The Clontarf scenario in Kingdom of Dyflin was in some ways a departure for the Shields & Swords II series, as it was significantly bigger than what had come before: more hexes, more units, longer playing time. That was also something like the point of the thing - it's "the big one", in the same way that another scenario might be "the one with the rearguard" or "the one with the ambush".
I'm a firm believer in the maxim there is no such thing as an ordinary scenario, and that the worst thing a designer can do is come up with an order of battle and some terrain features and then call it a day. In some way, the scenario - regardless of system - should tell the story of the thing. This might be something obvious and showy like a special rule, or something more subtle, but it's gotta have something: it's gotta be "the one with the whatever".
With Clontarf, the whatever in question was that it's big and long, a real epic slobberknocker, an absolute meatgrinder. And I knew that to deliver on that, to tell that story, I needed something that felt big.
So, the unit and hex scale for these games is deliberately fungible, which springs from the fact that we're not really going to have reliable actual numbers from primary sources. What I aim to do instead is represent relative numbers - the ratio of forces between the two sides. I tend to represent this at a scale that allows for about ten or fifteen units per Wing, with two or three Wings per side (depending on the situation). This generally gives me between forty and sixty total units on the map, which works well with the 440 hexes on our standard 17" x 22" mapsheet - enough guys to make it interesting without making it too cluttered. Because the unit scale is fungible, I can zoom in or out on the battle as needed to give me those forty to sixty units.
But I couldn't really take Clontarf and "scale it down" to say sixty total units because then it wouldn't feel any different than any other battle in the series - it wouldn't communicate the essential bigness. And I couldn't just put more units on my standard hex grid, because then it would feel cramped. Cramped isn't big, cramped is uncomfortable: the thing needed room to sprawl a bit.
Of course if I was doing this as a brand new standalone game, the solution would be to go to a 22" x 34" map that could accommodate a greater number of units. But since this was an expansion to The Great Heathen Army, I was restricted to the counters and maps that came with that game. All three maps were 17" x 22", but if we put two of those together, we'd have one long 17" x 44" area with 752 playable hexes. I knew 440 hexes could comfortably accommodate 60 units - a ratio of about 7.3 hexes to each unit. Maintaining that same ratio for the enlarged playing field would allow for 103 units. I ended up with 114, which is 6.5 hexes to each unit, but again, that's a pretty good roominess ratio.
Now that I knew about how big my canvas was and about how many units I was going to paint with, it was time to actually figure out my order of battle. Now, this is something I've done a couple dozen times before with these games, and I'd have to say that Clontarf was by far the most difficult. It wasn't that I couldn't find detailed sources - because none of these battles have detailed sources. And it wasn't that I had trouble figuring out how many units I wanted or what type of units they should be. The trouble was that I was working under an additional restriction, namely that it had to conform to The Great Heathen Army countermix.
Leaving out the Command and status markers, there were 148 unit counters to draw from, so it's not that I didn't have sufficient numbers. It's that the GHA countermix was designed specifically to fit the eight battles in that set, without any thought toward a future expansion, let alone one where I would need over a hundred units for a single battle. While I didn't have much trouble at all with the other three scenarios in Kingdom of Dyflin, with Clontarf I found myself frequently butting up against the vagaries of the exact, peculiar, and not-at-all symmetrical mixes for each Wing. I wish I could say that I came up with some brilliant epiphany, or some approach to the problem that made it all click into place, but really it was just plain hard, stubborn work: doing the thing over and over again, crumpling up sheets of paper, constantly struggling against the straitjacket I had accidentally made for myself, until a few weeks later I had something that worked.
Is it the exact mix I would have come up with if I was doing the thing from scratch as a standalone game? No, of course not. I would have made things much easier on myself! The Viking deployment in particular is a bit less granular than I would have liked in an ideal world. But does it tell the story of the battle and represent probable disparities in troop quality? Yep.
The next step naturally is to figure out the VP thresholds for each side. This is done by first determining the maximum VP available to each side - that is, adding up the VP values of all the opposing units. For example, in the Ashdown scenario in Great Heathen Army, there are 35 VP worth of Viking Units, and 40 VP worth of Saxons. The Viking Player's threshold starts at 16 VP ("16 VP, and 5 more than the Saxon Player") - 40% of the total VPs available. The Saxon Player's threshold starts at 17 VP - about 50% of what's available - and a secondary threshold kicks in at 21 VP, or 60%. Generally I tend to give the historical victor a higher threshold, especially if the scenario is giving them certain advantages (either through force structure or special rules) over their opponent. For example, in that Ashdown scenario, the Saxon Player can actually score double VPs in the early turns of the game, so it makes sense to give them a higher threshold.
Note that this isn't necessarily representative of the historical casualties, because the greater number of these casualties were likely inflicted after the losing side has broken and started to run (war is, after all, a rather grisly business). What I'm aiming to model here in most cases is the breaking point - everything up to the moment when the battle gives way to slaughter. With Clontarf, I didn't necessarily want to end it at that moment, but to let it linger, and to give a player who is on the ropes the chance to claw their way back to victory.
Now, as it turns out, in the Clontarf battle the 61 units on the Viking side and the 53 units on the Irish side are each worth 92 VP. The battle was a decisive victory for the Irish forces, but it was won with extremely heavy casualties, and I don't think that result was a foregone conclusion - it really could have gone either way based on the scant information we have. So it seemed reasonable given my parameters to set the VP threshold for both sides somewhere around 50%, and I settled on "45 VP, and 5 more than the opponent".
But while that ratio is consistent with the numbers I use for other VP thresholds, it is of course much higher than usual. Because most of the other battles in the series have between forty and sixty units in a battle, there's a general range of available VPs, and as a result most VP thresholds fall somewhere between 15 and 30 VP. 45 VPs naturally makes for a longer game, and this might be exacerbated by the high ratio of levy units (worth 1 VP) to veterans (worth 2 to 4 VP) in Clontarf. Yes, I wanted a longer battle, but I didn't want something that overstayed its welcome - the thing still needed to move.
There were three ways that I managed the pace of the scenario. First, the use of the Advanced Wing Integrity Rule that Dyflin introduces to the series meant that those enemy wings crumbled a lot faster - pushing against the right points in the line would see several units eliminate themselves on the following turn.
Secondly, I let the players do more on each turn. In most scenarios in the series, you get to issue two Commands to one Wing each turn. Very rarely will you get to issue Commands to two Wings, and in those scenarios, you generally only have one Combat Command available, which means that only one of those two Wings is going to be attacking. But for Clontarf, each player activates two Wings, and has two Combat Commands - greatly accelerating the mayhem.
Finally, I glommed onto a bit from my favorite medieval source, the Cogadh Gaedhel re Gallaibh, which made note of "terrible, piercing, fatal, murderous, poisoned arrows". Initially I had a special rule which simply gave a +1 DRM to Firing Units, increasing the chances of Suppressing the enemy. But that didn't really feel particularly piercing, fatal, or murderous to me, not to mention terrible or poisoned. What I settled on was skipping Suppression entirely, and having successful Fire inflict step-losses directly on the target. This made the Fire Command a lot more viable throughout the battle, and along with the two elements mentioned above kept the pace from flagging despite the battle's longer duration.
One might think that doing a scenario whose whole point was "it's bigger" would simply be a matter of blowing it up like a photograph. And, I mean, I did do that, but I also had to compensate for the distortions this introduced, so that it would retain the essential character and appeal of the system.