The last 4 sessions of Compromise and Conceit have been conducted using the basics of the system I have been developing, with the main properties of this system being:

  •  All actions are skill-based, with no difference between saving throws, skill checks, magic use and combat
  • All results are determined by a difference between the result of the roll and the target of the roll, and the maximum effect an action can have represents its power
  • hit points and available spells are determined by skills rather than by dice rolls
  • all actions are resolved using 2d10
The 4 sessions to date have given a chance to assess this system in practice, and I think it works. There are a few small flaws, but the main good points appear to be:
  • It is fast: we can run the introduction to the adventure, setting out, travel, a single large combat and a brief denouement in 3-3.5 hours, which I think is good
  • It is simple: there is only ever one die roll to determine the success of any action, and the mechanics of resolution being always the same means that I can assess the outcome of actions quite rapidly
  • The two properties of skill-based magic and wounds affecting combat/magic mean that the system does capture a feeling of desperation as people tire or get wounded, and magic is a powerful and frightening phenomenon
So far we have had 4 battles: 1 teaser with a witch and some zombies, one small conflict (inside a larger battle) with a couple of experienced soldiers and their minions; one ambush by a wizard/assassin and her minions; and one big pitched battle. All have been resolved rapidly, and although the first and third weren’t very challenging, the fourth was a very close call with the characters only being saved at the end by their Iroquois allies. That last battle was between the 5 characters (all 4th level) and 15 soldiers (1st level) and three leaders (all about 3rd level). The characters used some magic items, quite a bit of their more basic magic, and their stealth skills, and came out of it heavily fatigued and badly wounded. 
The main faults I have found so far are:
  • The difficulty classes for spells are a bit arbitrary and need to be tightened up. I think an absolute judgement on level and a level-based difficulty assignment may be a good idea
  • I am working with a system of DCs for spells in which failing to meet the DC on a skill check means that the spell still works but the character takes a fatigue. This makes magic perhaps a little too powerful. I am considering setting a lower DC for partial failure (which could also apply to other skills) equal to the DC minus the character’s level. So for example, if the DC is 15, partial failure occurs for a 4th level character on rolls of 11-15. This will make magic viable but will make people think twice about using their most difficult spells frequently.
All in all, if I can tighten up the magic system I think this system  will work well!