While keeping your powder dry and not giving anything away would you mind sharing which system you used to war game these scenarios?
Not at all, but it's one I came up with from first principles. Unfortunately therefore, I can't point you at anything you can look at.
The root of it is a plot of the action (as per any naval history book), to scale, gamed over 5 minute intervals.
Very important to keep one's thoughts separate, and record what's going on for each side, so I try to ensure that both sides act on what THEY see and want to do, not what the enemy (or yours truly) might want them to do. Obviously, there's scope for 'fiddle factor' there, but I try to assume that commanders act rationally based on their orders and what they know.
Even at a battle as vast and confused as Jutland, I can only think of two violations of that rule (and even then both are debatable) - Arbuthnot's cruiser charge and Evan-Thomas' not turning around.
When firing occurs, except in exceptional circumstances it's 1 RPGPM. Most navies fired faster than that, but they didn't often fire continuously at a high rate for more than a few minutes, so it roughly averages out, as I don't make minute-by-minute allowances for the effects of smoke, or for that idiot A/B Johnson jamming a gun etc...
Hits are random, but the probability is based on range, sea state, visibility, number of guns operational, rangefinder accuracy, general state of damage and factors to do with the movements of the ships (i.e. a ship that's just completed a major turn isn't going to be shooting very well for a few minutes). I manually (but still using a random number) try to include 'odd factors' such as the occasional mechanical failure/systematically poor training etc...
I never increase the number of hits manually, only reduce it.
The location of any hit is also random, plotted on a profile view of the ship in question, hence I can be quite specific.
What happens as a consequence is based on what the shell hits. If it's armour (and the result isn't obvious - e.g. a 15" shell hitting 4" armour), then the ballistics of the gun and the resulting velocity of impact comes into play, using the US Navy empirical formula, and a modifier to allow for different qualities of armour and shell (e.g. British shells in 1915 show worse results for off-normal hits than German ones).
There's a bit of 'fuzziness' around the question 'Does it penetrate?' as there are such things as partial penetrations (e.g. if the formula says the shell would penetrate 9.7" and it hit a 9" plate, I might say 'the plate broke, but kept the body of the shell out')
Even if it gets through, there's another random factor - does the shell explode? (and if it only just makes it through, the fuse/shell might be broken, so probably not).
However, even if it doesn't explode, it might still do some damage.
The effects of direct magazine hits or turret hits are fairly easy to model, but general flooding, equipment failures and damage to other areas is also noted, and fed back into the next round.
Obviously, a good deal of this is taken care of by a computer program, which also means there's less scope for me 'improving' on the randomness along the way.
Is it perfect - certainly not.
Does it produce an interesting and broadly believable result - well, you'll have to see what happens...