Name dcss:brainstorm:misc: Time
Summary Passage of time
Further informationDiscussion about various timers, their uses, advantages and disadvantages. Related: Unnanounced portals, God Mechanics, Lugonu, Interrupted Activities Reform
Added by b0rsuk
Added on 2010-04-29 23:54


Some effects in Crawl depend on passage of time. Potions or scrolls may work for certain amount of time. Breath weapons can be used once in a while. Magic Points recharge over time. God wrath pales with time. Characters become hungry over time. Piety for most gods changes over time. And so forth.

Based on effect on player, they can be divided into two categories:

  • In player's favour (Gods become less angry, MP recharges, breath weapons recharge, piety of good gods increases). I'm going to call these beneficial.
  • against player (hunger, timed portals close, gods growing impatient, potions or scrolls running out, summoned allies running out of time). I'll be calling these harmful.

Based on duration, the effects can be divided into short and long. I'll be calling these respectively tactical and strategic.

From abuse/scumming point of view, I think it's a good rule of thumb to use turn-based counter for majority of harmful effects. Player will always be struggling to beat the clock and no pathological behaviour emerges. But I admit there are occasions where the rule may be broken. Used my definition, Sage effect from from card (increases XP gained for a skill) would be classified as harmful strategic. Yet the effect lasts for 3 levels, presumably so that player doesn't feel rushed. I don't see a problem in this case. Additionally, Yredelemnul's ability to capture souls (I forgot how it's called) is currently time-based: player has to kill chosen monster within a time limit. While not abusable strictly speaking, turn count in Crawl is currently player centric. Players using “weaker”, but faster attacks may feel unjustly punished for their playing style. Having an objective turn counter would be nice.

As for beneficial effects, in my opinion it should depend on whether it's tactical or strategic (duration). Basing beneficial tactical effects on turn counter sounds fine - it's reasonable to expect player to wait a couple of turns for breath weapon or MP to recharge or for wounds to heal. Doing so very often is generally limited by food. But I think there's a problem with beneficial strategic effects. In particular:

  • Gods giving piety for passing turns encourages player to spend excessive amounts of time in one spot to “farm” piety. This has been dealt with in case of Sif Muna (now piety for learning magic skills), and good gods to some extent (they don't accept Mummy followers). I understand good gods are going to switch to some other kind of timer in the future.
  • A common strategy to deal with wrath of many gods is to “hole up” in safest possible environment and wait. Among other negative consequences, this strongly rewards reading spoilers.
  • Demonic Guardian, a new demonspawn mutation, can require large amounts of time for the demon to respawn. Particularly at level 3 of the mutation. Ideal play suggests to wait that much before entering any new area - I think it encourages players to bore themselves.

Exploration Timer

The idea popped in discussions about Stealth god and alternate (to killing) methods of gaining experience. Counting explored tiles sounds like a good way of measuring game progress. Different branches can be given different weights. Because of issues with good gods, it has been proposed to base piety gain on exploration. Basically, “time” passes only as player character enters new areas. This has the effect of not allowing players to gain benefit in cheap ways like simply waiting.

I suggest using Exploration timer for most or all beneficial strategic effects, including Demonic Guardian mutation and god wrath. It cuts down abuse and boring play while being (in theory) inobtrusive. And it's just a cool idea I'd wish to see tried :-).

Curiously, it's not out of question to use exploration timer for Abyss as well. For the uninitiated, a recurring idea is to make Abyss progressively nastier as player spends more time in it. Currently Abyss is infamous because for a reasonably experienced player character (generally around character level 20) it's very easy. There's no challenge, no real resource drain, not that much landscape or creature variety, and there's even some loot to be gained. High level characters often have nothing to worry about when banished. Looking for the Abyssal rune is less fun than popping pimples. What does Exploration Timer have to do with Abyss ? It depends on approach taken. Turn Timer discourages even resting in quieter parts of the Abyss. I don't think it's necessary to punish resting in Abyss - the exit is not getting closer this way, hunger and piety loss is also a factor. It's interesting to consider using an exploration timer here - player would be able to rest (limited by occasional new level, hunger, piety) but would still be getting tougher and tougher challenges as he moves through the Abyss.

Looking far into the future, it seems possible to move some events currently linked to level generation (like Boris reappearing) to exploration timer instead. That is, if player partially explores a level and kills Boris, goes to explore another branch and returns, Boris might be found again on the same level !

Bottom line: I'm trying to change the perception of the concept. Rather than just balancing a god or two, you should perceive it as an another way to measure passing time and game progress. I'm in favour of a general exploration timer, which could be used for a broader class of effects.

So shouldn't I stand in one place and shout at stuff until it stops coming, since my durations are unlimited? — tgw 2010-04-30 21:09
You misunderstood. I never proposed to change all effects to use exploration timer. By my definition an effect like potion of might is classified as tactical (short duration) and harmful (because time works against player). “Harmful” affects work fine with timers based on turns or time units. See above. b0rsuk
Very good analysis, I like the structural thinking. I have plans to change god wrath, waiting it out is counter-intuitive, boring and really un-Crawl-like. However, I don't think an exploration timer is good for this. The weaker reason: switching gods becomes harder and harder. There's also some meta-gamey effect involved as players who plan on god switching would try to explore convervatively. Some might say this is good (but note the mechanic would be non-intuitive), but… The stronger reason: if possible, we should use timers that are related to the desired effect. In the case of god wrath, doesn't it seem much more natural to have all godly punishments reduce the timer?
Exploration timer is so suitable for non-killing xp, because new monsters (=new xp) can only (well, mostly) be found in unexplored areas. — dpeg 2010-05-01 10:24
Exploration timer would be just a measure of time passed. So no “switching gods becomes harder and harder” - gods' wrath wouldn't be proportional to number of squares seen in the entire game. Gods would just stop punishing until you start exploring. Instead of “Punish player every 20 turns passed” it could be “punish player every 20 squares explored”. This is good in many cases and avoids the “trap syndrome” - traps only having an effect when they succeed in killing the player (which is not a lot of fun). Any temporary damage from traps is likely to occur when it's quiet, and player can just rest to heal. Time passing only while player is exploring would mean that punishments would happen 1) in potentially dangerous area 2) not in area chosen by player, especially not in a tiny 1 square cavern surrounded by deep water. These are gameplay benefits. As for theme, I can totally see gods waiting until player leaves his hideout and especially waiting for the right moment to strike. As for reducing the timer - you misunderstood. It could work like the Dungeoneer example on the bottom of the page. Exploring accumulates Threat points, which can be later spent by other players (or god in case of Crawl) to punish player.
For gods whose wrath can cause permanent damage, turn or action timer can be ok. Similar for Banishment, bad mutations. Attacks by divine servants and blasting with divine energy can be potentially deadly at any time, so it could be ok to just use turn timer as well. But Slow, Starving, Confusion and similar disabling effects should absolutely use exploration timer as they're unlikely to be occur meaningfully otherwise. I can't imagine playing Crawl without exploring, so I don't get how you could “explore conservatively”. b0rsuk

Turn Timer

Turn timer is generally good to use for tactical effects, either beneficial or harmful. It may encourage unintended behaviour when applied to beneficial strategic effects. Until recently, player turns were used all over the place for purposes of measuring time, which caused anomalies in some cases.

Time Unit Timer

A relatively new invention. Objective time units are a better choice when duration of tactical effects is supposed to be independent from player actions.b0rsuk

Most durations (haste, might, etc.) were changed to work off absolute time instead of player turns for 0.6. The unit for “absolute time” is the same as the unit of you.time_taken. Normally, a turn takes 10 units. Regeneration is time based. Hungering is turn based. — rob 2010-04-30 08:48

Other timers

  • New Level Timer (used by Sage card)
  • Experience/kill timer (it can be said piety for many gods uses this; it is also an interesting thought experiment to imagine clevel/skill progress without kill experience)
  • Skill gain timer (Sif Muna)
  • Gold Finding Timer (proposed for some new gods)

b0rsuk 2010-04-29 23:54

Absolute timer

Right now, the turn counter displays player turns (actions) instead of absolute turns. It creates several problems:

  • It's hard to know the time taken by very fast actions like wielding weapon and switching jewellery. For a newbie, switching jewellery looks like it takes 2 turns and he will avoid doing it in combat. I've learned by experience that it's faster, but I still don't know exactly how fast it is.
  • You can't keep track of absolute time. You know that this corpse will rot in 200 turns, but it can be more or less actions depending on your movement and attack speed and what you are doing.
  • It creates some inconsistencies as some effects have a duration based on actions, and others on time.
  • It makes speed run completely backward. By counting the number of actions instead of the number of turns, it rewards slow efficient actions (big weapons). A naga with the same aptitudes as a spriggan could theoretically be a good speed runner. Chei isn't a big penalty to speed run.

I suggest we replace the turn counter (which is an action counter) by a real turn counter which displays absolute time, with a decimal. We can keep counting actions for the dump log, but we don't need it on the interface, I don't see any use for it.

I've added an option show_real_turns (default off) in trunk to test this.

galehar 2010-10-11 22:31

How would this interact with the scoring mechanism, since turn count is so integral to final score? For example, for the longest time I thought that the turn counter *was* counting absolute turns pssed, so I kept myself swifted and flying as much as possible in an effort to reduce my number of turns and thus increase my score. The day that I found that the turn counter was independent of time that had passed was one of the happiest days of my crawl career, because I could finally stop renewing those spells every three seconds. It is for precisely this reason that I don't like the idea of rewarding turns passed rather than actions taken.
However, I do like the idea of being able to see how much time has passed, for the other reasons you state above (food rotting, etc.). Perhaps part of the issue is that the word “turns” is overloaded. Like you say, I'd prefer to rename the current turn counter to “action counter” so that new players can avoid precisely the scenario that I describe above. This would also us allow disambiguate the interface for the new turn counter, since most people who turn on one would probably also turn on the other. — wensleydale 2010-10-12 00:14
The option I added just change the display, not the scoring. It might bring some confusion about how scoring is calculated, but as your experience show, it wasn't that clear to begin with. I don't know how to make the scoring calculation more obvious. Maybe add a “progress screen”, with number of kills, number of actions, and other scoring related informations.
I think the real turn counter does clarify the tactical situation. The action counter makes newbies believe they are playing something like chess and that every action take the same time and everybody has the same speed. Displaying both counters is a bad idea. We have to decide which one is the more useful, make it the default and leave the other optional. — galehar 2010-10-12 09:33


Dungeoneer, (http://www.boardgamegeek.com/boardgame/5576/dungeoneer-tomb-of-the-lich-lord, a fantasy card game similar to Talisman) uses a form of exploration timer. In the game, each square a player character enters causes him to accumulate certain number of Threat points. Those points can then be spent by other players to play harmful effects on the player. This prevents excessive ganging up on the winner and kingmaker syndrome, as well as allowing players to play safe.

Logged in as: Anonymous (VIEWER)
dcss/brainstorm/gameplay/time.txt · Last modified: 2011-12-22 00:28 by XuaXua
Recent changes RSS feed Donate Powered by PHP Valid XHTML 1.0 Valid CSS Driven by DokuWiki