Scaled Continent: Lessons Learned

• There is such a thing as too many players at the table, especially when they all have henchmen and you can't see each other.  ACKS with DM+3 is a very different game from DM+5.  At DM+3, we saw cohesion and lack of player capability overlap.  At DM+5, we saw disintegration of cohesion (see: Tim's assassin forays away from rest of group) and massive capability overlap (contention for front-rank slots by many fighters).  Grim Fist was DM+4, and that worked out pretty well.  Alex Macris' games seem to run at DM+6, but he built the system, so one would expect him to DM it optimally.
• Playing over internet without video seriously aggravated this issue.
• Fighter-centric game problems might eventually be solved by ACKS Heroic Fantasy book?  Or not adventuring in narrow corridors.  It's interesting that our group seems to really like fighters; I think that of eight players, four either were playing a melee fighter or had a melee fighter henchman (or both in two cases), while only two had dedicated ranged fighter or ranged explorer henchmen.  Dungeon-biased party composition.
• Play to the things that motivate your players (duh?).  Treasure and exploration for their own sake, while romantic, don't appear to cut it with my player-base.  Money and exploration are both means, and there was a lack of interesting ends.  I have a slight aversion to unifying distant existential threat, because of a past experience where one player hyper-focused on it and set the group's agenda, but avoiding it entirely was probably not the right solution.
• I believe this also contributed to lack of players taking scheduling initiative, which meant the irregular scheduling mode of operation broke down and the game became regular.  Timezones basically precluded weeknights anyway, though.
• Lacking elements that I believe my players enjoy:
• NPCs with agendas.  For all that we complain about the Elminster effect and how ACKS isn't like that, having a handful of high-power (possibly malicious) NPCs around is actually good in ACKS.
• Coherent narrative; partly impeded by total playerbase size, partly by my intended style of game.
• In platoon-scale Domains at War, lizardmen can crush something like three times their number of human mercenaries as a result of HD advantage (2d8+ vs 1d6) and racial +Morale, and lightning bolt is probably less good than fireball.  Sorry about that, Mark.
• I may have made the barrier between civilization (Voltager) and wilderness (Continent) too big - 500 miles of ocean is "daunting".  Conversation with dad suggests that this may have also contributed to lack of scheduling initiative.
• Trade under the standard rules still isn't quite doing it for us.  Probably needed bigger markets.
• One player's feedback suggests that ready availablility of RL&L for purchase made death inconvenient rather than terrible, and hence the game was less exciting.  The primary consequence of dying was probably that you were basically done for the session because you needed weeks of bed rest; the side effects folks rolled were mostly nuisances rather than serious handicaps (especially when RL&Ling the giant hawk henchbeast).  Perversely, this problem is best solved by lowering market classes.
• Treasure maps pointing to lairs rather than full dungeons actually worked pretty well.  The only trouble is that a lair with as much treasure as the treasure map roll yielded almost certainly meant dragon, cyclops, or something else rather formidable.
• Hex stocking wasn't a big problem, but I did fail to use dynamic lairs again.  Random wilderness encounters were reasonably random and respected by players without dominating the game.  Nobody tried hex-clearing (yay?).
• Again, world needed to be more active.  This was coming, with monsoon season and political upheaval, but not to the once-a-session-at-the-pizza-break levels that I had planned.
• DM / genre fatigue is a thing, and timing when starting a new game is important.  If you take too long between when you have the idea and when the game starts, something else will start to look appealing.  In this case I waited a couple weeks trying to find a local group before falling back to internet.  This was a mistake, I guess?  Holiday unavailabilities did not help matters at all either.