Difference between revisions of "Yet Another Stupid Death"

From CrawlWiki
Jump to: navigation, search
m (Preventing Hubris)
m (Adjust Your Threat Assessment)
 
(4 intermediate revisions by 2 users not shown)
Line 11: Line 11:
  
 
If you do decide to engage in slow play, it should be methodical. You should be thinking about (at least) two things:
 
If you do decide to engage in slow play, it should be methodical. You should be thinking about (at least) two things:
# What can monsters do to you 1 turn, or 2 turns, from now? Can they kill you if they do maximum damage? Think about effects that change your position, any [[status effect]]s monsters can inflict on you, as well as the max damage you can take. ('''N.B.''' In the late-game: when fighting many monsters, especially with high [[AC]], the chance of taking max damage is incredibly small. The chance of taking max damage twice, even less likely. Thus you may find it acceptable to continue fighting even in "potential one-turn death range".)
+
# What can monsters do to you 1 turn, or 2 turns, from now? Can they kill you if they do maximum damage? Also, think about effects that change your position, and any [[status effect]]s monsters can inflict on you.<br>('''N.B.''' In the late-game: when fighting many monsters, especially with high [[AC]], the chance of taking max damage is incredibly small. The chance of taking max damage twice, even less likely. Thus you may find it acceptable to continue fighting even in "potential one-turn death range".)
 
# What options do you have? If you spot trouble early, you may want an effective-but-cheap option. If you're at 1 HP, you probably want to use a [[scroll of blinking]].
 
# What options do you have? If you spot trouble early, you may want an effective-but-cheap option. If you're at 1 HP, you probably want to use a [[scroll of blinking]].
  
Also, you may want to adjust your [[rcfile]]. If you make autofight slower, you may catch mistakes you otherwise would've missed. There are two options of note: <code>autofight_stop = X</code>, which is the % HP where autofight cannot be used, and <code>autofight_warning = X</code>, a delay (in milliseconds) between autofight commands.
+
In addition you may want to adjust your [[rcfile]]. If you make autofight slower, you may catch mistakes you otherwise would've missed. There are two options of note: <code>autofight_stop = X</code>, which is the % HP where autofight cannot be used, and <code>autofight_warning = X</code>, a delay (in milliseconds) between autofight commands.
  
 
===Adjust Your Threat Assessment===
 
===Adjust Your Threat Assessment===
One way experienced players avoid YASD... is experience. More specifically, the experience in "knowing what and what isn't a threat". A player might autoexplore, find a D:4 [[ogre (monster)|ogre]], and think "D:4 ogre? That's too risky to allow into melee range - it can do 37-39 damage per turn, enough to kill me in two hits". By constantly iterating this kind of thought process over time, one can play fast ''and'' consistently.
+
One way experienced players avoid YASD is experience - more specifically, the experience in knowing what is and what isn't a threat. A player might autoexplore, find a D:4 [[ogre (monster)|ogre]], and think "D:4 ogre? That's too risky to allow into melee range - it can do 37-39 damage per turn, enough to kill me in two hits". By constantly iterating this kind of thought process over time, one can play fast ''and'' consistently.
  
 
Like before, if you play 1000 games and don't learn anything from them, being "experienced" won't be of much help.
 
Like before, if you play 1000 games and don't learn anything from them, being "experienced" won't be of much help.
* One way to help is to look back at your deaths. If you play online, you might be able to download "ttyrecs" and play them with [https://github.com/Rytisgit/DCSSReplay external software]. It is recommended to go turn-by-turn: try to find the moment where you go from "you can always escape (even if you need items)", to "you have a chance of dying, no matter what". Then, go back and see what you could've done a turn before to prevent it, 5 turns before to prevent it, etc.
+
*One way to help is to look back at your deaths. If you play online, you might be able to download "ttyrecs" and play them with [https://github.com/Rytisgit/DCSSReplay external software]. It is recommended to go turn-by-turn: try to find the moment where you go from "you can escape with 0% chance of death (even if you need items to do so)", to "you have a chance of dying, no matter what". Then, go back and see what you could've done a turn before to prevent it, 5 turns before to prevent it, etc.
* Then, apply what you learned. Sounds simple, since it is. But make sure your theory applies in practice. You may think of an idea after a post-YASD analysis, then that idea doesn't hold up in actual gameplay.
+
*Then, apply what you learned. Sounds simple, since it is. But make sure your theory applies in practice. You may think of an idea after a post-YASD analysis, then that idea doesn't hold up in actual gameplay.
* Also, look for advice, whether on the wiki or from other players. They can provide a shortcut towards experience, e.g. give tips on what is a threat.
+
*Also, look for advice, whether on the wiki or from other players. They can provide a shortcut towards experience, e.g. give tips on what is a threat.
  
 
==External Links==
 
==External Links==

Latest revision as of 22:12, 2 June 2024

Yet Another Stupid Death (commonly abbreviated as YASD) is used when a player has done something, usually avoidable, that leads to his or her demise. Originated in NetHack.

It is often caused by hubris - an overestimation of one's player skill or character strength.

Preventing Hubris

This article contains advice from other players, which may be subjective, outdated, inaccurate or ill-advised. Take advice as you see fit, and read at your own risk!

Due to Crawl's gameplay loop, it's easy to experience hubris. By the mid-game, you will constantly encounter, and subsequently mow down, enemies who are trivial to fight. This encourages you to play fast, this makes players feel strong. And for a while, mindlessly autoexploring and autofighting does work. But it's only a matter of time before bad RNG kicks in, or you see a monster that's too difficult. If (and only if) you continue to fight recklessly, you will die.

Play Less Quickly

A common advice is to "play slower". This can work - and often does - but not without caveats. If you play slower without knowing why you are playing slower, you've done nothing productive. Also, playing slower can make you impatient down the line, causing the opposite effect.

If you do decide to engage in slow play, it should be methodical. You should be thinking about (at least) two things:

  1. What can monsters do to you 1 turn, or 2 turns, from now? Can they kill you if they do maximum damage? Also, think about effects that change your position, and any status effects monsters can inflict on you.
    (N.B. In the late-game: when fighting many monsters, especially with high AC, the chance of taking max damage is incredibly small. The chance of taking max damage twice, even less likely. Thus you may find it acceptable to continue fighting even in "potential one-turn death range".)
  2. What options do you have? If you spot trouble early, you may want an effective-but-cheap option. If you're at 1 HP, you probably want to use a scroll of blinking.

In addition you may want to adjust your rcfile. If you make autofight slower, you may catch mistakes you otherwise would've missed. There are two options of note: autofight_stop = X, which is the % HP where autofight cannot be used, and autofight_warning = X, a delay (in milliseconds) between autofight commands.

Adjust Your Threat Assessment

One way experienced players avoid YASD is experience - more specifically, the experience in knowing what is and what isn't a threat. A player might autoexplore, find a D:4 ogre, and think "D:4 ogre? That's too risky to allow into melee range - it can do 37-39 damage per turn, enough to kill me in two hits". By constantly iterating this kind of thought process over time, one can play fast and consistently.

Like before, if you play 1000 games and don't learn anything from them, being "experienced" won't be of much help.

  • One way to help is to look back at your deaths. If you play online, you might be able to download "ttyrecs" and play them with external software. It is recommended to go turn-by-turn: try to find the moment where you go from "you can escape with 0% chance of death (even if you need items to do so)", to "you have a chance of dying, no matter what". Then, go back and see what you could've done a turn before to prevent it, 5 turns before to prevent it, etc.
  • Then, apply what you learned. Sounds simple, since it is. But make sure your theory applies in practice. You may think of an idea after a post-YASD analysis, then that idea doesn't hold up in actual gameplay.
  • Also, look for advice, whether on the wiki or from other players. They can provide a shortcut towards experience, e.g. give tips on what is a threat.

External Links