reduce artifact drop rate

[Magnate]

It's about too many artifacts total. However, the amount of egos is not bad, so it can also be approached by changing how many ego items turn into artifacts. I understand that this is grandfathered in and there is reluctance to change it.

There is a very limited area between the early game and the mid game where the choice of useful weapons are very high. These include the slays and the branding weapons. Lately, I find a superior artifact before I find any of these weapons. Slay and branding weapons become solely a source of money. I think that about 75% of games you should be using one of these lesser egos at some point, even if it's only for 3 levels. Have other people noticed this also? Maybe my games were outliers.

Standart and randart generation rates should be the same. They should both be less.

As a side, there's no problem with special artifacts. I haven't found the palantir in several games, and that's a good thing in my mind.

Well, it is interesting to consider a change to how often ego items become artifacts. Personally I am not yet convinced that that is the problem, despite both your and Ewert's views. My own games still generate hundreds of ego items, most of which are junk. I think ego item generation should be turned down a bit first, before deciding whether or not to restrict artifact generation. But it would be helpful to get other views on this as well.

I agree with your point that @ should be using ego items in the early-to-mid game before finding artifacts. But the flip side of this is that many artifacts were junk when found, unable to compete with a decent Westernesse / HA / Gondolin. Clearly the balance is not quite right.

[fizzix]

I agree with your point that @ should be using ego items in the early-to-mid game before finding artifacts. But the flip side of this is that many artifacts were junk when found, unable to compete with a decent Westernesse / HA / Gondolin. Clearly the balance is not quite right.

Westernesse is often junk after the early game. Slay orc and troll are useless. HA and Gondolin are better, but they're pretty rare. I'd say I find my first of those high level egos after I've found about 5-6 artifact weapons. I can't remember the last game I wielded a westernesse weapon.

There certainly are a lot of things to do with ego creation. If TMJ is the issue, the quickest solution is to add more gradations in to weapon squelching. The fact that there's no setting to squelch a rapier of venom without squelching a gondlin with telep is a problem. Working from this side is how I'd prefer to deal with too many egos. (Don't get me started with not being able to squelch magical body armor without squelching magical MHDSM...ugh) But you can also make those lesser slays and branding egos only be created between dlevel 15-40 or so. Not that anything else is created in their place, the game just automatically squelches them. Say the monster carrying it realized it was junk and destroyed it.

There are lots of things to try. But, I guess other people need to agree that things aren't good the way they are to want to try them.

[cinereaste]

I like the suggestions that Magnate made about reducing the number of ego drops. In my last game there was a ton of ego junk in the middle levels, which mainly got sold in town. Reducing the number of ego drops would fix both the excess of ego weapons, but also slow down the artifact drop rate somewhat.

I also like fizzix's suggestion regarding squelching. A finer control over quality squelching would be great.

[Magnate]

I like the suggestions that Magnate made about reducing the number of ego drops. In my last game there was a ton of ego junk in the middle levels, which mainly got sold in town. Reducing the number of ego drops would fix both the excess of ego weapons, but also slow down the artifact drop rate somewhat.

I also like fizzix's suggestion regarding squelching. A finer control over quality squelching would be great.

Again, these are two very different issues. Squelching will ultimately be fine-grained but this is quite a lot of work, as nobody has quite worked out the desired UI yet (we just know it ought to be done through the knowledge menu). Adjusting ego and artifact generation is easier to start but harder to get right - but I do intend to move ego types to the alloc_prob framework, which will allow us to set max depths for them (i.e. no more slay orc weapons after dl25).

[fizzix]

having no more slay orc weapons deep would be a big help. However, I'd much prefer a system that instead of not allowing that item to be created, it destroys that items immediately after creation (giving no drop instead). Uniques can have a guaranteed drop if you want. I don't think any other monster should. This is probably very controversial and I doubt anyone will agree with me.

[krugar]

There's also some balance issues with ego items drops, I think. I can sympathize with the idea of some extraordinary items to have a very small chance of dropping very early in the game. But this bow dropping in 150' (L3) is perhaps a little too much. And right into the hands of a ranger...

  [Angband 3.1.2v2 Character Dump]

 Name   Rahtgaz III                              Self  RB  CB  EB   Best
 Sex    Male         Age            117   STR:     18  +1  +2  +0  18/30
 Race   High-Elf     Height          95   INT:     15  +3  +2  +0  18/20
 Class  Ranger       Weight         167   WIS:      9  -1  +0  +0      8
 Title  Strider      Social     Unknown   DEX:     16  +3  +1  +0  18/20
 HP     77/77        Maximize         Y   CON:     15  +1  +1  +0     17
 SP     14/14                             CHR:     10  +5  +1  +0     16

 Level               10   Armor    [15,+29]     Saving Throw         59%
 Cur Exp            893   Fight     (+4,+3)     Stealth        Excellent
 Max Exp            893   Melee     (+7,+8)     Fighting       Excellent
 Adv Exp           1150   Shoot   (+14,+12)     Shooting          Superb
 MaxDepth     300' (L6)   Blows      3/turn     Disarming            44%
 Game Turns      293652   Shots      1/turn     Magic Device          65
 Player Turns     22744   Infra       40 ft     Perception       1 in 20
 Active Turns     21371   Speed      Normal     Searching            27%
 Gold               166   Burden   89.6 lbs

 You are one of several children of a Telerin Ranger.  You have light
 grey eyes, straight black hair, and a fair complexion.

/* ... */
     
b) a Long Bow of Extra Might (x4) (+10,+12) (+1)
     +1 shooting power.

And it's not a monster drop mind you. Could be a little more understandable if it was. But this one was lying on the floor.

[Derakon]

One of the beauties of Angband is that fantastic things can happen sometimes. As long as this doesn't happen remotely regularly, there is no problem with this.

Well. There is the problem that the game won't let terrible things happen to you to compensate -- no Mature Blue Dragons on dlvl 3. But that's a separate issue.

[fizzix]

One of the beauties of Angband is that fantastic things can happen sometimes. As long as this doesn't happen remotely regularly, there is no problem with this.

Well. There is the problem that the game won't let terrible things happen to you to compensate -- no Mature Blue Dragons on dlvl 3. But that's a separate issue.

I don't have any problem with the double standard. The nature of permadeath requires it.

The bow drop is strong but not overwhelming. I would say this is about equal to finding a westernesse dagger on the ground there. Something that should occur once ever 50-100 games. It's a lot easier to judge (not necessarily fix) the artifact drop rates. This is because the complete set of artifacts is known and you can see what total percentage of them have dropped.

[Magnate]

having no more slay orc weapons deep would be a big help. However, I'd much prefer a system that instead of not allowing that item to be created, it destroys that items immediately after creation (giving no drop instead). Uniques can have a guaranteed drop if you want. I don't think any other monster should. This is probably very controversial and I doubt anyone will agree with me.

What is the difference between your proposal, and squelching?

[Derakon]

What is the difference between your proposal, and squelching?

You still have to personally determine the quality of the item before squelching it.

[Magnate]

You still have to personally determine the quality of the item before squelching it.

Ok, so fizzix's proposal is hard-coded, automated squelching, over which the player has no control? What is the difference between that and preventing generation in the first place?

[cinereaste]

Ok, so fizzix's proposal is hard-coded, automated squelching, over which the player has no control? What is the difference between that and preventing generation in the first place?

Correct me if I'm wrong, fizzix, but as I read it the difference is that it still takes up a drop from a monster. Say I kill something that will drop two things. If we are preventing (for example) weapons of slay orc from being ever being generated at this DL, then the monster will drop two things that are not weapons of slay orc, but I think fizzix is proposing that we don't change the generation process, we just filter the drops between when they are generated and when they actually appear on the floor. In my example if a trident of slay orc and a !speed were generated, only the !speed would ever appear. This has the result of reducing the number of drops from monsters.

[ewert]

That makes no sense versus just having a good squelch system, tbh ... just removes options from player. Either regenerate items until enough items are dropped (in effect increasing drops), or make a good squelch system (selectable by ego type for example). IMHO, and only if you want to go this route. Personally I wouldn't ...

[fizzix]

That makes no sense versus just having a good squelch system, tbh ... just removes options from player. Either regenerate items until enough items are dropped (in effect increasing drops), or make a good squelch system (selectable by ego type for example). IMHO, and only if you want to go this route. Personally I wouldn't ...

Cinereaste has it right. The reasoning is the following:

1. You don't want drops to be too powerful because you are automatically squelching weaker egos.

2. You eliminate the need for identification prior to squelching. Squelching egos still requires you to pick up (if squelching all non-artifacts) or wield (if squelching all non-splendid) or even identify (Robe of permanence, Shield of preservation).

The trick is to come up with the break points where there is a near guarantee that the weapon is useless. I'd actually do it by the item power algorithm, so that it can keep MoD of slay evil or BoC of acid. Weapons that could conceivably be useful for the endgame.

This is kind of getting off topic from the original point, so I'll try it back together with a summary of my views. Point 1 is the most important and I would like to know if people agree. Anyway:

1. There are too many artifacts generated
2: One solution is to reduce the number of egos that get changed into artifacts
3: Egos often wind up being junk, and increasing the percentage of egos to artifacts lying around means more junk.
4: Improved squelch features will help and these will eventually be put into the game when the UI is figured out.
5: However, squelching does not eliminate the need to pick up, wield, or identify many items to determine whether to squelch them.
6: Forcing the game to hard squelch items it deems worthless at some dlevel is a possible (although contentious) solution to the problems of 3 and 5.
7: (new) changing the ego -> artifact probability can be tied into the drop type. Floor, Vault, normal monster, unique monster, acquirement scroll. This can tailor the percentage of artifact drops that come from each source.

[Derakon]

If you were to do this kind of thing, my take on it would be that it'd be an adjunct to the current squelch system: "never show me even un-ID'd items that are worse than this quality level". That is of course open to some abuse, though, since it means that every item you see must be notable at some level even if you don't know what it is yet.

Note that the previous increase in general drop quality at the cost of drop quantity was basically an implementation of this. We don't see anywhere near as many items as we used to, but each item is much more likely to be interesting.