Roll Eyes
Lips Sealed
1 page
List results:
Search options:
Use \ before commas in usernames
Game Page: Doesn't exist yet

Ultima: Quest of the Avatar (Any %) (Single Segment)

Verifier Responses

A/V Quality: Good as you're going to get from an NES.  Some weird stuff goes on with certain floor textures, but based on what I'm also seeing on my NES and this game, that's just a limit of the NES's composite output, not caused by anything on the runner's end.

Cheating: None detected.

Timing: Control is taken on the world map at 1:04 on the supplied video.  The final button press (the adventure is at an end [ARROW]) is at 1:59:52.  Total: 1:58:48.

This is a long game because there's a lot of plot to it.  By design even, as this was Richard Garriott's answer to complaints that his games were so darkly about dungeons, demons, and killing.  As the runner points out: 8 stones, 8 runes, 8 virtues, 8 partial avatarhoods, 3 keys, 5 artifacts.  Plus 4 experience levels, and also companions and equipment to get gold and to fight through battles as quickly as possible.

So much of the time spent in this run is spent just traveling from point a to point b, which is done very well, as most towns and dungeons are visited only once, in a way that minimizes waiting around for moon gates.  It's very impressive in particular when the runner gets to Moonglow when he does for the black stone, seeing the moons align as he needs rapidly.  So the routing has been done with care in a way that minimizes backtracking.

Also important is the time spent is combat.  The runner must gain Valor in order to win the game, and that means battling.  So equipment is bought, and then bought again in order to make it go as fast as possible.  The runner chooses not to go for the Wand in the end, reducing the gold grinding needed. Possible battle savings there, as the runner notes, but Bow + Tremor + Energy does a lot as it is.  Things do die quickly.  There's just a lot to kill.

The virtue system is obviously important, but the runner pushes it to the max.  Knowing the time outs, and importantly knowing what doesn't have a time out, lets virtue be gained very quickly.  This also lets other savings be had, such as taking the 99+% discount from the blind shopkeeper.  I know this isn't by chance, either: I watched Feasel streaming his tool-assisted test runs, monitoring the memory locations where the virtues are stored, and seeing exactly when and by how much the virtues are affected by his actions.

Apart from minor combat luck differences, I believe beating this run would require a glitch, or a new glitch if you count as a glitch 'Thou art a wimp' guy's easy virtue gain.  This is a well-optimized run that shows a high level of skill in planning and in execution.

Accept this thing.

Timing: see above verifier. I guess the timing doesn't start from character creation but only with gaining control over the PC as usual?
AV is good.
No cheating detected.

The runner has planned and executes the run with precision and care. It is clear from his run comments - this quote is about traversing the monotonous dungeons:
Walking backwards in some parts can cut down the number of 90-degree turns. ... You can skip the ladder text upon entering a new floor simply by holding a direction on the d-pad during the floor transition, so for each ladder in the game you must memorize which direction to hold when taking it.
Other than this, the way the runner retrospectively lists a plethora of potential sources of improvement further creates a sense of confidence in the informedness and in-depth consideration that were present in the planning of this speed run.

The main character, the Avatar of Profit, seems to have an extremely calculating approach to achieving his avatarhood. He knows it's okay to scam blind merchants massively when you can later just talk to a campfire (don't ask) for 15 times in a row to maximize the virtue points for honesty. I'm picturing those virtue points creating a huge bulge in his sclerotically fat wallet.

There is really very little by way of mistakes or sloppiness in the execution of this run. The runner sometimes "over-kills" monsters by selecting more attack commands than were probably going to be required. It is faster (albeit by a mere fraction of a second) to select "auto". Perhaps it was for safety - nothing more dangerous than a wounded slime. Yeah I'm scrounging for more things to say.


AV quality looks great for an NES game. No cheating detected.

The route is well planned, combat is well executed, dungeon crawling is fast and efficient. It's especially clear how much planning went into the dungeon crawling, because his route is planned out to the turn, not just which hallways etc.

But what I think is even more impressive than his dungeon crawling is the town planning. There's just so much to do all over this enormous map, and I can't think of any improvements without completely redoing the strategy (like one of the options he mentions at the end of his comments). But I especially like this route because he plans in his needed random encounters at the right times, and since he needs to do them anyway it works out perfectly (for things like showing up at Moonglow at exactly the right moment). He goes to most cities only once, and never more than twice. I think this is a game where death warping for gold could be abused more, but dying takes time and I'm impressed the route was optimized enough that he only deathwarped once.

There are some minor mistakes, most of them cost about 1 second. After Britain he has to cast Wind 1 extra time, a mistake which costs about 5 seconds. But the wasted moves are all on that order, and total well under 5 minutes. It says something that (except for that Wind cast) I had to really pay attention to notice time losses; this is a very optimal version of this route. I wonder if some time could be saved in the Abyss when he casts Tremor a few times when there are 3 or fewer enemies alive, but even so, at 13 seconds per tremor, the few times this happens would still be under a minute.

Timing: I would count the class selection screen as part of gameplay, so I would start the clock at 0:22 and have it end at the last time he advances the dialog, 01:59:52, for a net time of 01:59:30. I certainly didn't know what those silly questions meant the first time I played it, it's not like "choose your class! Ranger, Paladin, Druid", etc.

This run was a pleasure to watch. Decision: Accept.

Decision: Accept

Congratulations to Jeff Feasel!
Thread title:  
The SDA spokes!

Congratulations Speedy Feasy.
Wooo!  Congrats feasel!
difficult and stupidly random
Awesome.  I'm glad you guys liked the run.  I'm also glad people appreciated the optimzations in the dungeon-crawling part.  That is easily the most intense section, since there's just so much memorization and it all comes at you so quickly.  Doing the fewest number of turns and always remembering which directions to hold to skip the ladder texts led to a fair amount of resetting.  The Abyss could have gone faster with better luck, but I think we're talking a rather small amount.  It's really tricky to decide whether you want Tremor or Energy or Quick, especially if the initial Tremor didn't kill many guys.  Sometimes I bet on Tremor even with a low enemy-count because I'm hoping for exceptional luck.

I'm usually pretty critical of my runs, but to be honest I thought this was pretty much as good as you're going to get without a new glitch or a route improvement that I didn't think of (both of which are quite possible).

Using Attack instead of Auto:  Yes sometimes I am losing a couple frames by choosing Attack instead of Auto for the last couple characters when I think the battle will end before it gets to their turn.  I guess I'm just being safe.  I have had ridiculously bad luck encounters where I've got just one enemy left, but my first 3 characters just can't seem to hit it and it comes down to my last character's attack.  Choosing Auto for the last character because I didn't think his turn would come, and then having the Auto action turn out to be something useless like moving or shooting a crappy spell, would be far more costly than losing a few frames.  Still, though, you make a good point, and I'd probably be a little less risk-averse if I ever went back to improve this run.

Redundant Wind:  After Britain I cast Wind to go Up-Right-Up when I could have just done Right-Up.  However I found that if I just do Right-Up then sometimes when I get back in the balloon after visiting the next town the wind will not be the correct direction (up) and I'll have to cast Wind an extra time.  I also feel like the pattern of Wind spells here will have an effect further down the line, and I'm less likely to get random direction changes when doing the next leg of the balloon journey.  Alternatively, I could have just cast nothing at all when I first got into the balloon because the wind was already taking me to the right, however sometimes it pays to cast Wind pre-emptively since (i think) it resets some timer and prevents you from having to deal with random wind direction changes.  In general I just try to be consistant about where I cast Wind, even if it isn't strictly necessary, because I feel like it helps produce more predictable results w.r.t. random direction-changes.  TBH I never fully figured out the way the wind's randomness works, so there probably are some minor optimizations to be had there.

Thanks a lot for the feedback!  For those who don't want to wait, I've got the run on youtube:
Congratulations feasel!
Well done feasel, love the game love the run love the player.