The Acquirement lottery

[Timo Pietilä]

Originally Posted by Estie

Problem with giant pits is the storm giants. Even if you kill them fast, meleeing a pit with many of those means your staves and wands and low rods will be gone. Dropping all in a corner before opening the pit works, but is still very risky and leaves you with fewer options if something happens while you kill the pit.

I have turned to ignoring giant pits because of the nasty light blue Ps.

Staves are not affected by electricity. Rods and wands are. Staves are destroyed by fire, so fire giants can do that.

You don't want to be in LoS of many at once, because they cast balls of lightning. That destroys rings, rods and wands at the floor, and rings especially can be useful. If you have weapon that can kill them quickly you might want to fight them around a corner so that not many of them have LoS to you at once.

[Magnate]

Originally Posted by Timo Pietilä

There are giant pits even deep in dungeon. They have titans inside and they drop good stuff. Slay giant is useful for emptying those and getting the loot which is important.

Obviously acid/fire/cold/electricity brand is better than that (in that order against giant-pits).

Ogres are also counted as giants, so Lokkak is giant. Then there are deep giant uniques that hit hard, and those resist everything (or nearly everything)

I'd say slay troll is less important than slay giant, because any weapon with enough blows and to_dam can kill normal trolls in single turn, but same is not true against biggest giants. Same applies to slay orc, any weapon with enough blows is weapon of slay orc.

Hmm - I was trying to consider an "average" value, across all possible scenarios of finding them. Finding slay orc or slay troll early makes a HUGE difference to your game - lots of kills means lots of loot and xp. It's essentially a stepping stone to your first artifact (or whatever). You are absolutely right that later in the game both of those have negligible value while slay giant remains quite useful (albeit as a swap - it's almost certain that you'd have a branded or artifact weapon by then).

[Magnate]

Originally Posted by PowerDiver

Perhaps we use the same word with a different meaning? Just because you use something 1 game in 10 in some unimportant way does not make the item useful.

I occasionally use iron spikes [thrown for 1d1 damage], perhaps 1 game in 10, to save on my iron shots. I doubt it has ever materially affected any game. If I had a means of setting initial squelch automatically [no per-game keypresses] I'd put iron spikes on the list. That means iron spikes are junk AFAIAC.

Your logic continues to fascinate me. I agree that something one would choose to squelch from the start of every single game is a good definition of junk - one might as well remove it from object.txt so it cannot be generated. That's exactly what takk did with broken sticks and shards of pottery in 3.0.something

But something I use one game in ten is definitely not junk. Imagine, if you took all those things out of object/ego/artifact.txt - how little you would have left, and how boring the game would be! Everything you find guaranteed to be useful every time - this is angband, not chess.

Ultimately it's subjective - Timo would like broken sticks and shards of pottery re-added for flavour, while you would remove iron spikes and probably quite a lot more besides. I'm somewhere in between.

I still get angry when I think back to how much criticism I got from various veteran players when I dared to tweak a number of long-unchanged standarts to make them more useful more often. I destroyed game balance, started an arms race, etc. etc. It's not perfect by any means, but now we see Til-i-arc and Gondricam and other artifacts being discussed in threads and dumps when they would previously have been tossed without a thought.

Ah well, never mind. Don't feel you have to reply to that last bit, it was a general broadcast. There isn't a right answer to this stuff - there's no perfect way to balance objects, monsters, spells and everything else. All we can do is adjust a bit at a time and try not to overcompensate. But I do think removing stuff needs to be considered extremely carefully - I still think takk was wrong to remove the brands from the elemental rings, as they are now pretty much junk again (though that could be addressed by making them shallower), and there were other solutions to the brands being overpowered - I'd have made them x2 brands to start with, and watched to see if they then needed further adjustment. (This sub-thread was originally about whether to remove the less powerful slays like orc/troll/giant, which is why I got back to the topic of removing stuff. Definite no from me.)

YMMV; can't please all the people all the time. Etc.

[PowerDiver]

Originally Posted by Magnate

Your logic continues to fascinate me. I agree that something one would choose to squelch from the start of every single game is a good definition of junk - one might as well remove it from object.txt so it cannot be generated. That's exactly what takk did with broken sticks and shards of pottery in 3.0.something

But something I use one game in ten is definitely not junk. Imagine, if you took all those things out of object/ego/artifact.txt - how little you would have left, and how boring the game would be! Everything you find guaranteed to be useful every time - this is angband, not chess.

There is a difference between using an item, because it is there and it doesn't hurt to use it, and using an item because it makes a difference. Lets be very lenient about the def of making a difference. Making a difference might mean killing a single monster a single turn earlier, now or even later when no longer carrying the object because it saved a consumable when you had it. A dagger (+4,+3) generally makes a difference when you previously were using a dagger (+3,+2).

I use the spikes 1 game in 10, but I'd guess they makes a difference 1 game in a million. It is the latter number that matters, not the former.

I do not remember the last time I used a weapon of slay troll. I really don't. I have used *slay giant, not for the extra power, but for the slay, 1 game in recent memory when it showed up on a MoD I tossed and then went back for when I encountered a giant pit with multiple cyclopses and had no elec branded attacks. I don't know whether this counts as evidence that normal slay giant has any value. If the normal slay giant ego is phased out for *slay giant before the native depth of MoD, then this example doesn't count because it would be too unlikely for me to use a normal slay giant MoD.

[Max Stats]

Originally Posted by PowerDiver

I occasionally use iron spikes [thrown for 1d1 damage], perhaps 1 game in 10, to save on my iron shots. I doubt it has ever materially affected any game. If I had a means of setting initial squelch automatically [no per-game keypresses] I'd put iron spikes on the list. That means iron spikes are junk AFAIAC.

Do you compile your own game? I got annoyed at squelching the same items over and over with every new character, and I figured out a way to hardcode certain items that I always want to squelch, but it requires compiling from source.

You can do this by opening squelch.c and editing squelch_birth_init. It starts out like this:

        /* Reset squelch bits */
        for (i = 0; i < z_info->k_max; i++)
                k_info[i].squelch = FALSE;

You need to change it to look like this:

	/* Reset squelch bits */
	for (i = 0; i < z_info->k_max; i++) {
		switch (k_info[i].kidx) {
			case nnn:
			case nnn:
			<insert as many more as necessary>
			case nnn:
				k_info[i].squelch = TRUE;
				break;
			default:
				k_info[i].squelch = FALSE;
		}
	}

Each instance of "nnn" needs to be replaced by the serial number of an object you want to auto-squelch, which is taken from the second field of the 'N:' lines in object.txt.

Of course, this is not an ideal solution, but an intermediate measure until (hopefully) the ideal solution is implemented; i.e., the ability to put squelch settings into a prf file and have them applied to each newly created character. That way, you can have different default squelch settings for each character class; this method applies the same settings to every new character regardless of race, class, etc.

Oh, and for good measure, you might as well go ahead and change the next few lines from this:

        /* Clear the squelch bytes */
        for (i = 0; i < TYPE_MAX; i++)
                squelch_level[i] = 0;

to this:

        /* Clear the squelch bytes */
        for (i = 0; i < TYPE_MAX; i++)
                squelch_level[i] = 1;

This changes the default quality squelch setting from "no squelch" to "bad" so you can automatically get rid of all the negative equipment you find. Of course, if you are playing the nightlies, it seems that there is no more negative equipment; everything is at least +0, so this change is mostly moot (unless you get something damaged).

[Magnate]

Originally Posted by PowerDiver

There is a difference between using an item, because it is there and it doesn't hurt to use it, and using an item because it makes a difference. Lets be very lenient about the def of making a difference. Making a difference might mean killing a single monster a single turn earlier, now or even later when no longer carrying the object because it saved a consumable when you had it. A dagger (+4,+3) generally makes a difference when you previously were using a dagger (+3,+2).

I use the spikes 1 game in 10, but I'd guess they makes a difference 1 game in a million. It is the latter number that matters, not the former.

Agreed, with the caveat that this is strictly a discussion of utility, and does not consider the flavour dimension at all.

I do not remember the last time I used a weapon of slay troll. I really don't. I have used *slay giant, not for the extra power, but for the slay, 1 game in recent memory when it showed up on a MoD I tossed and then went back for when I encountered a giant pit with multiple cyclopses and had no elec branded attacks. I don't know whether this counts as evidence that normal slay giant has any value. If the normal slay giant ego is phased out for *slay giant before the native depth of MoD, then this example doesn't count because it would be too unlikely for me to use a normal slay giant MoD.

I would suggest that this doesn't actually mean that either slay troll or slay giant are not useful, but that they are not turning up at the right times. (This is why I digressed in my last reply to the standart changes - but I forgot to join it up properly.) It would seem that these slays need to appear more often earlier on in the game, when they're more likely to make a difference, and less often later on - this is ticket #1080.

You don't cite slay orc, so presumably you find slay orc early enough for it to make a difference in a proportion of games which you consider reasonable. Is there any reason why adjusting the generation of slay troll and slay giant couldn't achieve the same for them? The latter is harder, I'll admit, because giants tend to come later on, but trolls have a definite "phase" in the dungeon, like orcs.

[Derakon]

How about replacing Rings of Resist Fire/Cold with the elemental rings and having them show up around about 800-1000'? You might want to tone down the damaging portion of the activation a bit, and maybe the amount of extra AC given, but then they'd be showing up early enough that they might actually get used.

Once variable-strength slays are available it'll be easy enough to re-create ring-slot elemental brands and see how weak they have to be to be balanced.

[Magnate]

Originally Posted by Derakon

How about replacing Rings of Resist Fire/Cold with the elemental rings and having them show up around about 800-1000'? You might want to tone down the damaging portion of the activation a bit, and maybe the amount of extra AC given, but then they'd be showing up early enough that they might actually get used.

Using Eddie's definition of "make a difference" in his other reply, I have never ever known the attack activation of an elemental ring to make a difference. It's nice flavour, but I only ever use the activation for the temp resist. But yes, I agree that making them turn up earlier would be what made the difference.

Once variable-strength slays are available it'll be easy enough to re-create ring-slot elemental brands and see how weak they have to be to be balanced.

Adding a new slay or brand is now only two lines: one extra line in src/object/list-slays.h, and one in src/object/list-object-flags.h. I would not feel comfortable adding any back to the elemental rings in direct contravention of takk's changes though (the wikipedia term "revert war" comes to mind!). He made a decision to address a vexing issue, and I respect his leadership even if I don't like the decision. (If I can persuade him to try x2 brands, that's another matter ...)

EDIT: I just added five x2 brands - no point saying how easy it is and then never getting around to it. I changed Paurnimmen to use the new x2 cold brand, as it's the only remaining off-weapon brand in the game (excluding randarts). When I rewrite randart.c, off-weapon brands will normally be x2, not x3.

[PowerDiver]

Originally Posted by Magnate

You don't cite slay orc, so presumably you find slay orc early enough for it to make a difference in a proportion of games which you consider reasonable. Is there any reason why adjusting the generation of slay troll and slay giant couldn't achieve the same for them? The latter is harder, I'll admit, because giants tend to come later on, but trolls have a definite "phase" in the dungeon, like orcs.

First off, let me say that I have no problem with junk. I want more junk, not less. I do not like the reduction in weapon drops in the early game. One person's junk may not be another's. My problem is the need to identify before squelching junk. That has always has been the problem IMO.

As to slay orc, the point is that orcs come earlier. Once you have a branded weapon, there is less value in a slay that affects monsters with no resists. You are more likely to have a branded weapon by the time you encounter trolls in numbers because they are deeper. Another point is that it has to be on a max-swing weapon. You need to get the slay when most weapons are at most 5 lbs. Slay orc or troll on a long sword doesn't make up for losing a blow. I don't use slay orc much, but I have vague recollections of using it on a spear. I wouldn't argue with someone who called it junk, but I wasn't sufficiently sure that I wanted to say so myself.

I don't see how to make the slays very useful, except for giving them monster-specific ESP. Even then, I doubt I would carry one once I was using a branded weapon.

If you want to make these slays more valuable, I don't think it will help much to make them shallower. Slay troll even at DL1 is still not particularly useful if you find a branded weapon before you encounter your first score of trolls. Instead, you need to make artifacts and other egos sufficiently deep that you don't find any before you start killing trolls in numbers. Every time you change an artifact weapon to make it appear earlier, you further marginalize weapons of slay troll.

The nature of these problems, where you sample from a distribution and count the number of times you improve the best seen so far, is that you really can't change the total number of useful items much at all unless you crush the variance. The only practical way to make slay troll more useful, useful defined as you might carry it and use it, is to make other things comparatively less useful, probably by making them deeper so you are less likely to use them when you find them.

[Derakon]

I've sort of been gently agitating for pushing back the good gear anyway. The 1000'-2000' region gets a lot more interesting when good gear is scarce; in standart games that's usually when you start finding a whole bunch of bread-and-butter artifacts that drastically increase your power level. Delaying that means taking on the monsters in that part of the game with egos instead.