IDing and scrolls

[Siegurt]

So one of the the things that's come up in another thread is that, as is, scrolls are worth use-IDing out of combat en-masse (in contrast to potions, which are not)

There's a couple reasons for this, but chief among them are:

1. Important tactical scrolls aren't rare (And therefore aren't a big deal to waste)
2. Important strategic scrolls aren't rare (And therefore aren't a big deal to waste)
3. Bad scroll effects are temporary (or at least easy to remove)
4. Rare scrolls aren't that important (And therefore aren't a big deal to waste)
5. Important and rare scrolls aren't really wasteable (Acquirement is always useful the moment you get it, and should you happen across an EW3 early enough to 'waste' it on an early weapon, it's still not a terrible waste, because there's plenty of EW1 & 2, and having a more powerful weapon early is a good thing)

What I propose is this:

We make there be some categories of scrolls which *are* a big deal to waste. What i would like to see is for the enchant scrolls to become altogether rarer, and have larger effects (They don't have a large enough effect individually to become rare as is)

If each enchant scroll was 5 times more powerful and occurred 1/5th as often, it would be a rather important scroll, and wasting it by read-iding it early would be a fairly large oppourtunity cost.

The problem I see with this, and what I think would be a good solution:

1. Currently, ID scrolls are pretty taxed on IDing your potions in a timely fashion, if it became important or relevant to use ID scrolls on scrolls as well as potions, you'd deprive yourself of some very important scrolls and/or potions early, making yourself dead a lot more often. This could be fixed by increasing the number of ID scrolls, but that pretty much defeats the purpose of having them in the first place, and at some point everything is ID'd and your still flooded with ID scrolls, which is annoying. My preference here is to introduce additional situation ways to ID scrolls/potions, similar to how you can ID jewelry by getting into the right situation. For example taking fire damage could ID immolation, being teleported or fighting a blink frog could ID blinking and/or teleport, etc. This allows you to gradually learn what the consumables are, without either tediously use-IDing them in a safe situation or having to flood the game with ID scrolls (In fact with sufficient ID mechanisms in place ID scrolls could be reduced to uncommon or rare, and used for more unusual items)

2. With read-ID being a poor choice, it's hard to ID scrolls of identification, of course if situational-mechanism-based identification were in place, Scrolls of Identification could simply identify themselves whenever you had an unidentified item in your inventory.

3. Situational, mechanism-based mechanics have a possibility of having the exact mechanism being hidden from the user, which invites "Spoiled" behavior (Where out of game knowledge can be used for gains) of course the easy solution to this is to simply document the mechanisms in-game (presumably in the help section and in the description of the item itself)

4. Mechanism-based mechanics have the possibilty of being tedious, if "take some fire damage" is a way to identify a type of scroll, some players will inevitably attempt to take some fire damage every time they get a new type of scroll, to see if that might be it. To discourage such behavior, the conditions of mechanism-based identification either need to be ones that aren't easily arrangeable by the player, or ones that would simply happen normally when progressing through the game with no extra tedium needed. "Step into shallow water" would be an example of a bad mechanic, because it's not one that comes with any risk for the player, and can be directly caused by the player with tedious travelling. Those types of conditions would need to be avoided.

I'm sure there's other problems with this that I haven't thought of, and a lot of people just don't like the ID game at all, but I'd rather see it provide interesting choices than be removed bit by bit, personally. There's also some question of whether we could find enough mechanisms for identifying that were not "bad", I would also personally like to see the difficulty of auto-IDing something raise with the usefulness of it, and ideally be thematically appropriate (If I could ask for everything that I want:)

[Siegurt]

And I put this in the wrong forum, could a Mod move this to game design discussion?

[dpeg]

Interesting ideas. You make a really good point that "items of low value and high frequency" is a bad concept for id purposes; I wasn't aware of that.

I note that there is a potional version of situational identification: this is when monsters use potions on themselves. Something similar could be done with scrolls.

Allow me to step back and consider what an id system is good for: at the beginning of the game, you collect consumables and know nothing about them (barring background pre-identification). You will want to sensibly use those items: tactically relevant ones first, but the others not too much later.

What might speak against the strategy of immediate information gain? (Use up unknown consumables right away to identify them.)
* The item might be harmful.
* The item might be precious and thus wasted.
* The item might not identify.

What might speak for delaying use-id?
* Can try to use probabilities.
* Monster use can provide clues.
* Shops can provide identities.

All of this is countered by the very strong point that if you die with consumables that would've helped, then you waited too long.

My goal in this case (and in most others of strategical relevance in Crawl), is to make the perfect approach depend on the situation. How could that come about?
* You can delay use-id if you are safe.
* (More) items could self-id through exploration (monster use, what Siegurt writes).
* early portal vaults could have some loot pre-identified
* consumables could occasionally generate with a good/bad ego (partial information)

I think that there are blunt ways to affect id strategies: suppose that, like the guaranteed food cashes, there would be guaranteed id helpers ("guaranteed" means the depth range). This could be a stash of 3-5 pre-identified identify scrolls in a domestic vault, or a small set of other useful pre-identified consumables (good or bad), or a shop selling scrolls and/or potions. However, I hope that we can improve the situation with more subtle measures.

Sorry for this stream of conscious, more to come later. Especially about curses.

[Magipi]

Probably I missed an important sentence or two in the opening post, but I don't see the reason for your proposal. ID-ing is pretty annoying as it is, and to make it even more annoying seems like a very bad idea.

[dpeg]

Magipi: Many thanks for your important contributions to design. I would be happy to consult you in all other questions Crawl-related.

[Siegurt]

Interesting ideas. You make a really good point that "items of low value and high frequency" is a bad concept for id purposes; I wasn't aware of that.

Well, and it's not any "items of low value and high frequency" it's actually when all items in a category are either low value or not-low value and high frequency. Some or even several items of low value and high frequency for a given category are perfectly fine, some not-low value and high frequency are fine, but without some items of high value and low frequency, there's no incentive to not simply blow through all of a given category attempting to ID them all, as there's no consequence for waste.

[and into]

Well since this was also brought up in other thread too, here's one stab at how you could implement randart potions and scrolls:

+ Randart ! and ?, upon generation, draw up to three times from a pool of normal scroll / potion effects, and either once or not at all from a pool of more powerful effects (there could be some unique things or just a bit of overlap with certain rarer effects, such as those from cards).

+ If you have already IDed the normal potion or scroll effect, then upon pick up your randart ! or ? may be partially identified as having those properties. However, the more powerful unique effect that the randart consumable might have will always be unidentified, unless you use ?identify on it.

+ The partial identification does not work the other way (cf. randart jewelry)—i.e., if you ID a randart potion that gives !speed, !resistance, and !lignification, you will *not* get knowledge of each of those potions from doing this.

+ randart potions are special cased so that they will always give you glow, although the exact amount will vary; however, a single randart potion will never give enough glow to cause adverse effects just from quaffing it, even if (for instance) you get !speed, !invisibility, and !resistance from one potion.


A few random ideas for a few unique randart consumable effects:

Potions:

+ Arcane Potency: Basically like berserk rage but for magic. You get !brilliance effect, and for a period of time your magic bar turns purple and you have infinite magic. However, upon expiration of the effect, all MP is drained and you are paralyzed for X turns.
+ Stabilized Flux: Until effect ends, any time you move, you may instead blink semi-controlled in the direction chosen. Blinking in this fashion causes no glow. This is not affected by -cTele, but will be suppressed by stasis.
+ Unstable Flux: Until effect ends, any time you move, you may instead blink randomly. This teleportation effect cannot be controlled (cTele has no effect on it). Blinking in this fashion causes no glow.
+ Insight: Until effect ends, you can memorize a random spell; the spell can be memorized, regardless of your current success rate in casting it. You will also permanently gain knowledge of some number of randomly chosen items that you previously could not identify.
+ Farsight: Change form into a completely undetectable astral projection. Until effect ends, enemies cannot see or hurt you in any way, but you cannot attack in any way, either. You can pass through all walls instantly and see through all walls. You cannot use stairs.
+ Hyperactive Attention: Until effect ends you will be mesmerized by any enemy that comes within 5 spaces of you.


Scrolls:

+ Almighty Forge: Improves an artefact, or creates a randart from a chosen non-artefact piece of equipment. (Certain properties, like *rage and -Tele or -Cast, will not be generated from this effect.)
+ Doppelganger: For a short time, you get a ghost copy of your player to fight alongside you.
+ Companionship: Generates a permanent ally (can use stairs, etc.) of neutral alignment and at least animal intelligence, with an HD based on your char lvl.
+ Challenge: Generates a powerful enemy (based on your char lvl) that is carrying a high quality item. The enemy is summoned on a long timer, but if killed before expiring you get the item (otherwise it takes item with it as it departs). Can create uniques who otherwise were not slated to be generated in your game.
+ Sanctuary: Instantaneously teleports you to inside the Ecumenical Temple.
+ Xom's Favorite Gag: Xom pretends to be Okawaru / scroll of acquirement, and plays a game of dress up with you. Until Xom gets tired of it (i.e., until you explore X amount of new squares in the dungeon), Xom replaces some of your equipment slots with grievously substandard randarts that cannot be removed. Being named by Xom himself, these temporary randarts should have ironic monikers making fun of the other gods. ("-6, -2 Quick Blade of Trog's Potency," "-2 Robe of Elyvilon's Chaste Virtue," etc.)

[Siegurt]

Note that personally, I don't like the idea of randart consumables. I personally think consumables, as tactical or strategic one-use things should be stackable, and homogeneous in effect. Having a potion or scroll that had several mixed effects that was also super rare and randomly generated leads (in my opinion) to a dilution of scroll/potion effects and a tedious buildup of randart consumables (as you try to hold on to those ultra-rare ones for just the most optimal situation), which ultimately leads to inventory shuffling as you try to swap out your 'best' set of panic buttons. Non tactical ones are even less interesting, and combining tactical and non-tactical consumables sounds just plain awful.

I *do* like some of the ideas for new potion/scroll *effects* in and_into's post, There's no reason "Arcane potency" couldn't be a potion (or scroll) on it's own (Not randart-y at all)

And while it is tangentially related, I'm not sure 'randart consumables' really addresses my point about use-IDing very well.

[conted]

As I said in the other thread, the premise that there are no rare and valuable scrolls to warrant scroll-IDing scrolls is simply false. Scrolls of summoning, silence, holy word, brand weapon, and vulnerability are rare and valuable all game long; and scrolls of teleportation, blinking, fog, and fear are rare and valuable in the early game.

[Siegurt]

Actual "rareness" from the code:

  Code:

 random_choose_weighted(
                1800, SCR_IDENTIFY,
                1250, SCR_REMOVE_CURSE,
                 // [Cha] don't generate teleportation scrolls if in sprint
                 802, (crawl_state.game_is_sprint() ? NUM_SCROLLS : SCR_TELEPORTATION),
                 331, SCR_FEAR,
                 331, SCR_MAGIC_MAPPING,
                 331, SCR_FOG,
                 331, SCR_RANDOM_USELESSNESS,
                 331, SCR_RECHARGING,
                 331, SCR_BLINKING,
                 331, SCR_ENCHANT_ARMOUR,
                 331, SCR_ENCHANT_WEAPON_I,
                 331, SCR_ENCHANT_WEAPON_II,
                 331, SCR_AMNESIA,
                 // [Cha] don't generate noise scrolls if in sprint
                 331, (crawl_state.game_is_sprint() ? NUM_SCROLLS : SCR_NOISE),

                 331, SCR_IMMOLATION,

                 270, SCR_CURSE_WEAPON,
                 270, SCR_CURSE_ARMOUR,
                 270, SCR_CURSE_JEWELLERY,

                 // Medium-level scrolls.
                 140, (depth_mod < 4 ? NUM_SCROLLS : SCR_ACQUIREMENT),
                 140, (depth_mod < 4 ? NUM_SCROLLS : SCR_ENCHANT_WEAPON_III),
                 140, (depth_mod < 4 ? NUM_SCROLLS : SCR_SUMMONING),
                 140, (depth_mod < 4 ? NUM_SCROLLS : SCR_SILENCE),
                 140, (depth_mod < 4 ? NUM_SCROLLS : SCR_VULNERABILITY),

                 // High-level scrolls.
                 140, (depth_mod < 7 ? NUM_SCROLLS : SCR_BRAND_WEAPON),
                 140, (depth_mod < 7 ? NUM_SCROLLS : SCR_TORMENT),
                 140, (depth_mod < 7 ? NUM_SCROLLS : SCR_HOLY_WORD),

                 0);


The point is they aren't valuable *enough* to care at the stage of the game where you are likely to waste them by read-IDing.

As you can see from the weights, teleport is the most common scroll other than ID and remove curse, and in the next most commn group scrolls, we have fog, fear, blinking and blinking, which makes them all what I would call 'common' or at least 'likely to have a stack hanging around regardless of whether you readID one of them'

The actually-rare scrolls you mention are brand weapon, holy word, torment, summoning and silence. Holy word and torment are situational enough that I rarely care if I have one of those on hand, if I do, I'll use it, but if I lose one *it just doesn't matter because I was never depending on it*.

Vulnrability is only useful in a very very limited set of circumstances (IMHO the places where it's most useful are in Zigs and if you play uncarefully with haste and/or invis)
Brand weapon is useful in some games, but in a large number of them, I'm using an already-branded weapon, or an enhancer staff. In the smaller percentage of games where I could make use of brand weapon, it's never a make-or-break situation.
Summoning I primarily use as a defense/escape scroll, which is already covered fairly well by fog/blink/teleport, it's again a "nice thing to have" but not critical.
Silence is the only one of those that isn't duplicated by some other effect (other than the spell) and has a consistantly useful effect in combat, which could potentially be critical in some of them, and it's useful in few enough circumstances, that wasting one is typically not a problem.

My argument isn't that these effects aren't useful at all, and that it's not a shame to waste them, only that they aren't valueable enough to hold on to un-ID'd scrolls long enough to get enough ID scrolls to scroll-ID them (rather than simply read-IDing them) Having unknown common scrolls for a large chunk of the early game is a larger oppourtunity cost than wasting a few of the scrolls (even the the more valuable ones) and increasing the value of some scrolls would bring the choice back into balance where it's not always better to Read-ID all your scrolls somewhere around level 2-6. (Also note that some of the scrolls you mention can't even exist in the time period we're talking about)

[daggaz]

I love brand weapon scrolls and am always careful to scroll-ID rare scrolls in the later game if I havent uncovered it yet. Its often a game-changer in terms of damage output. Sucks to burn one early as they are pretty rare. Not every character has the same needs or playstyle, and this will make judgement calls on 'rarity vs usefullness' very hard to call.

Personally I kind of like the ID mini game as it is, requires a certain knowledge of the game to optimize and there is enough variability involved with drops that it can still have major effects on your early or even mid-game. Auto-id on scrolls kind of takes the mystery out of things, tho I could see learning the spell, like blink, should ID your blink scrolls. Plus as mentioned, its hard not to add tedium-bait with this kind of mechanic.

I would argue as well that its not entirely risk- or cost-free to read-ID things, and once again this depends highly on player and character type. Sure, there are _usually_ a lot of remove curse scrolls, but not always and boy does it suck to end up with a -3 -4 weapon or a -2 armour in the early game and be unable to peel it off. Makes you want to ID that gear before hand, but there are an awful lot of valuable potions in the queu already. Or just getting a launcher stuck in your hands and running out of ammo, etc..
And that blink scroll that you just burned? Probably would have come in real handy soon given the nature of this game to throw surprises around corners.

If there is one scroll that is totally risk or cost free to waste, it is random uselessness. They arent randomly useless, they are ALWAYS useless, and pointless. They should be removed or given actual mechanical effects (besides summon butterflies). But I suppose thats another thread.

[Grimm]

And I put this in the wrong forum, could a Mod move this to game design discussion?

Done.

[Leafsnail]

This entire proposal seems to be based on loss aversion. Sure, under this system wasting an enchant scroll would be five time worse, but it's also five times less likely to happen. I would still read ID my scrolls early on and accept the occasional sub-optimal enchant use.

I'm really not at all sold on the spontaneous IDing, either. It sounds like I should get out a checklist each time I get a new scroll so I can start crossing off possibilities.

[johlstei]

I agree with Leafsnail, I don't think this really changes the optimal behavior and is just playing off human logical fallacies. The scrolls you only find one of over a long period are pretty good candidates for scroll-id anyway.

This is also in direct conflict with the idea tossed around earlier to split out the strategic consumables into a different category than the tactical consumables. It seems like the scrolls you could make significantly stronger/rarer are all strategic ones, which I really wish we could insulate from item destruction(destruction of tactical consumables is okay in my book, for strategic ones it really hurts). I guess any change in that direction will always make the ID game less interesting , a side effect I hadn't thought of.