Since I note this forum has yet to discuss this game, I might as well come up with some ideas in the hopes of seeing a TAS of this game eventually. Even if it's done by me. (I'd be more likely to run this game first than MMBN2, the other one I hinted at. Although I'd probably wait until after doing a live-console run of this game, which I'm in the process of planning and which inherently has to use a more complex strategy.)
I'm just hoping the 'random dungeons' element of the game won't make emulator movies desynch like mad.
Now, some people are probably aware that a Japanese run of this game was accomplished. For those that don't, the details of that run are at
http://dq-x.com/dqmta.htm. Unfortunately, to my knowledge, there's no video of the accomplishment. (It was reported the total time of that run is 2:03.)
From what I was able to discern by putting the page through a Babelfish translation, their final party is KingLeo, Balzak, and MadGoose, and their Starry Night strategy seems to abuse the move LureDance. In fact, a LOT of their strategies rely on abusing moves that prevent the enemy from doing anything.
While their final strategy is admirable, I personally feel that it fails to be optimal. Part of the reason is that they rely so much on random tamers wandering in the gates. I'll take adjustments on this, but what I was thinking was a party something like Grizzly, Reaper, and DanceVegi.
Grizzly and Reaper have fairly fast attack growth (more importantly, they can be bred really early), and DanceVegi learns LureDance naturally as well. The advantage of DanceVegi over MadGoose is that DanceVegi can also learn PalsyAir, by having FireWeed as a parent monster. And it turns out two of the final boss's three monsters can be affected by PalsyAir (MetalKing's the only no-go, and there's other ways of taking IT out). In fact, it is the rare monster indeed that is immune to PalsyAir, and of the monsters before Starry Night, I believe none are immune to both PalsyAir and LureDance. With a little (forced) luck and the right approach, I'm convinced that we can ignore (or close to ignore) worrying about the resistances of the player's monsters and just stop the enemy before they can do anything.
One other thing I'd like to note quickly. Unlike some other RPG runs that manipulate critical hits and the like, I feel that forcing critical hits (which ignore defense in this game, a potentially more deadly outcome than simply double damage) constantly is suboptimal. Instead, I'd recommend the skill EvilSlash, which acts as an automatic critical (damage-wise - there's no critical hit message with this skill) if it hits, though I believe manipulating around the lowered defense (so attacks go after someone else) might come into play.
While I try to come up with a more detailed plan, and possibly modify the final monster set, I'll let others share their opinions here.
