issue block game and you roll one, you're not going to be sweating it too much, because there's a chance that I'll botch all my rolls while yours will succeed. Whereas if I have three times as many units on a Napoleonic battlefield, provided I've put them in the right place and move them at the right time, there's a serious chance I might turn your flanks, encircling your forces and destroying them.

The problem posed by this model, however, is that all my units are essentially fungible, which tends to make the thing more abstract and chess-like. I didn't want to have specific counters tied to specific blocks, tracked off-map in some fashion – it just seemed like too much folderol for what I'd get out of it. (And besides, display sheets add to the production budget!)

Which is where, finally, my old friend "randomized units pulled from a pool" comes into the picture. Let's say I'm playing the French and I have three blocks with eight pips between them coming into battle; well, I put all the French units into a cup, shuffle them around, and pull out eight. They're going to range in quality from poor to good to great, but I won't know the exact mix until it's time to commit to battle.

This isn't representing uncertain troop quality in terms of the average value of each unit, but rather its performance, morale, et cetera in a given engagement. I'm gonna know what's going into the cup, and therefore I'll have some idea what's gonna come out of it, but I won't be precisely sure until the moment of truth.

That pool however isn't going to be the same every time. If I win a battle, I'm going to remove some of the lower-quality units from the mix, replacing them with higher-quality units. And if I lose, the overall quality of my unit pool is likewise going to degrade. The extent of one's victory or defeat will naturally determine how dramatic these changes are.

I think this is a simple, effective, and clean way to model shifts in morale, the transformation of raw recruits into hardened soldiers, and the importance of momentum – the more you win, the more likely you are to win, and the more you lose, the more likely you are to lose - without needing to keep track of modifiers and such. And while the intended operational models of these two designs are starkly different, I think if the core battle mechanics (including the randomized unit pools) in Siege of Mantua prove effective, I'll be able to build on that work when I return to Oblique.