I played through another game of Dertosa tonight using the same armies, deployments and plans as the first play test, but with the results this time proving to be far closer to the historical outcome. The Romans again pressed forward early to take the Carthaginian key zone, but this time the levy troops opposing them could not counter-attack quite so effectively. The infantry battles were again attritional, with Rome having the better of the exchanges in the centre and the centre left, but with Carthage performing strongly in the centre right.
The cavalry battle on Hasdrubal's flank was decided in the fifth turn as Gnaeus Scipio died in a rally attempt, but the centre abruptly gave way and carried the Numidian horse, the elephants and three of the (spent) average heavy infantry units with it before Hasdrubal could gain much advantage out of the cavalry's success.
At this point it became a grim struggle between the opposed right and left central zones. The Roman right was under terrible pressure from the veteran infantry to their front and the cavalry to their side and rear - so much so that they lost three units shattered in two turns. It was a similar story for the Carthaginian right, also under pressure to front, side and rear. For a time the Carthaginians kept rolling high enough on morale tests to stay in the game, but in turn 9 two veteran units were shattered on a double hit and the rest of the army beat a hasty retreat.
The final tallies were as follows:
From shattered: 2 x VHI, 1 x AHI, 1 x LHI for 26 points.
From routed: 2 x AEL, 1 x LLC, 1 x LLI, 1 x ALI, 3 x AHI, 3 x AHC, 2 x VHI for 68 points.
From generals lost: AC for 6 points.
Total Roman VPs came to 100 exactly.
From shattered: 2 x AHC, 3 x ALE, 1 x ALI for 36 points.
From units spent: 5 x ALE, 1 x ALI for 18 points.
From generals killed: UC for 6 points.
Handicap: 31 points.
Total Carthaginian VPs came to 91 points.
Result: a narrow victory to Rome.
This was much closer to how I think Patrick's article envisaged the battle as having gone. If Rome had not lost Gnaeus Scipio and had had two or three fewer units shattered, I think that would have been about perfect.
I still have a few doubts about the scenario, especially in regards to the deployments. The levies are still fleeing late and are needing to have given up their key zone to ensure the morale test failure. I wonder if they should perhaps be classed as light infantry for morale purposes, which would see them begin the game with a morale of -1? Anyway, this test shows that Lost Battles can come pretty close to replicating the historical result without requiring too much 'fiddling of the books', and without too much in the way of wild dice luck this game, either.
All in all, a satisfactory result.