GOOD and GREAT

[CJNyfalt]

Nor do they address the issue that having plusses to hit/dam/AC doesn't really make most weapons or armours any better than +0, except in the very early part of the dungeon before ego items begin to appear. For the vast majority of the game the definition of "good" is wrong.

So I would add to fizzix's proposals the following:

(i) Redefine "good" to include low-end egos for weapons and armour (single resist, single non-* slay, etc.), plus DSM/BoC et al., and also to include a set of high-end consumables (staves of destruction, stat potions etc.), along with the existing jewelry and early dungeon books

Well, as an unpolished idea: How about doing away with the plusses totally? Or at least remove the enchanting scrolls? As I see it, the problem isn't artifacts too early or too weak artifacts or too weak or strong ego items. It is that the so called 'good' weapons and armor are boring.

It would be nice to have something like this:
- All weapon and armor gets a durability rating
- Scrolls of enchanting becomes scrolls of repair
- Acid (and other damaging attacks) hits durability - not pluses
- Pluses are only generated on ego items
- Damaged armor and weapons gets penalties.

[Magnate]

1) I think artifacts need to be spread out more. They're currently designed to span the entire dungeon level. Some weapons, like the thancs, are weaker than most egos. Restricting them to level 75+ is a little silly. I think artifacts should be viewed as an upgrade to an ego item you might otherwise see at that level. I think artifacts should be determined separately. They should have their own, somewhat independent calculation. It also needs to accommodate failure. You can't say "this should be an artifact" up front, because there are a limited number of them.

I agree with this, but I think it's hard to get right. The 'thancs were beefed up in 3.1.x because they were so often considered junk when found - if we can get cleverer about when they're found, we can weaken them again and undo some of the power inflation. But it's got to be something more than just sticking max depths onto the existing algorithm.

2) It makes more sense to drop a slay dagger on dlevel 5 than it does to drop an average executioner's sword. The sword is junk at that point. Similarly plate mail is problematic because of encumberance reasons. It's really hard to get around this point. I don't think the weapon base drops are broken, except perhaps that DSMs are a bit too common. I'm not sure if you meant to change that, I might be reading too much in.

You are right that this is really hard, because not all base items are equally useful, and not all base items are equally boosted in utility by the same plusses or ego template. I do think the base weapon drops are broken, because +0 blades of chaos ought to start dropping around dl35-40 or so, and they don't. In fact I think I've never seen one, because by the time a BoC drops it's almost always good or great.

pseudo-code

Roll_for_good

if (GOOD) roll_for_ego (try for slays)

if (EGO) roll_for_goodego (try for brands, *slays, blessed)

if (GOODEGO) roll for greatego (try for gondolin, westernesse, HA, extra attacks)

The problem with this, as you said yourself, is that it's the hard-coding that you argued against. Each ego type would need to be hard-coded as ego/goodego/greatego. You said

I think I still prefer an approach where the only information needed is a level value

- do you simply mean a minimum dungeon level where it can start dropping in-depth, or do you mean something more like a power rating?

[fizzix]

I agree with this, but I think it's hard to get right. The 'thancs were beefed up in 3.1.x because they were so often considered junk when found - if we can get cleverer about when they're found, we can weaken them again and undo some of the power inflation. But it's got to be something more than just sticking max depths onto the existing algorithm.

I would prefer a system where you find a thanc by clearing dlevel 10 once out of every 5-10 games. I think bullroarer and wormtongue should probably have something like a 5% chance of dropping a thanc each. I dont' know how to do this, but this is my intuition. I have no problem with most games not finding a thanc until its useless. In fact, any of us that do a lot of diving probably shouldn't find it.

You are right that this is really hard, because not all base items are equally useful, and not all base items are equally boosted in utility by the same plusses or ego template. I do think the base weapon drops are broken, because +0 blades of chaos ought to start dropping around dl35-40 or so, and they don't. In fact I think I've never seen one, because by the time a BoC drops it's almost always good or great.

I've seen one or two in my day. They're useless. What's the point? Often I don't even have enough strength to wield it properly. BoCs are too heavy to be useful until strength is near max.

The problem with this, as you said yourself, is that it's the hard-coding that you argued against. Each ego type would need to be hard-coded as ego/goodego/greatego. You said - do you simply mean a minimum dungeon level where it can start dropping in-depth, or do you mean something more like a power rating?

Power rating is probably ok, it already exists and, even though it's probably impossible to rate item goodness on a single axis, at least it's something. However, I *think* that you can get by only with a little extra cleverness on using info contained in min level. GOOD items exclude ones that have a suitably early min level. GREAT items exlude ones with a higher min level. For example a GOOD item may require a min level above MIN(30,level+10) and a GREAT item may require a min level above MIN(50,level+20). Now 30 and 50 are somewhat arbitrary, but let's look at what items would be droppable for GOOD and GREAT, levels are in parens, max level include when less than 100. Obviously there are some problems, specifically with rods, that seem to appear about twice as deep as they should for all but mapping, detection, identify, recall and speed.

GOOD
rings:
-------
Stat rings (30)
body/soulkeeping (30)
resist poison (40)
branding rings (50)
speed (75
see invisible (30)
delving (30)

AMULETS:
-----------
Regen (30)
ESP (60)
Res (60)
Sus (60)
Dev (70)
Weaponmastery (70)
Trickery (70)
magi (70)

scrolls:
-------
Prot from evil (30)
Dispel Undead (40)
Banishment (40)
Recharging (40)
*destruction* (40)
Mass Banishment (50)
*enchant* (50)
*remove curse* (50)
curse weapon (50)
curse armor (50)
holy prayer (50)
*Acquirement* (60)
Rune of Protection (60)

potions:
---------
Stat gain (-chr) 30
healing (30)
RLL (30)
dragon breath (40)
Aug (40)
*heal* (40)
Exp (65)
*enlight* (70)

wands:
-------
fire bolts (30-60)
acid bolts (30-60)
lightning balls (35)
Cold balls (40)
fire balls (50)
acid balls (50)
dragon's flame (50)
dragon's frost (50)
drain life (50)
dragon's breath (60)
annihilation (60)

staves:
-------
earthquake (40)
speed (40)
remove curse (40)
*destruction* (50)
dispel evil (50)
heal (70)
banishment (70)
power (70)
holiness (70)
magi (70)

rods:
------
slow monster (30)
sleep monster (30)
recall (30)
fire bolts (30-80)
polymorph (35-80)
acid bolts (40-80)
probing (35)
teleport other (45)
detection (50)
identify (50)
lightning balls (55)
cold balls (60)
mappig (65)
acid balls (70)
fire balls (75)
drain life (75)
healing (80)
restoration (80)
speed (80)

[Magnate]

I would prefer a system where you find a thanc by clearing dlevel 10 once out of every 5-10 games. I think bullroarer and wormtongue should probably have something like a 5% chance of dropping a thanc each. I dont' know how to do this, but this is my intuition. I have no problem with most games not finding a thanc until its useless. In fact, any of us that do a lot of diving probably shouldn't find it.

Takkaria's proposed "drop profiles" would allow us to do this pretty easily, so we could experiment.

I've seen one or two in my day. They're useless. What's the point? Often I don't even have enough strength to wield it properly. BoCs are too heavy to be useful until strength is near max.

Well, I was coming at this from the idea that more stuff ought to be used at +0 before enchanted stuff is found. Admittedly this doesn't have to apply to everything though. But your point makes me bristle with irritation again at how combat is so much about multiple blows and +dam. Makes me want to postpone item generation until that's fixed ...

However, I *think* that you can get by only with a little extra cleverness on using info contained in min level. GOOD items exclude ones that have a suitably early min level. GREAT items exlude ones with a higher min level. For example a GOOD item may require a min level above MIN(30,level+10) and a GREAT item may require a min level above MIN(50,level+20). Now 30 and 50 are somewhat arbitrary, but let's look at what items would be droppable for GOOD and GREAT, levels are in parens, max level include when less than 100. Obviously there are some problems, specifically with rods, that seem to appear about twice as deep as they should for all but mapping, detection, identify, recall and speed.

Hmmm. I don't think I like the general principle that the value of an item is always proportional to its min depth. But there's no particularly good case to demonstrate why it shouldn't be, so I'd be happy to give this a try.