item identification tangent


Although the central place for design discussion is ##crawl-dev on freenode, some may find it helpful to discuss requests and suggestions here first.

Ziggurat Zagger

Posts: 3037

Joined: Sunday, 2nd January 2011, 02:06

Post Sunday, 23rd January 2011, 22:47

item identification tangent

So as to not pollute the other thread with a tangent, it seemed like a good idea to move the item identification tangent to a thread of its own.

I assert that there is a single 'best' strategy for item identification in most games of Crawl, assuming reasonable play. While there are small variants possible due to unusual RNG behavior or particular starting equipment, ultimately the process of item identification has minimal impact on gameplay at least as compared to other roguelikes.

Scrolls
For identification of scrolls, I assert that use of scrolls of identify are almost never worthwhile until the midgame, and in the early game you should read-id every case as soon as possible with routine precautions. To take them individually:

Curse Weapon: Risk can be trivially averted by switching to a butchering weapon before testing scrolls, and scrolls to correct the problem are common. The risk is that if you prefer to use a non-butchering weapon you will waste an enchant weapon scroll on a weapon you don't plan to use, or that you will need to spend part of a level with a substandard weapon before you correct the problem. This is the single most dangerous scroll, but in almost all cases it is not.

Curse Armor/Jewelry: Useless and mostly harmless. Worst case scenario is that you have to use up a Remove Curse scroll, which is extremely common. Correcting the problem can even wait indefinitely until you want to swap, since the cursed item is still fully functional.

Blinking, Teleport: Unambiguous identification, and important enough to make waiting to identify scrolls unviable. Waiting for an emergency to happen is a terrible time to start identifying scrolls hoping for a teleport. Identification wastes one scroll.

Identify, Enchant Armor, Recharging: These behave the same, and unless you're a gambler at heart you're going to waste a couple at random to identify them. Targeting an unidentified armor or wand will unambiguously identify them, but is also unlikely to be where you'd want them used in the long run. You cannot delay identifying identify, but the other two can wait until you have plenty of identify scrolls if you can sort them out from the other important scrolls.

Remove Curse, Fear, Vorpalize Weapon, Holy Word: These will frequently refuse to autoidentify. Vorpalize Weapon and Holy Word don't usually show up early on, so early on you're trying to distinguish two stacks that might actually be important. After wasting one of each of Remove Curse and Fear, you waste one more of either to distinguish them, either by finding a rat or by getting something cursed while identifying other scrolls.

Acquirement, Detect Curse, Amnesia, Fog, Enchant Weapon I-III, Silence, Summoning, Vulnerability, Torment: Harmless to read-identify, and in the worst case wastes a scroll.

Paper, Noise, Random Uselessness: Useless and harmless, but will waste three scrolls of identify if you're trying not to read-id scrolls for some reason.

Discounting scrolls that appear after the early game, you'd need seventeen scrolls of identify to sort out your scrolls without read-identifying them. You aren't going to spawn seventeen scrolls of identify in the early game no matter what the RNG says, and they're better used on other things anyway. Missing out on an enchant armor or an amnesia is mildly annoying, but you really don't have a viable alternative if you haven't identified teleport, blinking, detect curse, etc. You need those scrolls identified quickly, and while it's still safe to do so. In almost every case, the risk associated with read-identifying scrolls is the waste of a scroll, which is either that same type of scroll or a specific other type that serves as a remedy. The same gameplay could be duplicated by simply not spawning the first one or two samples of the appropriate scrolls.

Potions
Potions are a slightly more interesting case. Potion identification is to quaff-id them in a cleared area until you find the essential ones, and then save the rest until you have a glut of scrolls of identify in order to avoid mutation.

Healing, Heal Wounds: Absolutely essential to every character. Poison and confusion are threats starting from D1, and you absolutely cannot wait until Sigmund or an orc wizard has you confuse-locked to start quaff-testing potions. You cannot wait on these, and you will not have sufficient scrolls of identify at this point for that to help. You will waste one of each.

Speed, Might: Essential to some characters. If you're one of those characters, you keep quaff-testing until you find them. You will waste one of each.

Confusion, Decay, Paralysis, Poison, Strong Poison, Slowing: Mash 5 to correct the problem, or in the worst cases burn one healing potion.

Degeneration: Usually irrelevant, but in the worst case you might have to keep quaff-identifying until you find restore abilities.

Restore Ability: Until the late midgame, the only sources of stat damage are shadows, quasits, and self-inflicted sickness or degeneration. Shadows and quasits should be avoided if you're critically low on the stat in question, but otherwise you won't need to identify this potion unless you're trying to identify potions. Minor stat damage can simply be ignored.

Agility, Brilliance, Invisibility, Levitation, Magic, Porridge, Resistance: Nice to identify if you randomly happen upon them while quaff-identifying for other potions, but otherwise they can wait. Quaff-identifying wastes the one you had anyway.

Mutation: The Sword of Damocles hanging over your head. A bad mutation can kill you, and others make the game painful to play. This is the reason you stop quaff-identifying once you've found the four that are most important. If it didn't exist, potion identification would be meaningless.

Cure Mutation: Wasting one of these is actually a meaningful sacrifice, since they are rare and can be used to correct a dangerous handicap. Once mutation and cure mutation are out of the way, there is no reason not to quaff-id every other potion in the game.

Berserk Rage: Wastes the potion and some permafood. Very rare, so you're not likely to have any benefit from identifying it early.

Gain Stat, Experience: No reason not to use these as soon as possible, so the fact that they're not pre-identified is irrelevant. Saving Experience until just after you level up is a good way to start getting hit by cold attacks.

Blood, Water: Take care of themselves.

Other
Weapons, armors, and staves can be use-identified, and with the unusual exception of the distortion brand there is no risk involved.

Jewelry can sometimes be use-identified, and if it does not, then you burn a scroll of identify. It is not safe to wear unidentified jewelry, and if you have a choice between identifying an unknown bit of jewelry or identifying an unknown consumable, the jewelry is more likely to be immediately useful.

Conclusion
Item identification doesn't really do very much in the game. Wasting resources is only a threat to avoid if there's a viable alternative method that avoids the waste, otherwise it's simply an unavoidable toll. Crawl's punishments for overly ambitious item identification is intentionally minimal, pushing the point where you want to identify items to the earliest possible point. Waiting for a scroll or potion shop, or for a book containing Fulsome Distillation, is not a feasible alternative because they are relatively rare events and you need those essentials very early on.

As I see it, there is no problem to be corrected. Reducing item identification as a gameplay factor is an intentional design choice, and there is little incentive to modify it further. On the other hand, since the item identification subgame is intentionally minimized, it isn't a very persuasive argument against other proposed changes. Item identification is already largely irrelevant, and changes to it are likely to also be nearly irrelevant.

For this message the author KoboldLord has received thanks:
evktalo
User avatar

Dungeon Master

Posts: 4031

Joined: Thursday, 16th December 2010, 20:37

Location: France

Post Sunday, 23rd January 2011, 23:15

Re: item identification tangent

There is a little more to it. For example: I like to quaff-id potions on the upstairs of an explored level. So if I get confused or poison, I just wait it out (btw, waiting out decay might not be such a good idea), but if I get speed or might, I go downstairs and start exploring, so it's not a complete waste. Also, teleporting on an unexplored level can be dangerous. On the other hand, reading magic mapping on a fully explored level not only waste the scroll, but does not identify it. That's why I like to read-id scrolls near the upstairs of an unexplored level, preferably with monsters nearby in case of fear.
Blood as the unique property of self-identifying if you carry it for a while.
Of course, you can dismiss all this as being minimal and irrelevant, but I just disagree. It's a big part of the very early game, and it has some strategy to it. And I'm sure other players have different ones. Some like to gamble with mutation for example. And when you do quaff mutation, sometimes you get good ones, and then stop id-ing in fear of cure mutation.
I think Crawl has a really good balance for item identification. If we add things that make it irrelevant, then it gets boring and uninteresting and everyone will want it to be removed. But Crawl would lose something in the process.
<+Grunt> You dereference an invalid pointer! Ouch! That really hurt! The game dies...

For this message the author galehar has received thanks:
evktalo

Dungeon Master

Posts: 1613

Joined: Thursday, 16th December 2010, 21:54

Post Sunday, 23rd January 2011, 23:45

Re: item identification tangent

galehar wrote:On the other hand, reading magic mapping on a fully explored level not only waste the scroll, but does not identify it.

It does, actually ("You feel momentarily disoriented.")

Dungeon Master

Posts: 3618

Joined: Thursday, 23rd December 2010, 12:43

Post Monday, 24th January 2011, 00:03

Re: item idenfication tangent

It might well be that the OP's analysis is correct and the indicated behaviour is optimal, but I play it differently.

First, how much I need to rely on use-id depends on
  • general safety (the higher, the less I need to use items for id purposes)
  • access to identify scrolls (if I have sufficiently many, I can spend them on consumables)
  • shops (buy-id is much superior to other forms)
Then there are special cases: if I have found a good wand (or armour item), I am more inclined to read-id. I believe that scrolls of identify are best used on jewellery, so if I find one (which does not auto-id) early, I am more inclined to read scrolls.

Of course you can say that your approach is clearly better... but my conservative approach to identification has saved a scroll of blinking or recharging so often for me.

As a designer, my dearest hope is that the answer to a question like "how should I undertake identification" is invariably "it depends". Therefore, what I say might be wishful thinking, but I don't believe matters are as clear-cut as you present them here.

Thanks for the careful elaboration in any case.

Ziggurat Zagger

Posts: 3037

Joined: Sunday, 2nd January 2011, 02:06

Post Monday, 24th January 2011, 03:51

Re: item identification tangent

galehar wrote:There is a little more to it. For example: I like to quaff-id potions on the upstairs of an explored level.


Oh, this is a good point. Now that you mention it, I do the same thing. It's always nice to get a little benefit off of that wasted speed or brilliance potion, even if it only lasts a room or two on the next level.

galehar wrote:Also, teleporting on an unexplored level can be dangerous.


This is a bit of an understatement! I wouldn't dare risk read-identifying scrolls on an unexplored level if teleportation and noise are still unknown.

galehar wrote:On the other hand, reading magic mapping on a fully explored level not only waste the scroll, but does not identify it.


I forgot magic mapping, didn't I? I'm sorry I passed over it while I was reorganizing them into categories. As Marvin said, magic mapping does autoidentify even if it's wasted uselessly. I've tried it in both 7.0 and in trunk. I tend to keep one magic mapping scroll in my inventory for labyrinth or bazaar levels, and the whatever doesn't get used up or destroyed by fire attacks gets used in Slime or Zot. I regard them as too valuable and delicate to carry around in multiples, but using up the last one in hand merits an immediate return to my stash for another. Basically like scrolls of blinking.

It's painful to give up that first scroll, but one identified scroll in hand is worth more than two unidentified ones sitting in a stash.

galehar wrote:Some like to gamble with mutation for example.


I'm reasonably confident that this is Fountain Quaffing behavior. Bad mutations like berserkitis or teleportitis virtually end your game right there unless you have a countermeasure on-hand, and good mutations are limited to a single pip of a useful but ultimately replaceable resistance. This is nowhere near a fair bet.

The calculus changes, of course, if you've got some insurance. For instance, if you've diligently used Fulsome Distillation on every mutagenic corpse in the entire game and now you're preparing to raid Zot with a stack of 120 potions of mutation. It also amuses the crowd if you're doing a Let's Play thread. This aren't really the game situations I'm talking about, though.

Do any of the players who streak games in the annual tournaments intentionally quaff potions of mutation in the early game? I'd bet against it, but I don't actually know.

dpeg wrote:First, how much I need to rely on use-id depends on

* general safety (the higher, the less I need to use items for id purposes)
* access to identify scrolls (if I have sufficiently many, I can spend them on consumables)


Yes, I would agree that an arbitrary case is weighed mostly on the basis of these two interests. Optimal play is to use-id items in safety until you have what you need to ensure general safety going forward, and stash the rest until identify scrolls are available. Eventually those identify scrolls will be available, and the question is whether you can safely afford to wait.

dpeg wrote: * shops (buy-id is much superior to other forms)


I agree that shop identification is superior to all other forms when you are lucky enough to get access to it, but shops are very rare and shops specifically devoted to consumables are even more rare. If there was a guaranteed bazaar around the same band of levels that can spawn ossuaries or sewers, it might be worth waiting to identify consumables there, but with the urgency of item identification for most characters it seems like waiting for a windfall is unreasonable.

Not that I'm suggesting we should have a guaranteed bazaar at those levels. I haven't really thought through the implications, but it would probably have to be a specially nerfed bazaar so it doesn't have only overpowered or unaffordable wares.

dpeg wrote:Thanks for the careful elaboration in any case.


Thank you for listening.

Return to Game Design Discussion

Who is online

Users browsing this forum: No registered users and 130 guests

Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group.
Designed by ST Software for PTF.