Worst Randart Ever...

[miyazaki]

Well, maybe not worst, but definitely the most boring. The thrill of seeing {special} died a limp death as I ID'ed this:

l) The War Hammer of Adorn (3d3) (+7,+5)
     Cannot be harmed by acid, electricity, fire, cold.
     
     Combat info:
     4 blows/round.
     Average damage/round: 34.7

[Mondkalb]

That's more a hammer o'darn ...
or a hammer of o darn!

[Zikke]

I had a very similar randart last week. Basic weapon, no special powers. Might as well be a {magical} item that can't be broken.


-______-

[Schroeder]

If only it was cursed and agro'd. THEN it'd be the worst.

[Magnate]

Well, maybe not worst, but definitely the most boring. The thrill of seeing {special} died a limp death as I ID'ed this:

l) The War Hammer of Adorn (3d3) (+7,+5)
     Cannot be harmed by acid, electricity, fire, cold.
     
     Combat info:
     4 blows/round.
     Average damage/round: 34.7

What version was this? That's a terribly weak randart, and shouldn't happen in dev versions of 3.1.1 (post r1248).

[Atarlost]

At least it's well named. A warhammer of adornment.

[PowerDiver]

What version was this? That's a terribly weak randart, and shouldn't happen in dev versions of 3.1.1 (post r1248).

Are you sure some mixed artifact like boots of wormtongue cannot evaluate to boring? If the power code weakens speed sufficiently, that could eventually have 0 power.

Speaking of which, did you ever include weight in the power algorithm? 30 lb of weight should equal -1 speed. That may make things like iron helms have negative power, but IMO that could be correct. I was perturbed to sell a cloak [1, +4] and see it was cheaper than an iron helm [5, +0].

You may need to modify the randart algorithm to guarantee substantial mixed qualities when the base power is close to 0, or else just enforce a minimum power value for any randart even if the standart evaluates to junk.

[Magnate]

Are you sure some mixed artifact like boots of wormtongue cannot evaluate to boring? If the power code weakens speed sufficiently, that could eventually have 0 power.

Where did you get the idea that speed is weakened? Relative to what?

Speaking of which, did you ever include weight in the power algorithm? 30 lb of weight should equal -1 speed. That may make things like iron helms have negative power, but IMO that could be correct. I was perturbed to sell a cloak [1, +4] and see it was cheaper than an iron helm [5, +0].

This is a combination of things. Mainly it's because base AC is rated at 1.5x +ac, because base AC can't be disenchanted. That 1.5x is too high, it should be more like 1.2x, but I don't have that sort of granularity available (I'm using a doubled integer scale, so 0.5 is the smallest increment).

It's partly because I did give up on trying to factor in weight. Heavy armours are already far far cheaper than they were before power-based pricing, and penalising them for their weight exacerbates this problem, while gaining very few benefits (the reduction in power for things like heavy helms and shields was lost in integer arithmetic so as to be negligible).

One way I plan to modify this is to make the power scale for base AC nonlinear. An iron helm isn't really worth as much as a +6 cloak, but +0 mithril plate is worth more than +7 ribbed plate. Once I do that I can return to the weight issue without it gimping the armour prices.

You may need to modify the randart algorithm to guarantee substantial mixed qualities when the base power is close to 0, or else just enforce a minimum power value for any randart even if the standart evaluates to junk.

Well, after 3.1.1 is out I plan to separate positive and negative attributes completely - negative attributes won't be part of the power rating at all (apart from for pricing purposes) - the randarts will be generated against all positive attributes, and then the right proportion of negatives will be sprinkled among them. This will prevent things like Wormtongue turning into junk.

I'm still interested in what version the OP was using. Even Wormtongue has never turned into quite such a boring item for me - though I guess it's possible for one of 130 randarts to be junk every few games ...

[PowerDiver]

Where did you get the idea that speed is weakened? Relative to what?

Compared to how it used to be.

It's partly because I did give up on trying to factor in weight.

There's no need to factor. 30 lbs = -1 speed is totally obvious. The only problem is that you are approximating non-linear things with an additive power ranking. That's a general problem, not a "what is the penalty of weight" problem.

E.g. I tossed leather scale of resist acid [useless wormtongue drop] even though it was my only source of rAcid, to wear a robe[2, +6] in my current game. It just wasn't worth the weight.

Heavy armours are already far far cheaper than they were before power-based pricing, and penalising them for their weight exacerbates this problem

If you believe in power pricing, it is not a "problem" to price heavy armors cheaper.

If an iron helm [5,+0] costs more than a leather cap [2, +0] you've still got it wrong IMO. I always toss the former to keep the latter once my pack fills to the point I need to make a choice. 3 points of AC isn't worth worrying about being speed (-1) even occasionally.

The reason to wear the cap [other than as a placeholder for additional magical AC enchantment] is not for the 2, but in case of saving some damage from an acid breath. Does your power algorithm take that into consideration? I.e. that any good or avg armor regardless of AC has a chance to reduce acid damage. First approximation is (value of rAcid) * (1/6) * (1/3) / (1/2) but that is wrong because multiple rAcid do not help but multiple armors do. I'd say every armor [without AC minuses] should start with value rAcid/3 or maybe even rAcid/2. You could think of some of that as the value of a placeholder for recipient of future ?enchant as well [an additional fixed significant value]. That might avoid some negative prices when armor weight is taken into account.

[Atarlost]

Is there some way to find out what the base artifact is? I can't think of anything nearly this lame that doesn't have an activation.

[Magnate]

Is there some way to find out what the base artifact is? I can't think of anything nearly this lame that doesn't have an activation.

Neither can I. I can work it out if the OP sends me the randart.log file (there are no names in the file, so you need to scan it for a warhammer with those plusses, and look up the artifact number).

[Magnate]

Compared to how it used to be.

That makes no sense. Speed items are cheaper to buy from shops than they used to be, but speed is no less powerful, nor less common, than it has ever been. I don't understand your point here.

There's no need to factor. 30 lbs = -1 speed is totally obvious. The only problem is that you are approximating non-linear things with an additive power ranking. That's a general problem, not a "what is the penalty of weight" problem.

30lbs = -1 speed is not obvious at all, as the amount of weight that gives -1 speed varies by STR which in turn varies by class and race, and the second and subsequent -1 speed penalties are not the same as the first either in weight or value. But you are right that an additive power rating has difficulty with nonlinear approximations.

E.g. I tossed leather scale of resist acid [useless wormtongue drop] even though it was my only source of rAcid, to wear a robe[2, +6] in my current game. It just wasn't worth the weight.

Leather scale is not heavy, as body armours go, even in the early game. That must have been a weak or heavily burdened character.

Please rest assured that I do think weight is important, and I will return to it (both for weapons and armours). It's just not currently factored into the power rating (apart from a tiny +1 power for items which weigh less than their standard weight). It is on my lists of things to work on next (tickets #798 for the long-term and #853 for the shorter term).

If you believe in power pricing, it is not a "problem" to price heavy armors cheaper.

No, but change can fail simply because it is too different to what went before. Power-based pricing has already shaken some long-held tenets about prices of certain things (ESP, speed), so I decided that heavy armours had been reduced enough without at the moment reducing them still further because of their weight.

If an iron helm [5,+0] costs more than a leather cap [2, +0] you've still got it wrong IMO. I always toss the former to keep the latter once my pack fills to the point I need to make a choice. 3 points of AC isn't worth worrying about being speed (-1) even occasionally.

Well, I haven't created this problem - iron helms have *always* cost more than leather caps. But your views on the importance of weight are duly noted - I'm curiously reassured that there isn't anything more significant to worry about (that's not an invitation to rant - I'm well aware of outstanding issues like plite and Narya vs Thorin).

The reason to wear the cap [other than as a placeholder for additional magical AC enchantment] is not for the 2, but in case of saving some damage from an acid breath. Does your power algorithm take that into consideration? I.e. that any good or avg armor regardless of AC has a chance to reduce acid damage. First approximation is (value of rAcid) * (1/6) * (1/3) / (1/2) but that is wrong because multiple rAcid do not help but multiple armors do. I'd say every armor [without AC minuses] should start with value rAcid/3 or maybe even rAcid/2. You could think of some of that as the value of a placeholder for recipient of future ?enchant as well [an additional fixed significant value]. That might avoid some negative prices when armor weight is taken into account.

Yes, it might. It's also a lot more code for a tiny refinement of the algorithm - I'm not unwilling, it just isn't top priority. The value of rAcid already reflects the fact that armour items reduce acid damage. The 1.5 mult for base AC also partially reflects that (though I'm now convinced that base AC needs a nonlinear lookup table like speed - I'll do that next time I work on obj-power.c). But yes, since we now have a base power rating for jewelry, there's absolutely no harm in having one for armour items. Nice idea - thank you.

[PowerDiver]

30lbs = -1 speed is not obvious at all, as the amount of weight that gives -1 speed varies by STR which in turn varies by class and race.

Once you get to 18/50 str and 180+ lbs of inventory it is accurate, and I daresay even most mages manage that before 2500'. Randart generation should certainly be in the context that includes more than half of the dungeon. Earlier on, weight should perhaps cost more, but that is mitigated a bit by a certain amount of weight for free. Since I think you are arguing that this is too big a penalty, your position seems confused to me. I can imagine no reasonable alternative to 30 lb = -1 speed. That should apply to weapons as well, of course.

[PowerDiver]

Yes, it might. It's also a lot more code for a tiny refinement of the algorithm

I don't understand this at all. Either define a constant or set a value of 1/2 rAcid after you do the other prep work. Then add to the power of all armor. That's only 2 lines of code.

[Zikke]

Where is this "randart.log"?