memorable randarts

[Timo Pietilä]

Btw, the generator doesn't over-value AC per se, it calculates AC per unit weight and values that.

Does that mean that +1 in ethereal cloak gets infinite value? Or that ethereal cloak can never have bonus to AC?

Maybe the code over-generalizes things in this case. AC has pretty much fixed value regardless of the item it is found and it should not count anything less than +15 as any value, because you can enchant item with spells and scrolls to that point.

[Estie]

Btw, the generator doesn't over-value AC per se, it calculates AC per unit weight and values that. You may still think some AC values end up too high.

Thats interesting. I suspect that the weight/AC ratio of that helm is actually decent when compared to a heavy mail. Of course a heavy mail is basically out of the question before a substantial amount of str adders has been found.

I think that giving weapons more realistic weights (around 2lbs) would mitigate the armor-weight issue.

(For the record, no, I dont mind zero value randarts at all, I am very happy with them as is, but yes, I am still in the wrong thread.)

[Magnate]

Then that counts as bug.

Man, people need to stop using this thread to discuss randarts.

Anyway, yes, there's always been a problem with weak randarts, because there's often no room for interesting abilities after choosing the base item and initial plusses. There's also a separate problem with "bad" randarts because there are so few acceptable "bad" mods in the game. The generator originally used Derakon's approach of +25 and -50 (or thereabouts), but everything ended up aggravating and being junk. Maybe that doesn't matter - but if you want more interesting bad randarts, we need more interesting bad mods.

It would definitely be possible to solve the first problem and start the generation of a weak randart with an 'interesting' ability and then find a suitable base item.

Btw, the generator doesn't over-value AC per se, it calculates AC per unit weight and values that. You may still think some AC values end up too high.

[Timo Pietilä]

I checked the helm, it was based on one of the paur gauntlets.

Then that counts as bug.

[Derakon]

I thought of that and I think many of those "bad" artifacts are actually quite close to zero power if you combine bad and good qualities. Beruthiel could be the one that one was based on. None of the purely good ones are close to zero power.

Weakest "good" artifact I can think of is Thorongil, but even that is quite a lot better than steel helmet with mediocre bonus to AC.

Remember that there's a decent bit of variance between the power of a standart and the power of the randart it's based on. A low roll based on a low-power artifact can produce "uninteresting" randarts too.

[Estie]

I checked the helm, it was based on one of the paur gauntlets.

[Timo Pietilä]

Actually, there's a point -- how are "bad" standarts handled by the randart generator? It should try to make an equivalent "bad" randart, not just a randart with equivalent total power. In other words, when evaluating standarts, the power of both the benefits of the item, and the penalties on it, should both be determined; then the randart should try for equal power on both. So e.g. Beruthiel might have a positive power of 50 and a negative power of 75; I wouldn't be surprised if currently the game tries to make a randart with power -25, when it should try to make a randart with positive power 50 and negative power -75 (with some variance of course).

I thought of that and I think many of those "bad" artifacts are actually quite close to zero power if you combine bad and good qualities. Beruthiel could be the one that one was based on. None of the purely good ones are close to zero power.

Weakest "good" artifact I can think of is Thorongil, but even that is quite a lot better than steel helmet with mediocre bonus to AC.

[Derakon]

But it would be special. I wonder, which item from original set has so low value that it generates randart without any value?

The Paur* gauntlets leap to mind (especially Paurnen), as do Beruthiel and Camlost, who have so many penalties that their values are frankly negative IMO. I don't know the variance on object power, but I suppose it's possible that an exceptionally poor roll for the *thancs or Thengel could also result in a powerless artifact. Gorlim is pretty terrible too, though at least it has +to-dam which ought to boost its perceived power significantly.

Actually, there's a point -- how are "bad" standarts handled by the randart generator? It should try to make an equivalent "bad" randart, not just a randart with equivalent total power. In other words, when evaluating standarts, the power of both the benefits of the item, and the penalties on it, should both be determined; then the randart should try for equal power on both. So e.g. Beruthiel might have a positive power of 50 and a negative power of 75; I wouldn't be surprised if currently the game tries to make a randart with power -25, when it should try to make a randart with positive power 50 and negative power -75 (with some variance of course).

[Timo Pietilä]

I have already reported this (artifacts with only extra tohit/todam/toac), and the problem is that many artifacts have a really low power. In that case, there's usually not much room for anything else than a weak ability or extra tohit/todam/toac. The same helm with +6 toac instead of +16 and feather falling would not have been much more special...

But it would be special. I wonder, which item from original set has so low value that it generates randart without any value?

I think the root problem here is that randart calc values AC way too high. 9+16 is next to nothing, it's immediate junk unless it is dlvl1 item with rarity "found always", and even then it was in item that is heavy so I would probably still not use it. +10 AC less with FF should be about ten times more valuable at those shallow levels that +16AC could have some value.

In randart generation anything less than +20 AC (as bonus, not whole) should count nothing at all.

[PowerWyrm]

IMO that is borderlining a bug. Artifacts should have something to make them special. This does not have anything.

I have already reported this (artifacts with only extra tohit/todam/toac), and the problem is that many artifacts have a really low power. In that case, there's usually not much room for anything else than a weak ability or extra tohit/todam/toac. The same helm with +6 toac instead of +16 and feather falling would not have been much more special...

[Philip]

Yeah, the Glaive of pain only has lots and lots of damage, but higher damage dice count for a lot, and honestly, +30 is insane. Damage is useful, unlike AC, which is not useful for much. It is handy against mobs of monsters with attacks that do HP damage, as opposed to inventory damage or character damage. Actually, the subsection of those monsters you fight hand-to-hand.

For an objects AC to be interesting, it would have to give at least 50 more AC than I would reasonably expect, since the Glaive has about 10 or so more damage than I would reasonably expect on an artifact. If the helm gave 60-80 AC or so, it would be an interesting object.

I agree about mithril helms, feels fairly natural. I can't think of anything else that is mentioned to be mithril, but gauntlets would qualify, too.

[Timo Pietilä]

And having high AC isn´t special in this sense, I take it, while having, say, fire resistance is. That means assigning no value to AC like suggested above would solve.

Well, there is Glaive of Pain which has nothing but insane damage (and now later pFear). Maybe if AC bonus would be higher than +20 which is unattainable with enchant item scrolls/spells it could qualify. At +16 however it is just barely better than ordinary steel helmet which someone had enchanted a lot. Only special features that thing had was ignore acid and immune to manastorm destruction (automatic from being artifact). Not even AC was exceptional.

BTW we need mithril helmets. Guards of the Citadel in Denethors court had those, so they should be rather common item (or at least possible base item meaning rarer than shallow helmets items, but more common than egos later).

Helmets are now:

Hard Leather Cap
Metal Cap
Iron Helm
and
Steel Helm

None of those ignore elements (acid).
(+ four crowns with base 0 AC)

[Estie]

IMO in order to become an artifact object needs to have something special. Something that makes people of the realm remember and recognize the thing.

And having high AC isn´t special in this sense, I take it, while having, say, fire resistance is. That means assigning no value to AC like suggested above would solve.
One could also require "at least 2 properties for each artifact"; I wonder if it would still be possible to get an AC only helm with extra weight or so.

[Timo Pietilä]

Well there are bad artifacts, too; so having zero value ones isnt horrible, as long as they are rare enough.

IMO in order to become an artifact object needs to have something special. Something that makes people of the realm remember and recognize the thing.

[Estie]

IMO that is borderlining a bug. Artifacts should have something to make them special. This does not have anything.

Well there are bad artifacts, too; so having zero value ones isnt horrible, as long as they are rare enough.

I wonder how much value the generator gives AC. Its probably impossible to get right, because the value of AC changes during the game, and similarly, of weight. Maybe assigning weight a negative value would be a good idea.

This helmet was junk for me when I found it: early in the game and already at speed -1 from weight, I wasnt going to add more load for a bit of AC. If it had been a leather cap +16, I might have equipped it till something better turned up (like a hat of infravision).

For the lategame weight is all but irrelevant (you go from speed 23 to 22 or so, who cares), but a small negative value adjustment shouldnt break the system there.

Another option is to assign AC a value of zero. As long as the distribution of AC for the artifacts remains the same, youll end up with between 200 and 300 AC for the end kit anyway and the difference between the 2 is negligable.