Even more thoughts on v4

[Magnate]

I think part of the problem comes from plain multiplication of several mediocore items. One weapon of HA is much better than 200 weapons of slay Orc.

Amount of weapons should not count as much as quality of weapons. Quality over quantity.

Maybe some sort of power level sorting and then halving the value of each weapon past that?

1/1 for best 1/2 for second 1/4 for next and so on.

Each item slot should have unique "quota" of things, and consumables are not counted like that (two potions of Life is twice as good as one potion of Life).

Too complicated?

While I'm not one to criticise complexity per se, I think most of what you want to achieve can be done by increasing the contribution of high-quality items over low quality items - i.e. to use something like object_value squared. (We have to use value rather than power, because consumables don't have a power rating.)

So if 200 weapons of slay orc are each worth 1000gp, that's a total contribution to the level feeling of 200 x 1000 x 1000 = 200M. If the HA is worth 50kgp, its contribution would be 50k x 50k = 2500M - over twelve times as much. (Under the current system the results would be 200k and 50k, respectively.)

But if anyone wanted to code up your slot-based idea, it would be interesting to try it out.

[Magnate]

I agree with those that say that fuzzy detection should detect squelched objects, which then disappear when they come in LoS. We could provide a message, "you ignore a xxx" when you see the object, so that it is clear why it is disappearing. I'm also a fan of more granular object detection. We could allow for fuzzy detection that allows the type of object (potion, scroll, boots, etc.). One that distinguishes between money and other objects. One that only detects objects with magical enchantments.

This is all good. I'm hoping that my new effects system will make it possible to do this with edit files (though I haven't worked out all the ramifications for ESP yet).

You could even have a powerful scroll of artifact detection that allows all artifacts to be detected and identified, with the downside that they will not be preserved if left on the level. A true double edged sword.

Nice!

Re: damage. I'm fine with tweaking some of the egos so that they have an m-bonus value. This would give us some additional increase in damage and to-hit as a function of dungeon level. Right now, I think it's hard to ego weapons to compete with artifacts. Regardless, Magnate originally decided to want to have egos become better by having additional affixes, rather than better affixes or level-weighted affixes. Since the entire affix approach was his gruntwork, I'm definitely willing to give this approach a shot.

Hmmm. I may have mis-communicated my thoughts on that. Or perhaps they have changed since we last caught up. I think what happened is this:

- I failed to anticipate how few of the affixes actually improve +hit or +dam, meaning that both were far rarer than in V

- I first thought to fix this by allowing repeated applications of those affixes which did boost hit/dam, and adjusting their probabilities

- But then we agreed that items would only have one each of the make, material or quality prefixes. This was a huge step forward for naming and flavour consistency, but exacerbated the hit/dam problem (because all but one of the affixes affecting hit/dam are makes, materials or qualities).

- So we cranked up the likelihood of the Slaying suffix, but it's clear that this isn't the solution.

I can think of two approaches that would help, which are not mutually exclusive:

1. Increase the bonuses from the make/material/quality affixes. These are mostly taken from EyAngband, where they were deliberately set as minor effects (no more than a couple of extra sides, or +1 or +2 to hit or dam). But this was because Ey still used the old hit/dam bonuses on ego items, which have now disappeared. So IMO it would be interesting to increase the range of plusses available on make/material/quality prefixes, up to +5 or so for the best ones - especially since items are limited to one of each.

2. Add an m_bonus element to the Slaying suffix, so that it gets better with dlev. I'm sorry if I have previously argued against this - I've been thinking about it for the last couple of weeks. My fear is that it will make the endgame awash with overpowered egos, but Derakon's nearly at 4000' and this doesn't seem to be a problem so far ...

[Derakon]

A few notes on my playstyle:

* I'm playing with level feelings off. This may have some minor impact on my ability to find good gear, but I've been playing this way ever since the level feeling was split into danger and treasure feelings, because it feels wrong otherwise.

* I'm not skipping huge amounts of most levels. In particular, I'm detecting basically the entire level, which lets me see where items are. I then go after the items and item-dropping monsters I can handle that aren't too close to dangerous uniques, angel pits, and similar huge threats. Sometimes this means I legitimately clear the entire level. Sometimes I get boxed in early and flee. In practice I'm clearing well over half of each level, and I've done a lot of levels now; by all rights I should have found a decent weapon by now.

I think that scaling Slaying with level sounds like a good idea. We can start small -- multiply it by level / 20 or something. I'd be leery of having the make/material affixes affect the dice more than they do now, mostly because it makes branding more powerful. Having more powerful bonuses deeper down might also work, though you'd still need to stack several such to get a competitive weapon, I suspect.

I have noticed that launchers often have good bonuses on them because the Power and Accuracy affixes are comparatively large

[fizzix]

I meant to respond to Derakon's comments about items from angel pits/graveyards showing up in sims. I'll do that here.

I agree this is a huge limitation in the simulation. My stats does distinguish between items that came from the floor vs. those that came from monsters. And it even distinguishes floor vs standard vaults. However, it does not distinguish between floor and pits, so if the item showed up on the floor of a graveyard, you would still see it as a floor item. Myshkin's stat module can distinguish between these things and, I believe, you can filter out all objects from pit monsters or found on floors of pits. Theoretically, I could do that now too, but I'm not sure how much it gains for the analysis.

It might be a good idea to use one of Magnate's sims and try to calculate what percentage of top egos come from graveyards/angel pits/demon pits/golem pits that are probably unlikely to be cleared and see how damaging this effect is.

[Magnate]

A few notes on my playstyle:

* I'm playing with level feelings off. This may have some minor impact on my ability to find good gear, but I've been playing this way ever since the level feeling was split into danger and treasure feelings, because it feels wrong otherwise.

* I'm not skipping huge amounts of most levels. In particular, I'm detecting basically the entire level, which lets me see where items are. I then go after the items and item-dropping monsters I can handle that aren't too close to dangerous uniques, angel pits, and similar huge threats. Sometimes this means I legitimately clear the entire level. Sometimes I get boxed in early and flee. In practice I'm clearing well over half of each level, and I've done a lot of levels now; by all rights I should have found a decent weapon by now.

I think that scaling Slaying with level sounds like a good idea. We can start small -- multiply it by level / 20 or something. I'd be leery of having the make/material affixes affect the dice more than they do now, mostly because it makes branding more powerful. Having more powerful bonuses deeper down might also work, though you'd still need to stack several such to get a competitive weapon, I suspect.

I have noticed that launchers often have good bonuses on them because the Power and Accuracy affixes are comparatively large

Well now there's another possibility: allow those two suffixes on melee weapons. That would make a big difference.

m_bonus doesn't work quite as a linear scale-with-level, it's still a random roll, but the results are skewed by depth, so M10 is much more likely to be +1 or +2 at dl1 and much more likely to be +9 or +10 at dl100. But you can still get +10 at dl1 or +1 at dl100.

I wasn't intending to make any further changes to damage dice on prefixes, but simply to increase the +hit and +dam bonuses on them. I'll package up a bunch of these changes, run some stats and see what happens.

At the risk of changing the subject, I'm interested in your antipathy to the split level feelings. Is it because when combined they give too much away? Or because the monster feeling is immediate? Or ...?

[Magnate]

I meant to respond to Derakon's comments about items from angel pits/graveyards showing up in sims. I'll do that here.

I agree this is a huge limitation in the simulation. My stats does distinguish between items that came from the floor vs. those that came from monsters. And it even distinguishes floor vs standard vaults. However, it does not distinguish between floor and pits, so if the item showed up on the floor of a graveyard, you would still see it as a floor item. Myshkin's stat module can distinguish between these things and, I believe, you can filter out all objects from pit monsters or found on floors of pits. Theoretically, I could do that now too, but I'm not sure how much it gains for the analysis.

It might be a good idea to use one of Magnate's sims and try to calculate what percentage of top egos come from graveyards/angel pits/demon pits/golem pits that are probably unlikely to be cleared and see how damaging this effect is.

Good point. I had forgotten that myshkin's module already distinguishes between all the origin types, so we can indeed filter out all the DROP_PIT and FLOOR_PIT items. I'll see how much difference this makes to the stats for v4.

[Nomad]

If I can get my ego naming system up and running (probably won't have time to work on it more until after Christmas, alas) it should be possible to have any number of affixes that boost to-hit and to-dam all display with the same name 'of Slaying'. (Or even implement special checks to decide whether an item merits the name Slaying, Accuracy, or Damage according to the overall value of its bonuses.) In which case, adding as many more affixes as necessary to boost the values shouldn't be to much of a problem flavour-wise.

Or, to come at it a different way, how about making scrolls of enchant weapon more readily available and perhaps increasing the odds of them working on weapons that already have high values? A lot of the time in v4, my best weapon ends up being one that started out with good runes but weak bonuses that I gradually enchanted to +10 or 11 to dam by collecting scrolls.

[fizzix]

Or, to come at it a different way, how about making scrolls of enchant weapon more readily available and perhaps increasing the odds of them working on weapons that already have high values? A lot of the time in v4, my best weapon ends up being one that started out with good runes but weak bonuses that I gradually enchanted to +10 or 11 to dam by collecting scrolls.

I'm ok with making them more common, but I don't propose we go back to stocking them in the stores. On the other hand, I feel like *enchant weapon* should have a good chance of enchanting up to +20 or so. I'm ok with the normal enchant only going up to +10, because several classes get that as a spell.

[Magnate]

If I can get my ego naming system up and running (probably won't have time to work on it more until after Christmas, alas) it should be possible to have any number of affixes that boost to-hit and to-dam all display with the same name 'of Slaying'. (Or even implement special checks to decide whether an item merits the name Slaying, Accuracy, or Damage according to the overall value of its bonuses.) In which case, adding as many more affixes as necessary to boost the values shouldn't be to much of a problem flavour-wise.

Or, to come at it a different way, how about making scrolls of enchant weapon more readily available and perhaps increasing the odds of them working on weapons that already have high values? A lot of the time in v4, my best weapon ends up being one that started out with good runes but weak bonuses that I gradually enchanted to +10 or 11 to dam by collecting scrolls.

Yep, that's good too - though enchanting should be optional rather than necessary, so I'm inclined to leave the availability the same and just tweak the success rate.

So it turns out that pits have a significant effect on finds after dl70 or so - see the attached graph for artifacts. The pits add a steadily increasing percentage of artifacts, from 0% to 23% for dl1-67 and from 26-75% from dl68-98. Artifacts are so common on dls 99 and 100 that the effect of pits actually decreases.

I assume that the graphs are the same shape for egos but with different numbers, but I'll report back if not (EDIT: I was right, the graph for themes is exactly the same shape - even the percentage added by pits is roughly the same throughout the dungeon). Since the pits are the same as they are in V, it should be roughly the same there. I quite like this - it makes taking on pits quite significantly rewarding. YMMV.

[Derakon]

At the risk of changing the subject, I'm interested in your antipathy to the split level feelings. Is it because when combined they give too much away? Or because the monster feeling is immediate? Or ...?

It's a bit "gives too much away" and a bit "ruins flavor". I loved how the old feelings could have meant there was nice gear, but could just have easily have meant that there was a nasty monster waiting to decorate his lair with your teeth. There's no ambiguity in the new feelings, so they aren't anywhere near as interesting. They really just feel like being told whether or not I should bother with a level right at the start. Since I don't want to play the game second-guessing if I should even be on a given level, I turned them off.

[Derakon]

The RNG has a sense of humor. Check out this, my fourth artifact all game:

the Chain Mail 'Eanwe' (-2) [32,+24] <+2>
+2 CON, speed.
Resist acid/elec/fire/cold/poison/light/dark/sound
Protect against fear/blindness
Regeneration
Aggravates

I got this while clearing out one of the really rather frequent angel pits I've been encountering, at 4150' (the pit only had one Istar, which is all that made it feasible). I'd be more than happy to use it over my Elvenkind armor with rNether, except for that whole aggravation thing. Dangit, RNG.

* On a more serious note, there are three artifacts in the artifact knowledge list that I've never seen: Kelth, Natar, and Careglin. Judging from my limited artifact spoiler they're all mid-range artifacts with a generation chance of 99, which AFAICT is the most common an artifact can get, not counting the spellbooks. So something is screwy here. I do have preserve mode on.

* I saw a dracolich in an ice dragon pit.

* I've finally upgraded my weapon! From that same angel pit, I got a Brutal Morningstar of Westernesse (2d6) (+8,+11) <+2>. Standard westernesse except it also has electricity and poison brands. When those kick in I'm doing 70 more damage/round than before; if they don't, I deal 8 less. I figure on balance I'm better off, especially since the +2 CON gets me ~80 more hitpoints. Not like this is remotely enough to actually go hunting the big game just yet though.

* *Enlightenment* probably shouldn't be a "good" droppable; for most characters its utility is equal to a potion of Their Spell Stat + a potion of Enlightenment, both of which are pretty common. Of course, my warrior is even less enthusiastic to get this from offing Ar-Pharazon...

* Accented characters show as two blank spaces in monster recall (c.f. Ar-Pharazon, whose description mentions Numenor twice). Mentioned for completeness's sake; if nothing else it should be an easy repro case for the UTF-8 issue.

* Found item:

Ring of Light <, +1>.
You do not know the full extent of this item's powers.
+1 light radius.
Also affects your searching skill.

On being identified, turns out it gave +4 to searching as the first pval.

* I found Wrath of God sitting in a lesser vault at 4250'. So that's one spellbook that wasn't dropped by its owner.

* I found a weapon with 2 Nightbane affixes. Intended? AFAICT this affix only puts the *Slay* Undead property on the item, so stacking it has no effect. For completeness's sake:

a Topaz Battle Axe of *Slay Undead* (2d8) (+16,+10) <+1>
+1 wisdom.
*Slays* dragons, undead.
Branded with weak lightning.
Grants the ability to see invisible things.

* It seems weird that dragonscale armor can be damaged by acid. Currently it basically appears to be "armor of -2 to-hit and 1 elemental resistance", which is underwhelming.

* Why do gloves get the Featherfall affix? IMO that should be a boots/cloaks/body armor affix. Maybe shields so you can glide like a majestic Olaf by holding it above your head. </obscure>

EDIT: * Found a Massive Elven Rapier. "A slender sword, perfect for thrusting. It is lighter and more accurate than usual, favoured by wood-elf hunters. It is considerably larger and heavier than usual." Some affixes should be mutually exclusive, I think...

* Burden information displays in subwindows. It just doesn't display in the main window.

* It seems like lately mimics are targeting high-value items. Potions of experience, rings of speed, scrolls of banishment, etc. Is this intentional?

I don't really like upgrading game versions mid-character; I'll finish Barlowe off one way or another before getting the latest and greatest.

[Magnate]

* On a more serious note, there are three artifacts in the artifact knowledge list that I've never seen: Kelth, Natar, and Careglin. Judging from my limited artifact spoiler they're all mid-range artifacts with a generation chance of 99, which AFAICT is the most common an artifact can get, not counting the spellbooks. So something is screwy here. I do have preserve mode on.

Oh yes, I meant to include this one in the "known issues in v4" section of my update post. Not yet quite sure how it's happening, because it only happens to a very small number of artifacts. I think it is specific to randarts - the only thing I can think of is the added kinds (INSTA_ARTs). But AFAICR the randart code still leaves them alone. I can't think what else has changed in v4 wrt randarts.

* I saw a dracolich in an ice dragon pit.

That's legit, I think - they breathe cold, don't they?

* *Enlightenment* probably shouldn't be a "good" droppable; for most characters its utility is equal to a potion of Their Spell Stat + a potion of Enlightenment, both of which are pretty common. Of course, my warrior is even less enthusiastic to get this from offing Ar-Pharazon...

Well historically this is considered the ultimate potion in the game (highest sale value), so I'm loath to mess with such a trope ... but I kind of agree with you.

* Found item:

Ring of Light <, +1>.
You do not know the full extent of this item's powers.
+1 light radius.
Also affects your searching skill.

On being identified, turns out it gave +4 to searching as the first pval.

Ugh, not seen this before. Yuck. Will try to reproduce.

* I found a weapon with 2 Nightbane affixes. Intended? AFAICT this affix only puts the *Slay* Undead property on the item, so stacking it has no effect.

At the moment we don't prevent any duplicate suffices. If you get Nightbane twice, you've effectively got one less affix (remember that the individual affixes won't be listed eventually, once we sort names properly). So it's not doing any harm.

* It seems weird that dragonscale armor can be damaged by acid. Currently it basically appears to be "armor of -2 to-hit and 1 elemental resistance", which is underwhelming.

Yes, DSM items are not intended to be uber every time, just flavourful. Sometimes they will work in fantastic combinations, other times they will underwhelm. Non-black dragons are vulnerable to acid, like everyone else ...

* Why do gloves get the Featherfall affix? IMO that should be a boots/cloaks/body armor affix. Maybe shields so you can glide like a majestic Olaf by holding it above your head. </obscure>

This was just for variety really, there were very few legal affixes for gloves. Happy for people to suggest alternatives, though again I don't think it hurts much.

EDIT: * Found a Massive Elven Rapier. "A slender sword, perfect for thrusting. It is lighter and more accurate than usual, favoured by wood-elf hunters. It is considerably larger and heavier than usual." Some affixes should be mutually exclusive, I think...

Yes, we have a ticket for precisely this (#1562). Though in this particular case I can imagine an elven smith forging something massive for a half-giant friend ...

* Burden information displays in subwindows. It just doesn't display in the main window.

Did it ever?

* It seems like lately mimics are targeting high-value items. Potions of experience, rings of speed, scrolls of banishment, etc. Is this intentional?

Yes, IIRC it's to make you go after them ...

[fizzix]

Well historically this is considered the ultimate potion in the game (highest sale value), so I'm loath to mess with such a trope

After mercilessly slaughtering so many sacred cows, this one gives you pause?

*Enlightenment* was historically powerful because it was one of the only ways you could find out what you resisted. That use is gone, and it should currently be considered to be slightly better than an !Int/!Wis potion. That is, about on par with !Con or !Str.

[Derakon]

That's legit, I think - they breathe cold, don't they?

That's as may be, but they don't fit with the other monsters, who are all clearly on the list of "ice dragon" tiers. Dracoliches are undead, breathe nether, and bite to drain experience, which all strongly differentiate them from their nominal peers in the pit.

Yuck. Will try to reproduce.

It should happen with any unidentified Ring of Light.

Yes, DSM items are not intended to be uber every time, just flavourful. Sometimes they will work in fantastic combinations, other times they will underwhelm. Non-black dragons are vulnerable to acid, like everyone else

I'd rather they be more powerful and more rare, personally. Right now they're actively worse than standard Resist Acid affixes. That's just wrong.

Did it ever?

I think so.

[Magnate]

That's as may be, but they don't fit with the other monsters, who are all clearly on the list of "ice dragon" tiers. Dracoliches are undead, breathe nether, and bite to drain experience, which all strongly differentiate them from their nominal peers in the pit.

Not only can I not speak for the person who designed the new pits, I can't even remember who it was off the top of my head - but ISTR that this is precisely the kind of variety that they intended. But YMMV, of course.

I'd rather they be more powerful and more rare, personally. Right now they're actively worse than standard Resist Acid affixes.

They're not, actually, because they provide an AC boost (though I grant that this is unlikely to be noticeable on small items, since it's percentage based and black is at the bottom end of the DSM spectrum). But comparing them individually to resist affixes is not really the point, because DSM will become more useful in combination, gain activations etc. Rest assured that I am very wary of ending up with two indistinguishable things with different names.

On a related note, are you finding DSM common?? It should be significantly rarer than the corresponding resist affix, for the low elements.