How does LEMMISI works ?

[flap]

I have made some experimentations and updated the wiki page. (It seeks glyphs first)

Do you have no more ideas on the LEMMISI priorities or rules ?

[Darvin]

LEMMISI, at its heart, is random. It is weighted to have a higher probability of revealing interesting things, but at the end of the day there's no assurances as to what - if anything - it's going to reveal.

Short of an absolutely massive and meticulously cataloged data-set, you aren't going to be able to experimentally filter out the noise.

[flap]

Well, I have tried is on about 15 dungeons, and found what I wrote on the wiki pages : in 3 to 4 cast you can hope to have seen every glyphs, 6 or 7 every altars and maybe stairs. And it takes many more to spot every monsters.
That's already good to know (before trying this, I had not realised that it was that efficient at finding glyphs and altars)

[Darvin]

Well, I have tried is on about 15 dungeons, and found what I wrote on the wiki pages : in 3 to 4 cast you can hope to have seen every glyphs, 6 or 7 every altars and maybe stairs. And it takes many more to spot every monsters.
That's already good to know (before trying this, I had not realised that it was that efficient at finding glyphs and altars)

You can absolutely make generalized statements about its priorities and effectiveness. Getting higher granularity and number-crunching for actual probabilities is a completely different ball-game. Short of the devs outright stating how the weighting is implemented (as they did with JJ's random punishments) I don't think we'll be able to get much better than that.

[Urthop]

Well, I have tried is on about 15 dungeons, and found what I wrote on the wiki pages : in 3 to 4 cast you can hope to have seen every glyphs, 6 or 7 every altars and maybe stairs. And it takes many more to spot every monsters.
That's already good to know (before trying this, I had not realised that it was that efficient at finding glyphs and altars)

You can absolutely make generalized statements about its priorities and effectiveness. Getting higher granularity and number-crunching for actual probabilities is a completely different ball-game. Short of the devs outright stating how the weighting is implemented (as they did with JJ's random punishments) I don't think we'll be able to get much better than that.

While that is true for the exact numbers, his statement is roughly about right for what I've found. Priority does seem to be Glyphs>Altars>Stairs>Shops>Monsters. I'd add a few more to the numbers though, 3/4 seems awfully low for all glyphs.

[flap]

Ah well, I might be wrong with the figures. But still, in my experience I often see all glyphs by the 4th cast.
I am really pleased to have done that bit of testing, because now I find Lemmisi way more useful than before !

[Darvin]

Almost always worth casting LEMMISI three or four times if you find it early enough. Still usually converted immediately afterwards, though.

[flap]

That's what I am discovering !

[Lujo]

Almost always worth casting LEMMISI three or four times if you find it early enough. Still usually converted immediately afterwards, though.

Even more than that if youre not counting on blackpace... for the spike, lol. Knowing where everything is is well worth it, and can end up saving you exploration. I consider this guaranteed scouting to be the strongest feature of the Transmuter in most cases (oddly enough).

[Darvin]

Even more than that if youre not counting on blackpace... for the spike, lol. Knowing where everything is is well worth it

Agreed; I'm on the "explore generously, then make up for the lost blackspace by leveraging what you've learned about the dungeon/glyph/monster/shop spawns" side of things.

I consider this guaranteed scouting to be the strongest feature of the Transmuter in most cases (oddly enough).

Doesn't seem odd to me at all. I consider the free WEYTWUT glyph one of the best reasons to prep JJ. Early popcorn bowling and declogging? Yes please!