Identification game (decks, wands, and potions)

[crate]

Secretly wands use randomly 2 or 3 charges if you zap them while they're not fully identified, and the game lies to you about this by always telling you the wand was zapped twice per use. I still don't understand what the point of this change was, since it doesn't actually do what the commit message claims it was trying to do (you in the past already identified hw/healing/usually tele, so it doesn't affect those wands at all, and the wimpy wands got more charges, so it's mainly just a nerf to fire/cold/etc which wasn't the claimed intention) and since it outright lies to players it's not improving clarity in any conceivable way.

The rest are alterations to carryovers from Linley's original design that should make sense if you think about it for a bit.

[dpeg]

crate: The idea is that you consider read-identifying an early wand. It does happen in my games. Ideally, whether to do this should depend on circumstances (what kind of wand, what's your current threat level, numbers of identify scrolls).

I have the impression your comments on clarity and nerfing try to deliberately the point about decision making.

[crate]

The only change in wand-id is you use id scrolls on fire wands or something before lair occasionally. Does it happen? Yes, but the drawbacks are way bigger than any benefit from the change.

Of course, there's also this to consider: why did this change happen after several different changes that specifically reduced the use of ID scrolls? A comparable change would be to make rings only work half the time unless you ID scroll them. That didn't happen; instead rings just auto-ID now. Why did wands go the other way? I mean it's clear in the past that there is no overarching design goal for a lot of these changes, but this one really bothers me. Crawl should never be deliberately trying to mislead unspoiled players like by telling them wands are zapped less than actually happened. If it were at least changed to "always 2 charges" (or always 3 charges, if you want) then I could live with it, even though I still don't think it would accomplish what it was supposed to do. But randomly 2 or 3 PLUS telling the player it was always 2 is unacceptable.

[Siegurt]

The only change in wand-id is you use id scrolls on fire wands or something before lair occasionally. Does it happen? Yes, but the drawbacks are way bigger than any benefit from the change.

Of course, there's also this to consider: why did this change happen after several different changes that specifically reduced the use of ID scrolls? A comparable change would be to make rings only work half the time unless you ID scroll them. That didn't happen; instead rings just auto-ID now. Why did wands go the other way? I mean it's clear in the past that there is no overarching design goal for a lot of these changes, but this one really bothers me. Crawl should never be deliberately trying to mislead unspoiled players like by telling them wands are zapped less than actually happened. If it were at least changed to "always 2 charges" (or always 3 charges, if you want) then I could live with it, even though I still don't think it would accomplish what it was supposed to do. But randomly 2 or 3 PLUS telling the player it was always 2 is unacceptable.

What if it was {zapped: 2ish ?}
or {zapped: 2+ ?}

[crate]

something that makes the game not lie to the player, yes

[dpeg]

Well, the number of charges is unknown anyway. The only drawback to the unspoiled player is that she thinks zapping the wand loses 2 charges when it is 2.5 in reality. (This means that identifying the wand is worth a tiny bit more than she assumes.) On that front, I don't see what's wrong with actually displaying the actual number, i.e. go from {zapped: 5} to {zapped: 7} or {zapped: 8}, depending on what was rolled.

On the id-system: I am sorry that you got your hopes up for all identification to be completely removed. I don't like that prospect at all and will try to come up with ideas to make it more interesting. As I see it, that's a matter of taste: personally, I like that the game starts with everything unidentified, because the id-minigame forces decisions at a stage when you have rather few non-tactical questions (few items, only little xp to spend etc.).

It goes without saying that no-brainers must be purged -- for example, trying to deduce jewellery functions by silly rituals is bad. On the other hand, while the decision to read-id a wand may be trivial to you, in my experience it does come up (all of this is about the early game, of course, when non-top tier wands are useful and scrolls of identify rare).

If all items are identified at the outset, the game will be (a) easier, (b) change the decisions about "what/when/how should I identify things" to "which of these will I use" -- which is precisely the kind of choice you have through the rest of the game, and (c) less fun. I don't care much about (a), I explained (b) and I can say what I mean by (c): currently, you obtain a new consumable in two stages (as the plain item first; at some point later identified, so fully usable). I like this kind of "extended gratification", seeing how the unknown items pile up etc. I also like the little surprise of finding unidentified items later in the game, and it'd be less fun for me to see right away it's a "potion of experience" or a "scroll of acquirment".

A good game can be made with or without an id-system. I really want one in my games. So I try to find ways to remove non-choices and replace them by real choices. (The bit about ?identify telling you the top card was also my idea, but I like the new wand behaviour better. I wouldn't have increased the number of charges, by the way, but that's fleeting -- what's a number?)

[crate]

I think it's interesting you bring up decks and wands. As I said my main gripe with the wand thing is the game is lying to me (isn't this a perfect example of the kind of number you want crawl to avoid? 2 should always be 2. It should not sometimes be 3). "Zapped: choko" would be an improvement here despite giving the player "less" information! My secondary complaint (and this is just a complaint, not feedback of any sort) is that it's frustrating to feel like the devteam doesn't know what it's doing (let's reduce identification! let's increase identification!).

---

But as an aside I don't see the current interaction with wands and ?identification involving any actual identification at all; you're never using a wand that has an unknown effect (well, except random effects wand, but identification doesn't change that). If you like that design, why not implement my half-serious ring suggestion: you know the type of every ring when you find it, but using a scroll on it will make the ring more effective. Now you have to choose which rings matter! This is precisely the situation with wands. If what you really want is for ?id scrolls to be instead "scroll of improvement" then sure, that's a fine idea, but it's not identification.

Decks are actually the one instance in crawl of what I think is actual identification done well (i.e. you are actually seeing what the effect of the item is). Ideally, if you want players to identify things, I think you'd extend it to at least potions, also:

Each orange potion in my game could be one of several things and I know which effects are in the orange potion pool. E.g.: haste, heal wounds, might, magic
Any particular orange potion can have any of those four effects. It is chosen independently of what other orange potions are. All un-ID'd ones stack together.
I can use a scroll of identify on my orange potion stack to identify a single orange potion from the stack. I then know exactly what that particular potion does.

The important things: the effects are all generally beneficial (you can do away with this but it's a lot less fun for many players: see the discussion on re-splitting Nemelex decks after the changes that made him give just war and escape), and I can use un-identified orange potions to give myself a combat boost. However, because the effects are useful in different situations, I will almost certainly be wasting some of my total potential power if I use un-identified potions. It keeps identification useful all game, since any new orange potions I find are unidentified.

The big problems are the UI hit (you'd be adding potentially a lot more stacks to the player's inventory, depending on how you implement this) and of course the difficulty to implement it in the first place. So I don't see this happening, unfortunately, since it's actually a good way to do identification. (I think the nethack-style thing where you just learn what the item does is no fun and should be eliminated.) (Extending the idea further to scrolls is problematic because identify is itself a scroll, but you can fix that in a few ways)

Some reasons the deck model is a lot better than the current (for instance) potion model are described below.

Quaff-id'ing in the current system leads to problematic behaviour. There are a few things involved in this.
1) You (almost) never want to quaff a stack of size 1 just to identify it. You can't do much with the knowledge that that stack you just quaffed is brilliance, since you don't have any more to use when they would be useful.
2) To avoid wasting potion effects, ideally you'd quaff-ID near a staircase which has something you would like to fight with a buff active on the other side. (Early examples might be an ogre or a centaur.) You quaff some potions (preferably from stacks of size 2+) on one side of the stairs until you have sufficient buffs to fight, then you go fight. If you don't get the buffs you're looking for, you just avoid the monster. The big problem here is the incentive to set up this scenario.
3) Once you've identified a stack that's it, you're done. This is problematic since it means you will either reduce the number of identify targets pretty quickly, or you'll have the player not using that stack for a long time. Maybe this isn't a problem, per se, but I think it's pretty clearly inferior to the deck-like system.

Compare to the deck model:
1) Not a problem because quaff-id literally doesn't do anything except waste a potion. You either use the potion in battle for its effect, or you just don't bother using it at all.
3) This is the entire thing that this suggestion would solve.

2) is potentially still a problem, with less reward (no long-term knowledge, the pool of possible effects is smaller) but you could do it much more often. However, you're still wasting permanent resources by quaff-id'ing like this (if I'm looking specifically for haste from my above pool, I'll need to quaff four potions on average to get it, instead of just one if I have a known haste potion) so I think the lower reward is enough that it's actually a net-negative to do this.

[Galefury]

Cool idea, I like it! I think it would be more interesting if you don't know the appearance of each group at the start of the game. Here's a suggestion for grouping the current potions, with purely bad potions removed. I think these would all have different characteristics regarding their unidentifed use and their interaction with identification. The group names are very tentative suggestions.

Potions of Power: heal wounds, brilliance, agility, might, invisibility, (lignification?, berserk?, flight?) - mostly good tactical effects, very large group
Needs a name, I'll call it SRM-group for now: speed, resistance, magic, (berserk?) - good tactical effects, more rare, more specific to certain situations, small group
Potions of Alteration: mutation, benemut, cure mut - mutation stuff
Potions of Improvement: restore abilities, experience, porridge, (flight?) - out of battle non-mutation effects
Potions of Aid: curing, cancellation - removes confusion and some other stuff
blood: blood - vampires make these

The Power group would be okay to just quaff unidentified in battle, especially early in the game. But you might waste a turn or get screwed over by lignification or involuntary berserk. Lignification should probably just be removed in this framework. Berserk might be a better fit in a group you would want to identify anyway because it being in a group has some side effects. For example you would want to have a melee weapon equipped before quaffing unidentified. Flight could maybe be in this group.

The SRM-group would be one you generally want to identify because you usually only quaff these in situations where you cant afford wasting a turn, and you really don't want to waste speed. It is fairly small to give some incentive to quaff these unidentified hoping for the effect you need. Alternately, berserk could be moved to this group to streamline the use of unidentified Potions of Power, in which case you would basically always want to ID SRM-group potions.

The Alteration group you either identify or ignore. If you have lots of ID it might make sense to go fishing for benemut or check if you have any cmut available.

The Improvement pool contains no bad or double-edged effects, and if you have identified some restore abilities potions and have a source of flight you can probably just chug these down in a safe place looking for experience. I think flight potions are guaranteed in some vaults, these and other guaranteed potions should of course just come pre-identified. Also flight potions have some use in battles, some use for getting into certain vaults, so could also be in the Power group.

The Aid pool removes confusion. I don't know if cancellation removes poison and rot, but cancellation would only come up rarely anyway. If curing should be the only potion in its group, cancellation could be moved to Power.

I think blood potions are not randomly generated anymore, so they should just be in their own group.

[njvack]

Mod note: I've somewhat arbitrarily split the ID comments into its own thread. Probably dpeg's message should have gone here too :/

Edit: better now. Continue.

[1010011010]

How would Inventory work? You pick a potion type, then an actual portion?

As for groups I would try and keep the same number in each all with one "bad" one.

  Code:

Power:         Heal Wounds, Strength, Agility, Berserk
Magic:         Magic, Intelligence, Haste, Ambrosia
Self-Mod:      Mutation, BeneMut, CureMut, Experience
Enchantment:   Flight, Invisibility, Resistance, Lignification
Aid:            Curing, Cancellation
Blood:         Blood


My rational is that all groups must be:
short and easy to remember (minor point).
In battle potions are grouped to gather as are long term ones.
Power and magic help melee and spell casting respectively but both contain potions desirable for the other. This is to give some worth to both groups regardless of what build is being played.
Bad potions exist to deter using random potions mid fight and increase the value of identification.

Power encourages Melee combat, HW and to some extent Agility is helpful outside of that. Berserk is dangerous when used out of place.
Magic encourages spell casting, Haste begin good for all. Ambrosia is dangerous when used out of place.
Self-Mod is the out of battle group, has good potions mixed in with all those potions of mutation. Weighted heavily towards mutation.
enchantment helps for situations. Lig could be something else since most people want to fly/invis/resist for particular threats before engaging.
Aid is mostly curing.

This does perhaps encourage behaviour like swapping in a amulet of clarity before using random power or magic potions, but that could be seen as wasting potions.

[johlstei]

I really like crate's ideas, though certain things I would exempt from it. Anything strategic could be scrolls or just found ided. (Porridge, experience, restore abilities, flight I'd say.) I was going to put cmut in there but I like the idea of an "alteration" stack.

[dynast]

people who dont Id their wands are just gonna zap until it runs out anyways and unless you Id it you will never know if your wand of random effects has 6 or 39 charges, so what difference does it makes if your wand of random effects says zapped 5 or 20 or doesnt say anything at all?

[crate]

If the game didn't tell you anything it would've been fine, but it should never tell you something (particularly a number) that is deliberately wrong. MarvinPA fixed it though.

http://s-z.org/neil/git/?p=crawl.git;a= ... d1d59fb7f0

[Lasty]

FWIW, I'm against the idea of having to identify potions individually. It would dramatically reduce the number of effects players have reliable access to, and replace an aspect of the game that requires a bit of skill and thoughtfulness (consumable usage) with something much more random and frustrating.

I don't really enjoy Nemelex's "intentionally activate a random effect that might help or kill you", and I would hate having it applied to all characters.

[crate]

You can of course increase the number of identify scrolls, or something (or the number of potions!). But yes, if you want actual identification (effect identification) to be a meaningful thing then you are necessarily saying that you want players to be using unknown effects. There's no way around that. You can't simultaneously have a meaningful identification game and also have the player always know which effects he has to use. Unless you have some sort of solution that I haven't thought of at all.

The latter, of course (no meaningful identification, always know which effects you have available) is much closer to the way crawl currently is and always has been, for what that is worth. Combining that with the difficulty of actually implementing something like deck-like potions I think it is probably better for crawl to adopt the "scroll of improvement" idea instead of trying to salvage any value from identification. (You could even, as a bonus, combine the scroll of improvement with the two enchant scrolls.) But if the desire is to maintain an actual identification game, then I have no ideas.

As an aside, I think saying you "replace skill and thoughtfulness" by using soft-identified potions (decklike potions) instead of hard-identified potions is silly. There's quite obviously still skill involved, unless you think there's no skill in using Nemelex decks. In some ways there is more skill (you need to consider how to leverage getting each of several different effects); in others there is less (can't try to optimize for a single effect as much). If the pool of possible effects per-potion is reasonably small then it's not clear to me at all which I would call "more skillful".

(Also suggesting that randomness replaces skill is similarly something I would not agree with. Imagine crawl with absolutely no randomness, and compare that to current crawl; I would suggest that, in fact, current crawl is much more "skillful" precisely because it's more random. "Frustrating" is a personal thing so I won't say you're objectively wrong there, but I don't think I would personally agree that decklike potions would be frustrating. But then, I really like Nemelex from a design standpoint.)

[Sandman25]

"Frustrating" is a personal thing so I won't say you're objectively wrong there, but I don't think I would personally agree that decklike potions would be frustrating.

When quaffing unknown potion from an "orange" stack which can have might/berserk/agility/etc. we have only 2 results:
1) All potions are similar in effect so it does not really matter which potion you quaffed (this is bad design IMHO).
2) Potions are different in effect and you get one effect (for example, berserk) when you really wanted another one (for example, agility). This can be frustrating and really encourages quaffing potions before entering new level.

I like "scroll of improvement" idea. It would be interesting to spend the scrolls on different items with different characters.

[Lasty]

In general, I'm inclined much more to aim for removing the ID game, and I like the idea of the scroll of improvement. That said, I don't think the status quo of slowly getting full access to knowledge of what effects you have at your disposal is awful.

I definitely didn't mean to suggest that the existence of randomness always decreases the amount of skill required -- for example, non-random Poker is the same as non-random War, by which I mean not even a game. I suppose what I mean to say is that the status quo offers you a set of options you can engage at will, and the skill and thoughtfulness take the form of trying to make sure you engage them to the best effect. In what I'm gonna call grab-bag potions, the skill and thoughtfulness will largely involve trying to avoid situations where you are forced to gamble on the outcome of using the potions, and in recognizing early situations where it makes sense to leverage them despite the possible negative outcomes, and in recovering from situations where you tried to use a potion and it went horribly wrong. I do not make the claim that the latter involves less thoughtfulness or skill (retracting my earlier statement), but I do make the claim that the thoughtfulness and skill involve fewer options, and are weighted towards avoiding and ameliorating the effects of using those options. In this context at least, I personally think it's more fun to have more good options and think about how to use them well than have fewer and more unreliable options and think about how to avoid being screwed by them. That said, I also think that restrictions breed creativity, and it might be that there's some version of this idea that would strike me as provoking creative choices by restricting certainty.

[Siegurt]

I really don't understand why the ID game gets such a bad rap.

As-is it seems fine to me, perhaps the "high evocations can make you not need to waste a ID scroll on a wand" thing is a little spoilery and awkward, but that's all that.

I disagree with all the premises that crate put forth in his suggestion about "deck-like-potions"

"You never want to quaff ID a stack that only has one potion in it", I never quaff ID anything except as a last resort in only in a situation where a currently not-id'd potion would make a difference, and in those cases quaff-IDing a single-potion stack is completely reasonable.

"Setting up a situation where you can take advantage of a buff from an un-ID'd potion", Why would I set up a situation where quaff-IDing a potion is a good idea? If the situation *arises* during a game (Like I flee up a staircase after being chased by nasties, and I don't have a different good staircase to use safely, and I have stack(s) of un-id'd potions to use and it's relatively safe for me to quaff-ID (Like I've already ID'd mutation and cure mutation) them) then it's a *good* thing IMHO that I can give my un-id'd potion stacks a whirl to see if they'll let me tackle the obstacle, if they don't I have to figure another way. Going out of your way to set up such a situation is counterproductive (Why would you set up a difficult situation that requires potion using if you don't have to, ID'd or not?) In most cases you'll want to keep your potions until you *can* ID them, or *have* to use them, setting up situations where you have to use an un ID'd potion is silly, and doesn't reward the player at all.

"Once you've identified a stack you're done" I consider this to be highly superior to the deck situation, decks are awful, not being able to rely on what effects I get means what I do with a deck is drop it, and never use it, ever. Potion effects would be worse, because there's no single potion effect that's worth 3-4 turns of action in a stressful situation, particularly if there's useless or bad effects that I have to endure along the way. The only time I use a deck is *way* late into extended, when I have a ton of ID scrolls lying around, and what I do with the deck is ID the top card, if it's something I think I can get good use of in an emergency, I keep it there, if not I burn it in a safe situation and repeat this until I have a card on top that's reliable to use in an emergency situation. This is a terrible, boring, spoilery, tedious use of the item, that's only relevant way late in the game, and i would hate if that became the "status quo" The point of consumables is to use them in appropriate situations when the power bump will make a difference, if you can't rely on the power bump then you will never want to use it, and it becomes totally dead space.

I also disagree with the statement that "If you want actual identification (effect identification) to be a meaningful thing then you are necessarily saying that you want players to be using unknown effects." It's currently meaningful to be able to have a reliable effect from a potion, yes, the need for identification is transitory and occupies a smaller portion of the game, and no it doesn't mean you want players to be using unknown effects.

If you increase the number of ID scrolls to the point where you can always have the top potion on your potion stack identifed, then you've changed nothing from the 'perma-ID' status quo, except to make it so that you have to do a stupid little interface dance to burn off potions you don't want and get ones you do on the tops of your stacks (and you have to burn off less useful consumables) If you have *less* ID scrolls in the game than it requires to do that, then the situation is worse, because now you have entire categories that you don't use in important situations, and just burn off "when it's not going to make a difference which effect you get anyway" Why would we want to turn hard decisions into throwaway decisions?

I get that the current situation is that ID becomes less and less interesting as the game progresses, but so does your starting weapon(s) and spells, I don't think that the game progressing to different stages and focusing on different things as it progresses is a bad thing at all, having more tactics more choices and more tools at your disposal later in the game is a good thing.

[dynast]

Why bother identifying potions if they are not deadly anymore? Paralyze got turned into something good, a life saver if you will; Confusion now heals you; Degeneration is whatever; Decay doesnt exist; Poison you got that one right; Mutation is a coin flip. In fact, the worst that can happen if you dont Id your potions is that you gonna waste a cmut. On top of that, monsters and shops will most likely identify the potions that are important to you.

[crate]

Why bother identifying potions if they are not deadly anymore?

Would you rather have 1 extra haste, might, agility, lignification, and restore abilities, or would you rather have five more ID scrolls? I'd take the former, so I scroll-ID potions.

I really don't understand why the ID game gets such a bad rap.

well like I said it's not actually identifying wands, so...

"You never want to quaff ID a stack that only has one potion in it", I never quaff ID anything except as a last resort in only in a situation where a currently not-id'd potion would make a difference, and in those cases quaff-IDing a single-potion stack is completely reasonable.

This isn't really what I meant by "quaff-ID"; I mean quaffing a stack for the primary purpose of identifying the potion (I do firmly believe that doing so is a correct thing to do in some cases in current crawl, and if you are going to do this you would want to get extra benefits from the potion if you can). Emergency potion use is a separate thing, and I did not comment on it. I do believe the number of potion effects is too varied (too many different effects) and disparate (the effects are qualitatively different) for emergency quaffing to be useful with the status quo other than in situations where you either
1) have actually soft-ID'd the potion already (huge quantity difference is probably curing, potion you've seen an enemy quaff, etc.)
or 2) are going to die if you don't
I don't really have much to say on those situations. My suggestion changes absolutely nothing about emergency quaffing (so why would I mention it?), except you can choose from a smaller number of possible un-id'd effects (which I think is more interesting).

(I do think that paying the price of sacrificing some potions in a particularly identify-poor game to know what potions you actually have is worthwhile though. Mikee and jeanjacques both quaff-ID specifically to identify potions pretty regularly, for two examples, and honestly the only reason I don't ever recommend the practice is almost everyone does it incorrectly.)

"Setting up a situation where you can take advantage of a buff from an un-ID'd potion", Why would I set up a situation where quaff-IDing a potion is a good idea?

See above.

"Once you've identified a stack you're done" I consider this to be highly superior to the deck situation, decks are awful, not being able to rely on what effects I get means what I do with a deck is drop it, and never use it, ever.

If crawl didn't give you so many more things than you need this would be pretty clearly suboptimal. Decks are not actually nearly as high-variance as people seem to pretend (especially non-escape decks). Using decks of war or summoning or destruction in fights is a pretty sizable power spike. But, of course, this can be adjusted by changing what effects are in each pool. Let's say one potion pool is: agility, might, berserk. Are you really never going to quaff an un-identified potion if it must be precisely one of those effects? They're all "make me better in melee" (yes, agility does other things too).

also disagree with the statement that "If you want actual identification (effect identification) to be a meaningful thing then you are necessarily saying that you want players to be using unknown effects." It's currently meaningful to be able to have a reliable effect from a potion, yes, the need for identification is transitory and occupies a smaller portion of the game, and no it doesn't mean you want players to be using unknown effects.

If you don't use unknown effects, then the situation is that you just effectively don't have any unknown potions, and then every time you use a scroll of identify on potions you "find" a stack of that size. Either you want players to use the unknown potions at least sometimes, or you are just adding an unnecessary extra interface step for no reason. The only thing identify does* in the situation where players never use unknown potions is it lets them decide if they want a given item find to be a stack of potions or to be some other stack (you use the ID on potions, or you use it on something else).

Maybe you think that's sufficient to warrant keeping identify around. I don't. It's silly.

If you increase the number of ID scrolls to the point where you can always have the top potion on your potion stack identifed, then you've changed nothing from the 'perma-ID' status quo

This isn't how it would work. You identify a potion in the stack, and then that potion becomes a different stack of its own (combining with previous identified potions of the same type), and then future un-ID'd orange potions stack with the un-ID'd ones already in your inventory. There is no "top potion" in a stack. Importantly, you can continue to identify your orange potions until you run out of orange potions, and the goal of course would be to have potion generation outpace identify generation by a significant amount. (This is, of course, one of the reasons implementation would be difficult, since no current items can work this way, and it would be a significant UI problem if people walk around with 52 items since you could then create 53.)

*actually it does one other thing, which is rewards players for doing lots of nethack-style soft-identify (comparable to price-ID), which I hate with a passion. Since item generation is not flat over all floors (and in particular since vaults place potions as loot), you're rewarded at least to some degree to note on which floors you found certain potions, and for noting what size each stack of potions you pick up is, and (the most basic thing here is just comparing stack sizes)... Yes, this is a small benefit, but it's the only way you get quaffing an unknown potion to be anything different from a completely random effect out of all un-identified potions.

Of course with my decklike potions you'd get around this since all effects of a given group would have some fixed probability (easiest is just making them all the same).

---

I guess I really don't understand just how much other players don't use Nemelex or something, I dunno (do you ever use Makhleb's abilities? Do you think those are good?). Decks are good if you're smart about them. Basically unidentified potions would be draw one, and identified potions would be more-or-less like draw three (or stack five, if you prefer). There are situations where you use draw one because you don't want to pay the cost of the other abilities^, there are other situations where the actual effect you get is more important so you pay the extra cost.

^This isn't a perfect parallel, mainly since the cost of triple draw is pretty negligible, so you can just use it for pretty much everything. You can adjust this by adjusting the number of ID scrolls. (Also since identify scrolls don't actually identify anything else in crawl currently I'm sort of implicitly assuming you don't use ID scrolls on anything but potions (or scrolls, which you'd address in some fashion if this got implemented). That's not actually a terribly important thing though.)

It is almost certainly silly to really discuss this idea with regard to crawl though, since I see approximately negative chance this gets implemented. I quite definitely think it's superior to the current system from a game design standpoint--since it does not compete with the option of just making everything pre-identified--(and I personally would enjoy it a lot more as a player, though I would say that's more subjective), but it's so dramatically different from how crawl has handled things that probably it's not worth considering as a change to existing crawl. Just like how I think non-persistent floors are superior to persistent ones, but I never suggest crawl actually adopt that.

---

edit: Siegurt, do you prefer having to identify potions at all? Or would you prefer them to just be pre-identified? I certainly prefer the latter to crawl's current state.

[Lasty]

This isn't how it would work. You identify a potion in the stack, and then that potion becomes a different stack of its own (combining with previous identified potions of the same type), and then future un-ID'd orange potions stack with the un-ID'd ones already in your inventory. There is no "top potion" in a stack. Importantly, you can continue to identify your orange potions until you run out of orange potions, and the goal of course would be to have potion generation outpace identify generation by a significant amount. (This is, of course, one of the reasons implementation would be difficult, since no current items can work this way, and it would be a significant UI problem if people walk around with 52 items since you could then create 53.)

I didn't understand this part of the proposal initially. I confess that I like this version significantly more: you preserve the ability to have guaranteed resources available, but you force the player to make choices about which resources they will make available, and as a side bonus you make quaffing unknown potions less bad. I'm not sure it's worth the work of implementing in crawl, but I can see the appeal.

[dynast]

Would you rather have 1 extra haste, might, agility, lignification, and restore abilities, or would you rather have five more ID scrolls? I'd take the former, so I scroll-ID potions.

I would rather identify decks because those can actually be deadly. Or my wands. And i would rather have all potions identified on pick up than as it currently is.

Maybe if distilleries didnt sell identified potions or maybe if you couldnt tell which potion a monster is quaffing in front of you, or maybe if paralyze, strong poison and so on were still potions...

Talking about shops, whats up with shops selling unindentified potions for around 20 gold? that only compels me even more to not Id so i can maybe buy haste for cheap.

[Sprucery]

I just want to say that if everything would be pre-identified, the game would be more boring.

To me, it is fun and enjoyable that every once in a while in an early emergency situation, you have to read unidentified scrolls and/or quaff unidentified potions and hope for a good effect to survive. It is also an interesting decision whether to try on new jewellery, to wield the shiny weapon or wear the tempting piece of armour or not.

I really like the current id game (in 0.15, so no experience with the new wand system though).

[Siegurt]

edit: Siegurt, do you prefer having to identify potions at all? Or would you prefer them to just be pre-identified? I certainly prefer the latter to crawl's current state.

I prefer having to identify them. And actually I like soft-identifying potions (using judgements to guess what unidentified potion stack might be what, based on how many of them I have and whether they're "common" or "uncommon" for the part of the game I'm in) as well.

Yes, I missed the bit in your original proposal that you meant that the newly identified potion would get pulled off and made into it's own identified stack (that makes it somewhat better than decks, since you can never ID the second or third card down in a deck, much less access them without using up the first one)

well like I said it's not actually identifying wands, so...

Well, it *does* give you information about the item that you didn't have before, you can argue that's not revealing it's "identity" though, I'd label that firmly in a grey area though, myself.

Let's say one potion pool is: agility, might, berserk. Are you really never going to quaff an un-identified potion if it must be precisely one of those effects?

In that particular case, I will probably not ever quaff an un-id'd potion *unless I am sure that I can survive berserking and the subsequent cool down* That drastically reduces the effectiveness of the item to less than any one of the potions individually, to the point where there's very little difference between having the unidentified stack and not having a stack at all.

I use those three potions for very different reasons. I use !agility very often to approach or flee from ranged-weapon users, obviously this is the worst possible time to get a !berserk, I use !might when I am faced with a group of opponents that I need to kill quickly, where there's at least one thing that threatens me in non-direct damage ways (like summoners, tormentors, things that mutate or disable me etc.)l, using a !berserk in this situation would be dangerous and often fatal, and using !agility would not really help matters (Since I'm not trying to reduce damage, I'm trying to kill things before they have as many turns) Finally I use !Berserk when I'm faced with an *isolated* nasty that I need to kill quickly, And !might would help with that, !agility would help slightly, but neither would be *as* useful as a berserk in that instance.

Nemelex is a different ball of wax, both because he spams the hell out of you with lots of decks (so wasting good cards isn't a big deal) and because he's giving you renewable abilities to select the cards you want.

[dpeg]

But as an aside I don't see the current interaction with wands and ?identification involving any actual identification at all; you're never using a wand that has an unknown effect (well, except random effects wand, but identification doesn't change that).

This is because we are using different meanings of "identification". It is a bit like the difference between complete and partial information in games: (I think) you restrict identification to giving you the effect; I am fine with also applying the word when you gain some gameplay-relevant information (like saving wand charges).

And yes, that concept could be extended, and I wouldn't be averse to that. (I have some half-baked system for curses and Remove Curse that make the scroll a commodity you have to manage if you want to use Cursed Items.)

If you like that design, why not implement my half-serious ring suggestion: you know the type of every ring when you find it, but using a scroll on it will make the ring more effective. Now you have to choose which rings matter! This is precisely the situation with wands. If what you really want is for ?id scrolls to be instead "scroll of improvement" then sure, that's a fine idea, but it's not identification.

I don't care so much about whether the use of ?id is "proper identification", I am already happy when it triggers some decisions. Because I took our id model for granted (more below), I resigned to meaningful choices about ?id use to only come up in the rather early game. That's why I wanted to add a new use, to compete with the others. Needless to say, that one change was not meant to save the id-minigame or anything, just a little contribution to turn the horrible "identify wand type by zapping at a rat two spaces away" into something which may be relevant at times. (I can imagine that in serious games, you might decide to identify your D:2 wand of frost, on a shaky character.)

[Deck-stye id model]Each orange potion in my game could be one of several things and I know which effects are in the orange potion pool. E.g.: haste, heal wounds, might, magic. Any particular orange potion can have any of those four effects. It is chosen independently of what other orange potions are. All un-ID'd ones stack together. I can use a scroll of identify on my orange potion stack to identify a single orange potion from the stack. I then know exactly what that particular potion does.

That idea is novel and has gained some traction at ##crawl-dev. It would turn ?id from an early game commodity into something that is relevant throughout the game. (This is probably, though not automatically, a good goal.)

The big problems are the UI hit (you'd be adding potentially a lot more stacks to the player's inventory, depending on how you implement this) and of course the difficulty to implement it in the first place. So I don't see this happening, unfortunately, since it's actually a good way to do identification.

Why so negative?
First, when trying this out, I'd opt for a very conservative approach (the concept is radical enough as it is), and do it for one, restricted class of items, like the tactically relevent potions you listed. That also keeps the interface in check.

(I think the nethack-style thing where you just learn what the item does is no fun and should be eliminated.)

I guess it's almost universal among roguelikes. Fun is subjective, so I'll grant that you won't ever find it fun. I tried to explain above why I think it's a good idea to have something interesting going on in the early game that peters out later on, no need to repeat.

Some reasons the deck model is a lot better than the current (for instance) potion model are described below.

Quaff-id'ing in the current system leads to problematic behaviour. There are a few things involved in this.
1) You (almost) never want to quaff a stack of size 1 just to identify it. You can't do much with the knowledge that that stack you just quaffed is brilliance, since you don't have any more to use when they would be useful.
2) To avoid wasting potion effects, ideally you'd quaff-ID near a staircase which has something you would like to fight with a buff active on the other side. (Early examples might be an ogre or a centaur.) You quaff some potions (preferably from stacks of size 2+) on one side of the stairs until you have sufficient buffs to fight, then you go fight. If you don't get the buffs you're looking for, you just avoid the monster. The big problem here is the incentive to set up this scenario.
3) Once you've identified a stack that's it, you're done. This is problematic since it means you will either reduce the number of identify targets pretty quickly, or you'll have the player not using that stack for a long time. Maybe this isn't a problem, per se, but I think it's pretty clearly inferior to the deck-like system.

1) is not a problem in itself.
However, 2) is. Then again, your concept also suffers from this. (The real culprit is being able to take time-outs via stairs. So perhaps using a staircase should cancel all temporary effects.) Of course, you can do the same with a sleeping ogre nearby -- and here the current system is better at punishing you for this.
As I said, 3) is not so much a problem for me: I don't mind identification becoming less prominent as the game goes on. I admit that having a bunch of bad items just for the id-minigame is a cost in itself. (The previous approach to salvage that -- by providing corner cases where good turns bad -- doesn't really work out. I think it could work if identified bad items simply don't generate anymore once you know them. Rationale: they now fall under dungeon rubbish, like all the other items that presumably exist but are not simulated by the game.)

Thanks for the interesting concept!

[crate]

Oh, I see what you mean with 2); I was only thinking of the situation where you wanted to quaff-ID a potion (which does not exist at all if you adopt decklike potions), but had not considered the situation where you have decided you probably want to use a buff for a fight but want to avoid fighting if you get a bad effect. I don't think the sleeping-monster situation is a problem: there's no luring-to-stairs involved, which is the behaviour that I really think should not be encouraged (using a potion, getting an effect you did not want, and then just not fighting is not a problem: you've used a resource--the potion--so you're still paying a cost). Removing statuses when taking stairs sounds reasonable to me and would also probably be a good idea because of things like haste spell anyway.

[Siegurt]

Hm, would that include negative effects?

[crate]

Why would you include purely-negative effects in the decklike potion system? There's no reason to keep them around for incentivizing identification (which was probably the rationale for retaining the old purely-negative potions for so long), and I don't see a benefit to including them in any particular potion pool. And I think it's safe to say that players in general don't enjoy using purely negative effects from unidentified things. I think it makes more sense to make every potion either good or double-edged like lignification or mutation; if that's the case, then you're still using up a permanent resource if you do the quaff-near-sleeping-monsters thing. It just might not be a particularly useful resource.

I don't see a real problem with quaffing near sleeping monsters anyway though; it's just part of the reward for not waking up the monster. The only effects I can think of that you'd actively want to wait out are things like lignification that have their own risks even in relative safety; if you quaff a beneficial potion that wasn't the one you wanted, you just exclude the monster and walk away (which is what you'd have done if you just didn't have potions at all). So this situation doesn't have the tedious pull-the-monster-a-long-distance thing going on that's the problem with stairs.

Even with things like waiting out berserk slow at least you can just press 5 (or use autotravel if you want to wait it out in a different location), which you can't do when pulling a monster along with you.

[Sprucery]

Would also all negative status effects expire upon using stairs?

As you know, it would be a very radical change to the current system if you couldn't buff yourself up before going downstairs.

[crate]

Presumably negative effects wouldn't be cured by stairs. It would be unintuitive, perhaps, but doesn't sound too hard to figure out. I think the gameplay ramifications would be good on the whole.

[mechanicalmaniac]

FWIW I actually prefer the pre-0.14 ID game to the current Larn-style "immediate and guaranteed identification." Identification of items felt more meaningful. 'Course, this could just be because I hail from NetHack, but then again I'm probably not the only Crawl player who hails from NetHack.

[KoboldLord]

I also hail from Nethack. The Nethack-style identification mini-game works great in Nethack, because it includes the breadth of options required to make that mini-game actually work. There are reasonable ways available in Nethack to profitably soft- or hard-identify your consumables before your identify scroll supply makes that concern obsolete, and there are actually serious potential drawbacks to just use-identifying everything as soon as you have a second copy.

Crawl no longer has the intricate subsystems that make Nethack-style identification work, and it hasn't had those subsystems for a long time, if it ever has. Crawl's version of the item identification mini-game has been on the way out for a long, long time, and nowadays it is little more than a routine double-unlock system to slightly slow down loot acquisition for the early game. Both styles of item identification are a valid design choice, but since Crawl is clearly going with a minimalist identification style it should probably go all the way rather than hold onto a few legacy mechanics that it isn't using properly.

[Sprucery]

since Crawl is clearly going with a minimalist identification style it should probably go all the way rather than hold onto a few legacy mechanics that it isn't using properly.

I also hail from Nethack and agree with what you said except for this last part. I still think that Crawl would be more boring if everything was identified from the start.

Actually, I think it would be an improvement if misc evokables started out unidentified as well. Using the evokable for the first time would reveal it's identity (and of other evokables of that type).