Shafts - Too common now?


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

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Tuesday, 30th April 2019, 16:24

Re: Shafts - Too common now?

tealizard wrote:VAF uses this weird reasoning a lot where he'll say there was an infinitesimal but nonzero chance that the player could have died and therefore he went on to win the game only by luck.


I think that he is worried of the opposite: there is a small but non-zero chance to survive, and it happens, so the player went on to win the game only by luck despite bad play (that resulted in the situation). I really do not think it worths to prevent it.

For this message the author sanka has received thanks:
VeryAngryFelid

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Tuesday, 30th April 2019, 17:56

Re: Shafts - Too common now?

If in a game somebody knows an optimal strategy I think it will not give great replayability for her. There must be some tension not only in the form of a one-armed robber game, where simply rolls determine the outcome (the game will degrade to this if you follow the optimal strategy) but also from not knowing which moves are the best.

Chess and go are good games because nobody really knows what the optimal strategy is. (Yes, we made AI that beats every human - but most likely they still do not play optimally.) And more importantly: I do not know what the best strategy in go is, therefore I keep wondering in every game what to do.


Crawl has many more potential situations and potential moves than chess. As such, the assertion that deaths under "theoretical perfect play" doesn't make sense. You're simultaneously warning of a risk of crawl being "solved" while pointing out a game with a much smaller pool of potential moves that has been around and more popular for centuries has yet to be "solved" even with supercomputers working on it. These assertions are not consistent with one another.

Again, the emphasis is not to have a 100% winrate. The emphasis is on avoiding introducing mechanics that arbitrarily lower the theoretical optimal without adding enough meaningful situations to the game to compensate.

Personally, I don't think shafts meaningfully alter win rate OR present particularly unique or challenging situations very often.

Now lets look at crawl for some dangerous situations: centaur/killer bee etc. on D2, or Grinder on D3, etc. I do think that there is a non-zero probability that a very good player dies in these situations.


To illustrate the impact something like D:2 bees has on the game, imagine this happening in every game. There are *always* bees on D:2 in this hypothetical. Does this make for more interesting pre-game and D:1 decisions, or does it constrain viable choices so far that it dominates player considerations of nearly every other factor?

I'll go ahead and say it. D:2 bees would overcentralize how people trying to win play the start of the game. D:2 shafts won't. This is a case for why D:2 shafts are okay and D:2 bees aren't.

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Tuesday, 30th April 2019, 18:35

Re: Shafts - Too common now?

TheMeInTeam wrote: Crawl has many more potential situations and potential moves than chess. As such, the assertion that deaths under "theoretical perfect play" doesn't make sense. You're simultaneously warning of a risk of crawl being "solved" while pointing out a game with a much smaller pool of potential moves that has been around and more popular for centuries has yet to be "solved" even with supercomputers working on it. These assertions are not consistent with one another.


"Solving" does not depend on the number of possible moves: in a game, there still can be a very simple strategy that beats it every single time even if there are infinitely many choices. (I argue that depth in a game cannot be achieved by just arbitrarily increase the number of choices.)

The "theoretical perfect play" assumption is not that you somehow calculate probabilities in a complex game. It is just that how well you can get close to it in terms of win rate, since literally 100% win rate would mean you solved the game, and for that, you do NOT need to know the probabilities of death for different actions, just that it is 0 for your actions.

Of course it impossible to achieve literal 100% win rate in crawl, but the edge case for me shows that we are too close to it: strategies with very high win rate are too common, so there is no incentive to improve! When we know a strategy that gives you 99.9% winrate the possibilities to improve just run out: we simply won't pay enough games to measure the difference in player strength anymore.

TheMeInTeam wrote: Again, the emphasis is not to have a 100% winrate. The emphasis is on avoiding introducing mechanics that arbitrarily lower the theoretical optimal without adding enough meaningful situations to the game to compensate.


You are right in this if I understand you well. I argue that we most likely need to reduce the theoretical optimal winrate in order to add enough meaningful situations that distinguish between strategies.

Of course a mechanic which just kills you 1/3 chance when you enter a new branch won't be good for this. But for example a monster which have 5% chance to kill you with the best actions we currently know could be: maybe there are better actions which reduce it significantly? The death need not to be a result of one roll, it needs to be a result of many many actions and rolls.

TheMeInTeam wrote:
To illustrate the impact something like D:2 bees has on the game, imagine this happening in every game. There are *always* bees on D:2 in this hypothetical. Does this make for more interesting pre-game and D:1 decisions, or does it constrain viable choices so far that it dominates player considerations of nearly every other factor?

I'll go ahead and say it. D:2 bees would overcentralize how people trying to win play the start of the game. D:2 shafts won't.



I guess I mostly agree with this. Always bees on D2 would be a too emphasized moment. However, if there is merely a chance, then that could result an unfair death exactly the same as if there are always bees! From the point that good play should prevent deaths there should never be bees on D2.

But bees on D2 also gives opportunities for better players for come up with strategies that lower the chance of death in these situations. So there is a low chance of bees on D2 with two output:
1. non-zero chance of "unfair", etc. death
2. opportunity for exceptionally good players to lower the chance of this death.

TheMeInTeam wrote:This is a case for why D:2 shafts are okay and D:2 bees aren't.


I do not really get it. A D2 shaft has a chance to drop you in the middle of a bee pack lower down. I have absolutely no idea if it has a higher chance than finding a single bee on D2 or not. But I think both is ok as long as it is not too frequent (currently not), exactly, as you said, to not be too central problem compared to the rest of the game.

Lair Larrikin

Posts: 23

Joined: Saturday, 6th April 2019, 23:22

Post Tuesday, 30th April 2019, 18:57

Re: Shafts - Too common now?

tealizard wrote:I'm not sure I can speak for the devteam on shafts, but the way I read the position (and I think this is correct) is: shafts are good, getting shafted twice is good, getting shafted three times is good. If you got shafted from d:1 to d:3 then to d:5 and so forth all the way down to d:15, that would be good. Maybe you would lose because of it, but it would still be good.

I mean, my experience with crawl convinces me that being able to go upstairs at all is bad. Being effectively shafted on every single descent from one floor to the next is good for the game, creates more interesting and balanced play. So all the whining about how hard shafts are is being directed at a very moderate, milquetoast perspective.


What is good about being shafted from D1 to D15?

Why is going upstairs bad?

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Tuesday, 30th April 2019, 19:07

Re: Shafts - Too common now?

tealizard may answer you differently, but here are my answers:

IveGoneSupine wrote:What is good about being shafted from D1 to D15?


Because a very dangerous situation is needed to distinguish good players and better players. The latter will die less frequently in this situation.

IveGoneSupine wrote:Why is going upstairs bad?


Because upstairs give a simple boring solution to many, many, many situations. If you do not have this solution then many situations would need much more creative play to minimize death chance: the game will be more interesting! While it is possible that other changes would be needed to not make the game overly lethal at some places, but I do not think it would be.

For this message the author sanka has received thanks:
duvessa

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Tuesday, 30th April 2019, 20:18

Re: Shafts - Too common now?

Of course it impossible to achieve literal 100% win rate in crawl, but the edge case for me shows that we are too close to it: strategies with very high win rate are too common, so there is no incentive to improve! When we know a strategy that gives you 99.9% winrate the possibilities to improve just run out: we simply won't pay enough games to measure the difference in player strength anymore.


Crawl players are not at/near 99%. Even the longest streaks depend on controlling factors, and there aren't enough of these to make strong inferences about win rate in general (you can have a run of good luck at 90% odds for example, if you played 20 games at 90% odds you'd still have > 10% chance of streaking 20 wins in a row).

Streaks longer than 20 have been attained by only a handful of people in the world, safe to say they're among the best people at this game. The top .01% of players attaining a winrate > 90% is not something that merits significant tuning in a game; an enormous proportion of crawl players have 5 wins or fewer, and quite a few have never won a game.

You are right in this if I understand you well. I argue that we most likely need to reduce the theoretical optimal winrate in order to add enough meaningful situations that distinguish between strategies.


I'm only opposed to this if the reduced winrate comes from pure RNG controlling the outcome (which would be meaningless noise). If it's coming from additional complexity/things to consider, it can enrich the game. I don't see shafts doing that in particular, but it could be done in principle.

But bees on D2 also gives opportunities for better players for come up with strategies that lower the chance of death in these situations. So there is a low chance of bees on D2 with two output:
1. non-zero chance of "unfair", etc. death
2. opportunity for exceptionally good players to lower the chance of this death.


What would happen is that exceptionally good players attempting streaks would reliably select for species/backgrounds that consistently survive or escape bees. Doing so would greatly narrow the pool of "viable starting species/backgrounds". These "new strategies" are largely things I expect elite players to have already considered, but now they would be constrained to either doing only those or risk dying to RNG intentionally.

I hold that such a scenario would make crawl worse rather than better.

I do not really get it. A D2 shaft has a chance to drop you in the middle of a bee pack lower down.


Nothing I said about bees holds regarding the shaft. In fact it's not clear to what extent shaft on D:2 would influence your starting choice at all. If it could be measured, it would not be a large effect because it's extremely unlikely that a random D:2 shaft will kill even bizarre species/background combinations like DEFi or NaVM.

Basically:

- One example centralizes player choices around a comparatively narrow set of viable possibilities
- The other has minimal if any apparent effect on what players choose/do in advance

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Wednesday, 1st May 2019, 03:51

Re: Shafts - Too common now?

Argument "we need shafts to differentiate good players from better players" is not valid, games use difficulty settings for that. In crawl it means good players should be able to infinitily streak 3rune games as MiBe and better players should be able to infinitily streak 3 rune games of HuFi, KoEn or even 15 rune NaWn of Chei.
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25

Blades Runner

Posts: 616

Joined: Thursday, 25th October 2012, 03:19

Post Wednesday, 1st May 2019, 04:54

Re: Shafts - Too common now?

duvessa wrote:also not seeing what problem this even solves? what's wrong with getting shafted multiple times in succession?


In case you truly don't know the answer to the question you're asking, I'll spell it out for you: The notion is that getting shafted multiple times in a row may be considered too harsh. If you did understand that, then really what you're saying is that you don't think getting shafted multiple times in succession is too unfair of a mechanic, and why not just say that?

There is absolutely nothing wrong with discussing the difficulty level of shafts in a thread entitled "Shafts - Too common now?"

I don't think the RNG nature of the game should change at all. The replay-ability and draw to the game is most certainly tied to the random elements. It makes perfect sense to discuss the degree and range of how much RNG can affect the game play.

If you don't want to acknowledge that difficulty level with RNG present in the game is a thing, then why is there a mechanism that sometimes stops multiple paralysis?

You didn't like my hypothetical game mechanic that with each step you have a 1 in 3 chance of instantly dying? Why not? Could you please qualify your position with something substantial? Why do you think chain shafts are good for the game? Are you in favor of chain paralysis?

For this message the author svendre has received thanks:
TheMeInTeam

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Wednesday, 1st May 2019, 09:33

Re: Shafts - Too common now?

VeryAngryFelid wrote:Argument "we need shafts to differentiate good players from better players" is not valid, games use difficulty settings for that. In crawl it means good players should be able to infinitily streak 3rune games as MiBe and better players should be able to infinitily streak 3 rune games of HuFi, KoEn or even 15 rune NaWn of Chei.


I personally do not consider gods (chei) part of the difficulty setting, it is an in game choice. NaWn had been streaked, in fact long streaks had been done with it.

Also, I do NOT think that everything should be made more difficult in crawl. Other "challenges" (like current Depths) would be best to be removed, which would decrease difficulty (you would need less patience for example). Shafts are obviously not "needed", I just think they provide a better challenge than many other things in crawl.

That is, they not only reduce win rate - they require better play to get high win rate. I only argue that without the former it would be really hard to achieve the latter. Of course our goal is to achieve the latter, but the former is a prerequisite which is very hard to skip.

TheMeInTeam wrote: Crawl players are not at/near 99%. Even the longest streaks depend on controlling factors, and there aren't enough of these to make strong inferences about win rate in general (you can have a run of good luck at 90% odds for example, if you played 20 games at 90% odds you'd still have > 10% chance of streaking 20 wins in a row).


TheMeInTeam wrote: Streaks longer than 20 have been attained by only a handful of people in the world, safe to say they're among the best people at this game. The top .01% of players attaining a winrate > 90% is not something that merits significant tuning in a game; an enormous proportion of crawl players have 5 wins or fewer, and quite a few have never won a game.


I argue that longer streaks are not rare because those players do not know what a good strategy is that could achieve even longer streaks. I argue that people are not that patient to make even longer streaks. So your chance example exactly shows that if the current known best tactic reaches above 90% win rate than the game is too shallow to distinguish good players anymore, because nobody wants to do 100< streaks to show that they reached an even better strategy. If our best tactics reach about 50%, then there would be much more space to improve.

About beginers: I do not aim to raise the floor to win. Unlike VeryAngryFelid, I am perfectly happy with situations where you can survive despite bad play due to a series of lucky rolls. I merely argue that the ceiling is too low.

TheMeInTeam wrote: What would happen is that exceptionally good players attempting streaks would reliably select for species/backgrounds that consistently survive or escape bees. Doing so would greatly narrow the pool of "viable starting species/backgrounds". These "new strategies" are largely things I expect elite players to have already considered, but now they would be constrained to either doing only those or risk dying to RNG intentionally.


I lost you here - I do not take race/class selection as part of the strategy, considering it an out of game choice (difficulty setting). Otherwise current game is clearly too shallow, I think the opinion is that DD could be almost indefinitely streaked, just nobody likes to do it.

I also think you missed in my example that it is not hypothetical - there are bees on D2 currently. They are just really, really rare. https://underhound.eu/crawl/morgue/Ykky/morgue-Ykky-20190501-022704.txt

They do not result in everybody playing DD or Troll or whatever. They can be survived sometimes.

TheMeInTeam wrote: Nothing I said about bees holds regarding the shaft. In fact it's not clear to what extent shaft on D:2 would influence your starting choice at all. If it could be measured, it would not be a large effect because it's extremely unlikely that a random D:2 shaft will kill even bizarre species/background combinations like DEFi or NaVM.

Basically:

- One example centralizes player choices around a comparatively narrow set of viable possibilities
- The other has minimal if any apparent effect on what players choose/do in advance


I think those races/classes that are more likely to survive a D2 bee are laregely coincede with those that have a good chance to survive a bad shaft. The reason they are both ok with me is because neither have a good chance to kill DEFi or NaVM now, because bees on D2 are extremly rare, and bad shafts are also extremly rare.

Please note that I do not want to raise the floor of the difficulty of crawl. In fact I am perfectly ok winning with worse tactics sometimes with luck. I consider it good for beginners. I am perfectly ok lowering the ceiling somewhat tough.

Edit: I added a missing NOT - I accidentally said the opposite of what I would have liked.
Last edited by sanka on Wednesday, 1st May 2019, 12:47, edited 2 times in total.

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Wednesday, 1st May 2019, 09:56

Re: Shafts - Too common now?

svendre wrote:
duvessa wrote:also not seeing what problem this even solves? what's wrong with getting shafted multiple times in succession?


In case you truly don't know the answer to the question you're asking, I'll spell it out for you: The notion is that getting shafted multiple times in a row may be considered too harsh. If you did understand that, then really what you're saying is that you don't think getting shafted multiple times in succession is too unfair of a mechanic, and why not just say that?

There is absolutely nothing wrong with discussing the difficulty level of shafts in a thread entitled "Shafts - Too common now?"

I don't think the RNG nature of the game should change at all. The replay-ability and draw to the game is most certainly tied to the random elements. It makes perfect sense to discuss the degree and range of how much RNG can affect the game play.

If you don't want to acknowledge that difficulty level with RNG present in the game is a thing, then why is there a mechanism that sometimes stops multiple paralysis?

You didn't like my hypothetical game mechanic that with each step you have a 1 in 3 chance of instantly dying? Why not? Could you please qualify your position with something substantial? Why do you think chain shafts are good for the game? Are you in favour of chain paralysis?


I cannot argue in the name of duvessa, but I have time and my own answers - maybe that could also help.

1. Multiple shafts are rare and landing in a bad position are rare. Really unfair deaths are very, very rare due to shafts.
2. Yes, unfair deaths may occur due to a shaft, even without multiple shafts. You may fall down a shaft on D2 and land in the middle of a pack of killer bees before picking up any consumable. But they are rare, and rare unfair deaths are ok.
3. So why rare unfair deaths are ok? The aim is not to create (even rare) unfair deaths. The aim is to create challenging, but winnable situations, with relatively simple, transparent rules. However, it is really hard. The price we pay for these situations is that sometimes we just cannot win. I fell down and survived shafts multiple times, and in some games, I did not even fall down (early on when it is dangerous), and I am not some good player, so clearly shafts are not some unsurmountable difficulty spike.

There are other such spikes: you may miss multiple times in a row. You may miscast multiple times. The enemy may hit you multiple times in a row despite having good defences.

4. The example you gave (dying 1 in 3 chance) is bad not because it lowers the win rate. It is bad because surviving does not depend on your actions. With shafts the chance to survive almost always depends on your actions, better players will survive more. The example with the bees and no consumables are very rare - it is the price we pay, but situations which are challenging but you have a good chance to survive if you can find good tactics are much more common, this is what we get.

5. I think that the reason behind special casing chain paralysis is different: with chain paralysis, you do not get any actions anymore. With shafts, you always have at least one action. I think that if after shafting the monsters get to act first, and there is good chance that they kill you in that turn, it would be overly harsh - exactly because it would limit possible counterplays (in fact it would be worse than chain paralysis). You want challenge = chance for death, but you also want interesting counterplays, which require some actions to minimize this chance of death. This is not true for chain paralysis or for your example.

To summarize: making the game harder is not a problem, so it is not very important to argue that shafts make it harder. Having rare unfair deaths are not a problem. The problem would be if shafts would never result in interesting situations: hard but survivable, or if these situations would be much more rare than the unfair deaths it may create.

For this message the author sanka has received thanks:
petercordia

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Wednesday, 1st May 2019, 15:00

Re: Shafts - Too common now?

I argue that longer streaks are not rare because those players do not know what a good strategy is that could achieve even longer streaks. I argue that people are not that patient to make even longer streaks. So your chance example exactly shows that if the current known best tactic reaches above 90% win rate than the game is too shallow to distinguish good players anymore, because nobody wants to do 100< streaks to show that they reached an even better strategy. If our best tactics reach about 50%, then there would be much more space to improve.

About beginers: I do not aim to raise the floor to win. Unlike VeryAngryFelid, I am perfectly happy with situations where you can survive despite bad play due to a series of lucky rolls. I merely argue that the ceiling is too low.


You can't actually solve for what you're saying with pure RNG though. You'd need new mechanics. Those can reasonably have RNG within them to consider, but they need agency or you necessarily lower the floor and the ceiling simultaneously. Also, for any element of complexity added to the game, it will impact new players more than experts. The latter understand mechanics better.

They do not result in everybody playing DD or Troll or whatever. They can be survived sometimes.


They are *extremely* rare. I know people can get them, but for example in 565 games I've never encountered bees quite as early as D:2. Maybe it MIGHT have happened on one of my D:1 deaths, but this is a ~.5% chance or less. Even if bees kill you 60% of the time they're encountered on D:2, we're talking about a scenario less likely than the jackal pack caster death.

But that's disingenuous. In my hypothetical they're going to be there every time, and if players know that it WILL change what they pick.

I think those races/classes that are more likely to survive a D2 bee are laregely coincede with those that have a good chance to survive a bad shaft.

What brings you to this conclusion? Only a couple species/backgrounds can kill or escape bees, almost anything can live a shaft.

There are other such spikes: you may miss multiple times in a row. You may miscast multiple times. The enemy may hit you multiple times in a row despite having good defences.


These are not fair comps. They tend to happen thousands of times per game, enough for most common statistical scenarios to happen and wash out. You can also react to them in the interim (you don't HAVE to miscast 5x, you don't even have to attempt 5 casts). Comparing high frequency of individually low-impact dice rolls to instances where a single RNG check can influence the outcome is not valid. These things force different considerations.

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Wednesday, 1st May 2019, 15:40

Re: Shafts - Too common now?

TheMeInTeam: I feel that you do not understand what I would like to say. It is very likely that I cannot explain myself clearly - sorry for that.

TheMeInTeam wrote: You can't actually solve for what you're saying with pure RNG though. You'd need new mechanics. Those can reasonably have RNG within them to consider, but they need agency or you necessarily lower the floor and the ceiling simultaneously.


I never talked about "pure RNG", whatever it supposed to mean. I specifically tried to explain that the aim is not to lower the win rate in arbitrary ways. I think it is a price that we should pay to reach our goal, so that if a mechanics lowers the win rate, that is NOT a problem in itself. In fact, that is an indicator that it COULD be a good mechanic. Since if it does NOT lowers the winrate, than most likely there is no challenge involved. That's why I think that the perceived problem of "shafts makes the game harder" is false: even if it is true it would NOT be a problem. The problem would be if shafts would never create interesting situations.

In other words: increasing difficulty (lowering win rate) is not a problem, and "but lowering it to 0% would be a problem" is not a counterargument. Nobody wants to do that, and everybody agrees I think that shafts are very, very, very far from achieving that.

TheMeInTeam wrote: Also, for any element of complexity added to the game, it will impact new players more than experts. The latter understand mechanics better.


I agree with it. I merely wanted to emphasize that I have no problem with designing challenges that are survivable with some luck, to give a hand to beginners. Also that's why I emphasized to create challenges on top of simple, transparent rules! So I argue anainst some internal state for shaft generation, for example.

TheMeInTeam wrote:They are *extremely* rare. I know people can get them, but for example in 565 games I've never encountered bees quite as early as D:2. Maybe it MIGHT have happened on one of my D:1 deaths, but this is a ~.5% chance or less. Even if bees kill you 60% of the time they're encountered on D:2, we're talking about a scenario less likely than the jackal pack caster death.

But that's disingenuous. In my hypothetical they're going to be there every time, and if players know that it WILL change what they pick.


I did not understand what your hypothetical example about bees always appearing on D2 has to do with anything. Nobody seems to claim that shafts kill you almost every game. We addressed the supposed problem that shafts might kill you. I argued that this is not a problem. I brought up that D2 might kill you with generating killer bees, and that I also do not see it as a problem. Both situations are so rare that it is mostly hypothetical (even if happened in some actual game), both has countermeasures just not 100% sure.

Now if in a very large number of games shafts would drop you in the middle of OOD monsters, or if D2 had killer bees every time, it would be a problem. But we did not talk about these situations.

Maybe I should have expressed myself more clearly with my bad analogy:

shafts = D2
very bad situation after a shaft = D2 generates killer bees for you.

Shafts are usually easily survivable and D2 is usually easily survivable. Very bad shafts may happen that could be very hard to survive. D2 bees may happen that could be very hard to survive. Both are very rare though, and you still have countermeasures to lower your probability of dying, even if in these bad situations it may not be close to 0 with the best play.

TheMeInTeam wrote:
sanka wrote:I think those races/classes that are more likely to survive a D2 bee are laregely coincede with those that have a good chance to survive a bad shaft.


What brings you to this conclusion? Only a couple species/backgrounds can kill or escape bees, almost anything can live a shaft.


I hope that it will be more clear with my clarification above about how I meant this. I suspect that combos that have a good chance to survive D2 killer bees are largely overlap with those that have a good chance to survive a very bad landing after a shaft (not just survive any general shaft). Of course it simply means that these combos has a greater chance to survive any D2 and any shaft overall (even if in general D2 or shafts are not lethal, so the chance is big for bad combos as well).

TheMeInTeam wrote: These are not fair comps. They tend to happen thousands of times per game, enough for most common statistical scenarios to happen and wash out. You can also react to them in the interim (you don't HAVE to miscast 5x, you don't even have to attempt 5 casts). Comparing high frequency of individually low-impact dice rolls to instances where a single RNG check can influence the outcome is not valid. These things force different considerations.


It depends on the fight. Suppose you fight a monster that can kill you in two hits, but has a low chance to hit you. If it hits you once, that is only one roll, yet this one roll changed the outcome probabilities more than a usual shaft.

Similarly: some monster have multiple attacks. If each of them has a low chance to hit you, there is still a non-zero chance that all of them will hit you (before you can react), and from your perspective it is one event (singlre roll) that resulted in a bad position.

My examples may have been misleading (sorry for not being clear) because of course most fights in crawl are not like that, but fights like this still happen fairly often.

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Wednesday, 1st May 2019, 16:22

Re: Shafts - Too common now?

I did not understand what your hypothetical example about bees always appearing on D2 has to do with anything. Nobody seems to claim that shafts kill you almost every game.


I used them as an example that intentionally contrasts with shafts. The intention was to show why shafts are mostly okay while something like the bees example wouldn't be, despite that the latter would certainly lower winrate for most possible species/background picks.

My intention was to point out that there are reasonable and unreasonable ways one might go about lowering winrate, with agency being an important factor in discerning whether a particular change is one or the other.

[quote/I hope that it will be more clear with my clarification above about how I meant this. I suspect that combos that have a good chance to survive D2 killer bees are largely overlap with those that have a good chance to survive a very bad landing after a shaft (not just survive any general shaft).[/quote]

I don't think so in this particular case just because the former is such a specialized situation (artificers and venom mages would have a much higher chance to live than average), while the latter has a much wider breadth of possible "bad" situations. Venom mage vs OOD shaft centaur is not going to help much. Felid can't run from bees at all but has the speed/stealth to move away from other dangerous stuff more often.

It depends on the fight. Suppose you fight a monster that can kill you in two hits, but has a low chance to hit you. If it hits you once, that is only one roll, yet this one roll changed the outcome probabilities more than a usual shaft.


It's "only one roll", but now you can simply walk away or burn a consumable. The surrounding factors are all still controlled, so even in this particular case you have less outcome riding on a single roll.

That said, we're in agreement about relative threat of most shafts.

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Wednesday, 1st May 2019, 16:28

Re: Shafts - Too common now?

You are absolutely right about killer bees being too special (so have their own weaknesses) compared to a bad shaft. I should have said "deeply OOD monster on D2, like a killer bee" or something.

About walking away or using consumable in a bad fight: usually, you also have actions after a bad shaft, that's how I meant this comparison. You can try to move or use a consumable, etc.

You are absolutely right that shafts are different because of the potential of creating uncontrolled situations. That is why they are good - I agree with this (if this was your argument). I merely wanted to take a look at the "increasing difficulty is bad" aspect.

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Wednesday, 1st May 2019, 16:51

Re: Shafts - Too common now?

We don't actually disagree regarding difficulty increases. My argument from the start is that there's a right and wrong way to do it. Initial impression was that you were arguing even "fake" difficulty would be okay if it brings winrates of elite players down, but it doesn't seem like that was actually your intention.

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Wednesday, 1st May 2019, 17:29

Re: Shafts - Too common now?

Oh, I must have expressed myself poorly. No, "fake" difficulty is absolutely not ok: the aim is not to lower the win rates with whatever means. (It is not even to lower the win rates with good means.) I just think that lowering the win rates is most likely a necessary condition for finding good solutions (and not a problem in itself, if it is not due to fake difficulty.).

Thanks for helping to clarify it.

For this message the author sanka has received thanks: 2
duvessa, TheMeInTeam

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Wednesday, 1st May 2019, 18:07

Re: Shafts - Too common now?

I don't get it. Good players can streak and thus we need shafts to differentiate the players??? It is like a competition between runners and since many runners can run long distance, we need sometimes to shoot at them from pistol. If a runner didn't evade the shot, he/she was not a good runner. Is it our only option? With the running we might force competitors to carry heavy bags or jump on one leg or climb a wall. I mean if there are players for whom every combo is too easy, we should not randomly make it harder, we should invent something which is not easy for those players. For example, let's add a species which is slow, weak and cannot worship gods.
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Wednesday, 1st May 2019, 18:10

Re: Shafts - Too common now?

^ There are already successful runs with species such as "felid mummy", regular 10 aut speed but felid HP and no extra lives/quaffing. Must have been a joy to play :p.

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Wednesday, 1st May 2019, 18:57

Re: Shafts - Too common now?

VeryAngryFelid:

I think the logic behind shafts from your point of view is the following, and you may not agree with some statements (although I am unsure which one you do not agree with):

1. We are in hunt of a game where the best tactics are interesting to find.
2. The best tactics in crawl involve a lot of luring and climbing up stairs, which solves the vast majority of situations.
3. Lot of luring and climbing upstairs is not interesting.
4. Fixing this with new race with reduced HP, movement speed, etc. is not easy, because the other parts of the game is not designed with these in mind. For example playing with half max HP would be harder for very good players, indeed. However, it would not result in better gameplay, since luring and stairs etc. still would be a good strategy. The extra challenge would be how to deal with a lot of monsters that can one-shot you, which I do not think would be a good challenge. Slow speed has better potential, but the ability to freely go upstairs still gives you a simple solutions to many situations.
5. Shafts generate interesting challenge, because it restricts your ability to easily retreat to safe territory.
6. Shafting is not like "randomly kill" the runers, because deaths where you have no ability to modify the chance of death are extremly rare. It is not a problem to have extremly rare but completly lethal situations, because it is a price we are willing to pay for having interesting challenge.
7. Situations where you can merely decrease (but not completly decrease to 0) the probability of dying are OK. It is a price we are wiling to pay to have interesting situations.
Last edited by sanka on Wednesday, 1st May 2019, 19:03, edited 1 time in total.

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Wednesday, 1st May 2019, 19:01

Re: Shafts - Too common now?

FR: New race: Downdwarf.

* Can not use upstairs or upwards escape hatches.
* Immune to shafts
* Racial tunnels: any branch entrace that you have seen but not used are available in the last level of your current branch.

For this message the author sanka has received thanks:
VeryAngryFelid

Lair Larrikin

Posts: 23

Joined: Saturday, 6th April 2019, 23:22

Post Wednesday, 1st May 2019, 19:56

Re: Shafts - Too common now?

sanka wrote:tealizard may answer you differently, but here are my answers:

IveGoneSupine wrote:What is good about being shafted from D1 to D15?


Because a very dangerous situation is needed to distinguish good players and better players. The latter will die less frequently in this situation.

IveGoneSupine wrote:Why is going upstairs bad?


Because upstairs give a simple boring solution to many, many, many situations. If you do not have this solution then many situations would need much more creative play to minimize death chance: the game will be more interesting! While it is possible that other changes would be needed to not make the game overly lethal at some places, but I do not think it would be.


Thanks for the response, sanka.

While I understand that you (and likely duvessa and tealizard) relish the challenge and are a very experienced player, what you're advocating for are changes that would impact players that do not have that level of experience. For someone brand new to crawl, getting shafted from D1 to D15 is not good. It doesn't differentiate anything, because a new player doesn't have that fundamental understanding of the game to be creative with. tealizard mentioned in another thread that he doesn't like the argument I'm using here, because he is of the opinion that new players adapt to whatever environment they come into, but I think that's a rationalization to excuse poor game design and it ignores how many players might try DCSS only to get frustrated by it and quit. Whether you want those players to stick around or not is subjective, but I think frustration can be indicative of poor design.

With regards to upstairs, I think that is player prerogative. You as a player can choose to not go upstairs as a tactic. Whether other players use that tactic has no impact on your gameplay or your enjoyment of the game. But to completely remove upstairs would require a drastic overhaul of the game. Considering, you know, you have to go upstairs during the orb run. And, as alluded to in the downdwarf post, DCSS isn't linear in it's dungeon design with branches.

Swamp Slogger

Posts: 182

Joined: Monday, 2nd July 2018, 16:47

Location: United States

Post Wednesday, 1st May 2019, 20:12

Re: Shafts - Too common now?

FR: Replace all stairs with escape hatches (3 up and 3 down on each normal level). Now, separation from stairs occurs on every descent, but the difficulty is mild since you only drop one floor. Also it's less frustrating then shafts, because it happens upon player action rather than a random event. If this is implemented, shafts can be removed.

Blades Runner

Posts: 593

Joined: Tuesday, 11th December 2018, 19:14

Post Wednesday, 1st May 2019, 20:23

Re: Shafts - Too common now?

All this aside, I do think a species that can *only* use hatches but is otherwise stronger than average would be interesting. Might need some tunneling and/or self shafting ability given how depths floors can separate and the inability to use normal stairs might otherwise soft-lock.

Or maybe racial ability TP, so you can keep trying until you land somewhere you're not trapped, with the usual risks of blindly TP :).

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Wednesday, 1st May 2019, 21:01

Re: Shafts - Too common now?

As an answer to non-linear crawl and unconnected levels, maybe I need to refine my Downdwarf idea. But I think that is CYC material.

Lair Larrikin

Posts: 23

Joined: Saturday, 6th April 2019, 23:22

Post Wednesday, 1st May 2019, 21:07

Re: Shafts - Too common now?

In all honesty, my only real point of contention is this:

Tension creation
Trap like effects happen with a fixed chance per tile revealed
Effects are either getting shafted or alarmed (mark plus noise)


Based on the title, the goal was to create tension. It doesn't do that. It just feels like punishing players for exploring levels fully or for pressing the "o," button. I don't know if tension is a realistic goal in a game that isn't really immersive.

For this message the author IveGoneSupine has received thanks: 2
Shard1697, VeryAngryFelid

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Wednesday, 1st May 2019, 21:25

Re: Shafts - Too common now?

I think tension supposed to happen after you fall down the shaft, not during exploration, fearing about being shafted.

Shafts rarely give you lethal treat. I do not feel that this is too punishing for beginners because there is a good chance that you can simply climb back. There are other situations which kill beginners frequently, like an unseen horror on D7, or, even more rare, killer bees on D2. I really do not think that shafts are the new number 1 beginner killer feature of the game.

If you are beginner, do not get intimidated by dying sometimes. It will be a better feeling when you finally beat the game.

In the past, the game used to be even less smooth in difficulty in my experience. I do not mind that many cases had been removed that created "unfair" situations, because they not really created interesting gameplay in the first place. (So I do not mind if a change makes the game somewhat easier for beginners. On the contrary.)

However, finding yourself in an unexplored level, without a way to go back, is interesting! This is an unusual challenge you should welcome! Opportunities to improve!

EDIT: I had to look up what immersive means, but come on, I play go and there is a lot of tension in it sometimes.

Blades Runner

Posts: 616

Joined: Thursday, 25th October 2012, 03:19

Post Wednesday, 1st May 2019, 22:26

Re: Shafts - Too common now?

sanka wrote:
svendre wrote:
duvessa wrote:also not seeing what problem this even solves? what's wrong with getting shafted multiple times in succession?


In case you truly don't know the answer to the question you're asking, I'll spell it out for you: The notion is that getting shafted multiple times in a row may be considered too harsh. If you did understand that, then really what you're saying is that you don't think getting shafted multiple times in succession is too unfair of a mechanic, and why not just say that?

There is absolutely nothing wrong with discussing the difficulty level of shafts in a thread entitled "Shafts - Too common now?"

I don't think the RNG nature of the game should change at all. The replay-ability and draw to the game is most certainly tied to the random elements. It makes perfect sense to discuss the degree and range of how much RNG can affect the game play.

If you don't want to acknowledge that difficulty level with RNG present in the game is a thing, then why is there a mechanism that sometimes stops multiple paralysis?

You didn't like my hypothetical game mechanic that with each step you have a 1 in 3 chance of instantly dying? Why not? Could you please qualify your position with something substantial? Why do you think chain shafts are good for the game? Are you in favour of chain paralysis?


I cannot argue in the name of duvessa, but I have time and my own answers - maybe that could also help.

1. Multiple shafts are rare and landing in a bad position are rare. Really unfair deaths are very, very rare due to shafts.
2. Yes, unfair deaths may occur due to a shaft, even without multiple shafts. You may fall down a shaft on D2 and land in the middle of a pack of killer bees before picking up any consumable. But they are rare, and rare unfair deaths are ok.
3. So why rare unfair deaths are ok? The aim is not to create (even rare) unfair deaths. The aim is to create challenging, but winnable situations, with relatively simple, transparent rules. However, it is really hard. The price we pay for these situations is that sometimes we just cannot win. I fell down and survived shafts multiple times, and in some games, I did not even fall down (early on when it is dangerous), and I am not some good player, so clearly shafts are not some unsurmountable difficulty spike.

There are other such spikes: you may miss multiple times in a row. You may miscast multiple times. The enemy may hit you multiple times in a row despite having good defences.

4. The example you gave (dying 1 in 3 chance) is bad not because it lowers the win rate. It is bad because surviving does not depend on your actions. With shafts the chance to survive almost always depends on your actions, better players will survive more. The example with the bees and no consumables are very rare - it is the price we pay, but situations which are challenging but you have a good chance to survive if you can find good tactics are much more common, this is what we get.

5. I think that the reason behind special casing chain paralysis is different: with chain paralysis, you do not get any actions anymore. With shafts, you always have at least one action. I think that if after shafting the monsters get to act first, and there is good chance that they kill you in that turn, it would be overly harsh - exactly because it would limit possible counterplays (in fact it would be worse than chain paralysis). You want challenge = chance for death, but you also want interesting counterplays, which require some actions to minimize this chance of death. This is not true for chain paralysis or for your example.

To summarize: making the game harder is not a problem, so it is not very important to argue that shafts make it harder. Having rare unfair deaths are not a problem. The problem would be if shafts would never result in interesting situations: hard but survivable, or if these situations would be much more rare than the unfair deaths it may create.


Thank you, you make some good points and I will chew on them further. I think we are like-minded about the idea that difficulty created by more interactive challenge rather than the need for extra pure luck is a superior difficulty. On this particular subject, I want to be clear that I do not think shafts are the most dangerous thing to happen. I do have a fair amount of experience, and I quite often survive shafting.

Where we may differ in opinion is that I feel the chain-shaft where you're taken down many levels, and before you can climb back even one set of stairs (or even know where a set of stairs is), being taken down several more levels, is more deadly than it sounds like you are suggesting it is. I think it is even more dangerous being shafted multiple times earlier on, before you've had a chance to stabilize your build or collect many consumables, or another way of putting it: "options" that you could leverage with skill even with the extra turn upon landing. Furthermore, you're assuming that you are only being shafted into levels never explored. For example, if a stealthy character probes downward more quickly without fully exploring each level, with intent to later circle back up to tackle things they aren't able to take on initially, you can have situations where you land on a level or levels where things are already active.

One counter-point about chain-paralysis is that unlike shafts, you can somewhat more easily control when you will or will not be subjected to the mechanism by your choice of paths through the dungeon/branches. This allows you to manage the situation by avoiding certain areas (like orc) until you find sources of MR. You have less control over shafts, they could just happen whenever for the most part. There's also no equivalent of MR for shafts. Despite these factors, at least someone has already made a determination that chain-paralysis was too harsh. I agree with that decision.

I don't think that the rarity of a chain-shafting is completely relevant. You could take my awful example of 1 in 3 steps cause you to instantly die, and say that this mode activates in only 1 of 100,000 games, and it would be super rare to die from it, but it still wouldn't be a good example of difficulty which is ideal.

Lair Larrikin

Posts: 23

Joined: Saturday, 6th April 2019, 23:22

Post Thursday, 2nd May 2019, 00:39

Re: Shafts - Too common now?

sanka wrote:I think tension supposed to happen after you fall down the shaft, not during exploration, fearing about being shafted.

Shafts rarely give you lethal treat. I do not feel that this is too punishing for beginners because there is a good chance that you can simply climb back.


sanka, you can see how I would interpret these as contradictory statements, right? Shafts provide tension after they occur but do not create lethal events, in your opinion. I've always felt tension requires something to be at stake. Otherwise, there is no reason to feel tense.

There are other situations which kill beginners frequently, like an unseen horror on D7, or, even more rare, killer bees on D2. I really do not think that shafts are the new number 1 beginner killer feature of the game.

If you are beginner, do not get intimidated by dying sometimes. It will be a better feeling when you finally beat the game.


I appreciate the above quoted text, and I would agree that an experienced player would see only those scenarios you mentioned as threats, but I think you're underestimating the learning curve of the game. And I think it's worth noting that a shaft event can easily put a player in front of the scenarios you mentioned before they have the resources to deal with them.

Everyone goes into rogue-likes with the understanding that dying is a part of the game and it's the biggest contributor towards learning how to win, but shaft events generated early in the game don't teach the player anything, especially when they are unavoidable. I made the same argument earlier in the thread and I think it stands pretty well.

In the past, the game used to be even less smooth in difficulty in my experience. I do not mind that many cases had been removed that created "unfair" situations, because they not really created interesting gameplay in the first place. (So I do not mind if a change makes the game somewhat easier for beginners. On the contrary.)


I'd love your opinion on my first post in this thread then, sanka. My original proposal was to shift shaft events from early game to mid-game.

However, finding yourself in an unexplored level, without a way to go back, is interesting! This is an unusual challenge you should welcome! Opportunities to improve!


We just have different opinions on what's interesting. I enjoy the exploration aspect and seeing what the randomly generated seed has to offer. I'll admit that in the rare circumstances where I got shafted and missed a portal, I enjoyed the game less. I also, anecdotally, have felt that shaft events contributed to a large portion of my losses, whether it was from just being dumped into a challenging situation or because of the amount of resources I had to use to survive that initial bad situation. But that's subjective and just my opinion.

EDIT: I had to look up what immersive means, but come on, I play go and there is a lot of tension in it sometimes.


In fairness, you're probably playing against another human in a game of go, which provides a different sort of tension. Although I see your point that tension is subjective.
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Thursday, 2nd May 2019, 01:22

Re: Shafts - Too common now?

sanka is more or less on target and I basically endorse his views on the questions IveGoneSupine asks me (the point about the importance of simple, transparent rules is particularly good), but I would add to the d1-d15 shaft scenario that this is a highly, highly improbable situation and that makes it more interesting. The only reason it might be worse for a new player is that a new player may not recognize the significance of the turn of events as they unfold so that a unique and valuable game would go unremarked. Of course, a newer player would probably be less able to cope with such a bizarre situation and make a worse show of it than a more seasoned player, but that's how it goes.

There's a reference here to something I said in another thread about arguments from the perspective of a new player, which IveGoneSupine characterizes as excusing "poor design." By far the most common way people excuse poor design is by saying things should stay the way they are or go back to the way they had been. My position on shafts, again, is that the effect is good but the delivery mechanism is bad. Ideas like escape hatches only (I believe this exists in x-crawl), downstairs only (the one true solution, available in hellcrawl), and shafts as branch entrances or occasional stair replacements (a moderate, new proposal so far unimplemented) are the way forward. It is not my opinion that new players adapt to the environment the game presents. That is a fact. New player frustration, in my opinion, is almost entirely a function of the strange front-loadedness of difficulty in dcss.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

For this message the author tealizard has received thanks:
duvessa

Lair Larrikin

Posts: 23

Joined: Saturday, 6th April 2019, 23:22

Post Thursday, 2nd May 2019, 02:44

Re: Shafts - Too common now?

I don't understand. You say there is a front loaded difficulty in DCSS, but propose that having only downstairs be the solution. Wouldn't that add to the front loaded difficulty? You're taking away one of the few intuitive ways a new player knows how to escape a bad situation.

It is not my opinion that new players adapt to the environment the game presents. That is a fact.


Not the ones that quit, my friend. Nor squeaky wheels like myself, haha
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Thursday, 2nd May 2019, 02:51

Re: Shafts - Too common now?

I did not say that no-downstairs is the solution to front-loaded difficulty.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

Blades Runner

Posts: 616

Joined: Thursday, 25th October 2012, 03:19

Post Thursday, 2nd May 2019, 04:26

Re: Shafts - Too common now?

tealizard wrote: I would add to the d1-d15 shaft scenario that this is a highly, highly improbable situation and that makes it more interesting.


Would you be a proponent of adding to shafts that each time you are shafted somewhere between D1:-D:3, that there is a 1 in 100,000 chance that you are immediately teleported next to Cerebov in Pandemonium? This would be highly improbable. Does that improbability make it interesting? If you find it interesting, does that also mean that you feel it would enhance the game?

For this message the author svendre has received thanks:
TheMeInTeam
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Thursday, 2nd May 2019, 04:36

Re: Shafts - Too common now?

See, this is what I'm talking about re: not appreciating the actual probabilities involved.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

For this message the author tealizard has received thanks:
duvessa

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Thursday, 2nd May 2019, 04:42

Re: Shafts - Too common now?

Interesting thing about probabilities from real life is that scientists pay unproportionally higher attention to events which can destroy mankind than their probabilities deserve. Just because the consequences are too harsh. Scientists don't say "we can ignore those events because they are very unlikely to occur".
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Thursday, 2nd May 2019, 05:01

Re: Shafts - Too common now?

Annoying thing about shatfs is missing times portals. Broken thing about portals is being shafted 2 floors when you have just entered a level or even started playing. Bad thing about shafts is that it punishes autoexplore and greatly favors from having scroll of magic mapping. Interesting thing about shafts is that there are no stairs in view and you can be surrounded. There is a solution which keeps good things and removes bad things:

When you enter a level for the first time, there is a chance that you will be instantly teleported away from those stairs. No more consequtive shafts, dependence on mapping is decreased, timed portals are still on this level, no XL 1 characters on D:3, autoexplore is not affected, you still can explore below full HP.
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25

Lair Larrikin

Posts: 23

Joined: Saturday, 6th April 2019, 23:22

Post Thursday, 2nd May 2019, 05:37

Re: Shafts - Too common now?

tealizard's concern about statistical probability is a fair one. The D1 to D15 shaft scenario was just something sanka was using to illustrate their appreciation of the shaft mechanic to me. It's unlikely to the point of never happening. Just a thought experiment we were using to see where each of us were coming from.

tealizard wrote:I did not say that no-downstairs is the solution to front-loaded difficulty.


Yeah. But my question was why propose a solution to the shaft issue that would make a related problem, that you acknowledge, worse? Maybe I'm not understanding you and you don't see new player frustration as a problem. Or maybe you don't see the two as related. I'm just trying to get a good grasp on where you're coming from.

VeryAngryFelid wrote:Annoying thing about shatfs is missing times portals. Broken thing about portals is being shafted 2 floors when you have just entered a level or even started playing. Bad thing about shafts is that it punishes autoexplore and greatly favors from having scroll of magic mapping. Interesting thing about shafts is that there are no stairs in view and you can be surrounded. There is a solution which keeps good things and removes bad things:

When you enter a level for the first time, there is a chance that you will be instantly teleported away from those stairs. No more consequtive shafts, dependence on mapping is decreased, timed portals are still on this level, no XL 1 characters on D:3, autoexplore is not affected, you still can explore below full HP.


That is a little bit of cool suggestion. Teleport trap on stair down. I'm going to spend some time thinking on that one. Maybe I'll look through the boards to see prior conversations about teleport traps or teleportitis.
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Thursday, 2nd May 2019, 06:34

Re: Shafts - Too common now?

As far as the shaft effect goes, the first thing that happens to a player when they start the game is they get "shafted" into d:1, i.e. if they do not fight their way out of the immediate vicinity of the entrance, they lose the game. Incidentally, this is where the vast majority of unwinnable game scenarios happen. It's also the site of some of the most interesting gameplay in the game. But because of the way it is presented to the player, a new player is unlikely to think of it the way I have described it here. And indeed, I've only ever seen fairly experienced players talk about the importance of the dungeon entrance. Therefore I doubt that new players have a hard time accepting the dungeon entrance or that they would have a hard time accepting a similar arrangement on every floor. People cling to stairs because it is what they know, not because they have an innate intuition that they should be able to cheez their way through the game with stairs.

What people actually balk at is the trap mechanism applied to the shaft effect. This is because it cuts against the sense of what is normal they develop through play. Even though no-upstairs is objectively more challenging than 2.5 shafts per game, people would accept it more readily if it were the default mode of crawl.

As for how to address the difficulty of the early game, the root of the problem is that the player character is fairly weak compared to worst case spawns. That is easy to address. The next layer of the problem is that the rest of the game is too easy so you end up with a bland game experience. The best solution to this is increasing the difficulty of the rest of the game, which is the direction hellcrawl pursues.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

For this message the author tealizard has received thanks:
stormdragon

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Thursday, 2nd May 2019, 09:07

Re: Shafts - Too common now?

Front loaded difficulty is caused by many factors, I may even miss some:

1. To make the game winnable for beginners, it must be winnable with bad skilling/strategy (I use strategy mean long term decisions like equipment, etc. as opposed to tactics here). Currently it is. But since we want to make skill decisions relevant, this means that any charcter that has somewhat sane skilling will be much stronger then the enemies.

2. Consumables accumulate. Worse players need them more, if we want them to have enough later then as you get better you will have so huge amount that makes many situations trivial (if you actually use them).

3. Crawl is non-linear. This has bad consequences, and the impact is very clear, as usually the game suddenly becomes easier when you find the first branch (lair), or maybe a few levels after it. It is not the fact that you can go multiple ways matters, but that there is no very clear order of difficulty. This is bad, because if there are two levels of the same difficulty, you clear the first, get stronger, and the second becomes easier: the earlier levels will be the harder ones. Crawl has many, many such levels: after lair the rest of the dungeon is usually easier than it should be.

4. Gods are really unbalanced.

Removing shafts from early dungeon will not help this at all IMHO, their impact is way too low on difficulty.

For this message the author sanka has received thanks:
duvessa
User avatar

Vaults Vanquisher

Posts: 454

Joined: Thursday, 1st November 2018, 02:33

Post Thursday, 2nd May 2019, 09:33

Re: Shafts - Too common now?

I just mean the hardest part of the game is the first few floors. The factors you're talking about contribute to the fall off in difficulty later and the weird learning curve.
This is where mechanical excellence and one-thousand four-hundred horsepower pays off.

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Thursday, 2nd May 2019, 09:53

Re: Shafts - Too common now?

Right - you are right both in that the first few floors are the hardest, and beginners rarely notice it, and that the reasons I listed are about a different fall off, which happens later.
User avatar

Shoals Surfer

Posts: 301

Joined: Saturday, 21st May 2011, 08:23

Post Thursday, 2nd May 2019, 10:20

Re: Shafts - Too common now?

svendre wrote:I don't think that the rarity of a chain-shafting is completely relevant. You could take my awful example of 1 in 3 steps cause you to instantly die, and say that this mode activates in only 1 of 100,000 games, and it would be super rare to die from it, but it still wouldn't be a good example of difficulty which is ideal.


No one is saying this. Why do you keep making up silly arguments and then shutting them down?

Clearly an insta-death mode has zero value, and shafting has some value. Whether that's worth keeping them, I don't know, but it's not the same thing at all.
(p.s. this is stupid some dev please make it not stupid) - minmay

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Thursday, 2nd May 2019, 10:31

Re: Shafts - Too common now?

sanka wrote:Front loaded difficulty is caused by many factors, I may even miss some:

1. To make the game winnable for beginners, it must be winnable with bad skilling/strategy (I use strategy mean long term decisions like equipment, etc. as opposed to tactics here). Currently it is. But since we want to make skill decisions relevant, this means that any charcter that has somewhat sane skilling will be much stronger then the enemies.

2. Consumables accumulate. Worse players need them more, if we want them to have enough later then as you get better you will have so huge amount that makes many situations trivial (if you actually use them).


This is not necessarily true.
1. Easy combos like MiBe can (not must) be winnable but hard combos (MuAM, for instance) require very good play.
2. I am not a bad player but I ran out of consumables with some combos (I remember at least FeXX and DrEn) and died because of that. My point is resource accumulation itself is not a problem, the player always can switch to harder combo which better suits their level.

I have impression that some players want MiBe become hard and non-streakable even for top players, this is a very bad idea provided crawl does not have difficulty levels other than combos.
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25

Blades Runner

Posts: 616

Joined: Thursday, 25th October 2012, 03:19

Post Thursday, 2nd May 2019, 14:12

Re: Shafts - Too common now?

I was saying that something isn't interesting (lending itself well to the game) merely because it is improbable. That *was* said directly, even quoted. The rarity was also separately used as part of some points in the defense of chain-shafting.

Technically in my seriously extreme example, it was not "insta-death" because despite the extreme odds, a person could roll the 1 in 3 over and over and survive. That it came close to an almost certain death, statistically, but not absolutely was significant, This distinction and why I made it seems to have gone right over some heads.

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Thursday, 2nd May 2019, 14:22

Re: Shafts - Too common now?

VeryAngryFelid: I think that if you play by conducts this is often true. Like you use autoexplore too much, you use a suboptimal god, you do not kite enough (especially with those damned cats), etc.

Felids are actually a very good example of the problems. Felid is NOT a hard race in the late game. However, it suddenly becomes a hard race if you use autoexplore.

If you do not, I think that mummies and felids still have an easier time later on (even if many cats are extremely, extremely annoying to play properly: you are fast but have few hp and only dodging as a defence = most degenerate playstyle), exactly due to the reasons I mentioned.

However, in a design forum I would like if there are mechanics which reduce the impact of these (so in a way I propose to make a game easier), and make it harder at the same time.

I do not think the aim is to make the experience of playing MiBe significantly different, even if, objectively, some modifications would make the game harder for it by not allowing as much kiting. Usually we do not use it anyway with MiBe because it is annoying and it is not needed anyway, the monsters are so pathetic.
Last edited by sanka on Thursday, 2nd May 2019, 14:46, edited 2 times in total.

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Thursday, 2nd May 2019, 14:37

Re: Shafts - Too common now?

VeryAngryFelid: one more thing came into my mind.

You keep worrying about not having an "easy" mode after making the game too hard by shaft, cutting down extra branches, downstairs only, etc.
You propose conduct plays and extra races instead.

I have too counterarguments:
1. Conducts that are not implemented inside the game are against the game philosophy. The game should be enjoyable without conducts. The problem is not that just that it is too easy - it is also that it is easy in the wrong way, the good tactics are not interesting gameplay.
2. Fixing this with harder races is really hard, as you could probably guess from my "Down dwarf" proposal. Simply giving less HP, while would make the game easier, is not a good solution to gameplay problems. Everybody tries to tell you that it is not the difficulty level that is the main target!

Instead, I propose going the other way around, namely:
1. Change the game towards better challenges: limit kiting, stair dancing, fix the difficulty curve.
2. Fix the "easy" races so that they remain easy! Because this is much easier. As opposed to giving less HP for a hard species (does not work, because you will get under one-shot range for many monsters, really changing good strategies for very annoying ones), giving more hp, better apts, etc. to minotaur does work.

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Thursday, 2nd May 2019, 14:48

Re: Shafts - Too common now?

sanka wrote:VeryAngryFelid: one more thing came into my mind.

You keep worrying about not having an "easy" mode after making the game too hard by shaft, cutting down extra branches, downstairs only, etc.
You propose conduct plays and extra races instead.


Well, my core proposal (several years ago) was to implement explicit difficulty levels but it was refused and I was told that crawl uses "species" as difficulty levels so yes, now I am proposing to add "impossible" difficulty levels via intended way - "impossible" species. Conduct is not a good way to setup difficulty, I agree with you here.

I have too counterarguments:
1. Conducts that are not implemented inside the game are against the game philosophy. The game should be enjoyable without conducts. The problem is not that just that it is too easy - it is also that it is easy in the wrong way, the good tactics are not interesting gameplay.
2. Fixing this with harder races is really hard, as you could probably guess from my "Down dwarf" proposal. Simply giving less HP, while would make the game easier, is not a good solution to gameplay problems. Everybody tries to tell you that it is not the difficulty level that is the main target!

1. I don't care about statistics so conducts are fine IMHO in single player game.
2. Fixing that is very easy, for instance, I had a branch where player might change PC/monster player max HP. If we can add species which is slower than Na, noisier than Qazlal at max peity, has lower HP than Fe etc., the game will be a nightmare even for top players and some mechanics like luring, stair/pillar dancing will be completely impossible.

However, going the other way around, namely:
1. Change the game towards better challenges: limit kiting, stair dancing, fix the difficulty curve.
2. Fix the "easy" races so that they remain easy! Because this is much easier. As opposed to giving less HP for a hard species (does not work, because you will get under one-shot range for many monsters, really changing good strategies for very annoying ones), giving more hp, better apts, etc. to minotaur does work.

1. Reducing max HP is not required, just make every aptitude -8 and speed 50 auts per move. Only partially a joke.
2. I believe we should not make game harder, it is already unwinnable by some players who play it for years. I mean if you want to have "impossible" difficulty level, it should not change "easy" difficulty in any way. But currently we have "top players still can streak NaWn, let's add a monster with 120 damage per attack", nobody seems to be worried that it will also affect players who cannot win even MiBe.
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Thursday, 2nd May 2019, 14:53

Re: Shafts - Too common now?

Speaking about difficulty levels in roguelikes I have just found https://www.gridsagegames.com/blog/2017 ... ifficulty/, looks interesting IMHO.

That said, it’s probably not a coincidence that two widely popular roguelikes also offer multiple difficulties.


From author of TOME:
Also to note that my difficulty levels are not simple number bumps (there are those too obviously), each level introduces new harder mechanics (like random bosses, being constantly hunted, …). To be fair I do not understand why devs do NOT put difficulty levels in their games heh ;)


From author of DoomRL:
It raises the replayability of the game, allowing you to go more difficult once you feel that the game is getting to easy. Initially it was added just as a tribute to Doom, but it turned out to be a great feature, so I gave it a lot more thought later on.
Last edited by VeryAngryFelid on Thursday, 2nd May 2019, 14:56, edited 2 times in total.
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Thursday, 2nd May 2019, 14:55

Re: Shafts - Too common now?

svendre:

I think that there are two counterarguments against "fixing" rare extreme shafting deaths, both suppose that they are really rare.

1. Simply shafting, even multiple times, even to D:15 does not kill you. And even in these situations, better play gives you better chances to survive. So the distribution is that while the challenge can go up to lethal, but with really, really dropping probabilities. Having rare and challenging (as opposed to 100% deadly) challenges are ok. Even if you cannot reduce the probability of dying to 0%, it is still interesting to try.

2. Cutting down this distribution somewhere feels arbitrary. That is, it complicates the rules, introduces hidden states, complicates the codebase, may interact with other mechanism in a bad way, etc. It is also really hard to cut in the right place: most likely you get rid of some lethal situations while preserving others, while getting rid of interesting challenges as well.

As an example, not allowing multiple shafts introduces a state of the depth of your last shaft. Extra state usually considered bad for aforementioned reasons.
You still could easily die unfairly by falling down into the middle of a vault or group of bees, etc., with one shaft, so you did not achieved the goal of "getting rid of unfair deaths by shafts".
You got rid of multiple shafts where you actually had a chance to survive with good tatctics, because multiple shafts did not kill you alone. You may land in a relatively good postion and try to make back your way: since it is really dangerous indeed, it is rewarding if you are successful!

Abyss Ambulator

Posts: 1131

Joined: Tuesday, 4th January 2011, 15:03

Post Thursday, 2nd May 2019, 15:09

Re: Shafts - Too common now?

VeryAngryFelid wrote:But currently we have "top players still can streak NaWn, let's add a monster with 120 damage per attack", nobody seems to be worried that it will also affect players who cannot win even MiBe.


I think nobody proposing adding such a monster in the dungeon. Shafts, removing upstairs are really different. I specifically said that monsters which can kill you with one shot are not very good enemies, because the tactics you can use against them are very limited.

I haven't tried the super bad race you propose (if you have source code I may compile at one point), but I doubt that they will lead to significantly better gameplay.
Last edited by sanka on Thursday, 2nd May 2019, 15:19, edited 1 time in total.

Ziggurat Zagger

Posts: 4432

Joined: Friday, 8th May 2015, 17:51

Post Thursday, 2nd May 2019, 15:14

Re: Shafts - Too common now?

sanka wrote:
VeryAngryFelid wrote:But currently we have "top players still can streak NaWn, let's add a monster with 120 damage per attack", nobody seems to be worried that it will also affect players who cannot win even MiBe.


I think nobody proposing adding such a monster in the dungeon. Shafts, removing upstairs are really different. I specifically said that monsters which can kill you with one shot are not very good enemies, because the tactics you can use against them are very limited.
http://crawl.chaosforge.org/Juggernaut
I would not care about shafts and removing stairs if it was on the hardest difficulty levels only or ideally customizable in game options but since it affects every combo it will make MiBe levels combos even harder than they are for players who never won.
Underestimated: cleaving, Deep Elf, Formicid, Vehumet, EV
Overestimated: AC, GDS
Twin account of Sandman25
PreviousNext

Return to Game Design Discussion

Who is online

Users browsing this forum: No registered users and 28 guests

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