Early Game Myth


If it doesn't fit anywhere else, it belongs here. Also, come here if you just need to get hammered.

Swamp Slogger

Posts: 147

Joined: Sunday, 14th June 2015, 07:19

Post Saturday, 12th September 2015, 11:23

Early Game Myth

So a lot of crawl 'wisdom' I see propogated here has a tendency to get hyperbolized to memetic proportion, and rules of thumb get exaggerated to statements of absolute fact about the game, and it seems like people have started actually taking them seriously and not as the tongue-in-cheek arguments that they are, with the one most commonly getting flaunted as the idea that the early game is the toughest, or even the only challenging part of the game, and that by reaching the staircase to Lair victory is a foregone conclusion if a player is worth a lick of salt.

I'm really curious what version of trunk I'm playing that's different than everyone else's, because when I reach Lair I'm not handed a stack of every consumable I need to escape any dangerous situation through the rest of the game, along with all the resists and equipment I need to get through Zot. And the version of Lair I get introduces enemies that can outrun you like Spiny Frogs, Black Mambas and Blink Frog packs that I can't just walk away from. Just as well though, as one-tile wide narrow hallways to funnel enemies are a rarity in Lair, so often I'm going to be facing packs of Blink Frogs, Death Yaks and Elephants that I can't turn into a series of one-on-one fights anyways. Don't forget new threats like basilisk petrification, hydras that will rip me to shreds if I don't have the defenses and offenses to handle them, and poison that I very likely do not have resistance to.

And believe me, I'm not trying to say that Lair is too hard, I'm not even trying to say that it's hard at all. But its silly that 'conventional wisdom' says this is a cakewalk after early Dungeon, where I can close doors, walk away into narrow corridors, and very little can walk faster than me.

And seriously, 'the early game is the hardest part of the game?' For races and backgrounds that lack useful tools at the start, sure. But pick a Minotaur, a Berserker, or Zot forbid, both, and substantiate the claim D1-D9 is the most challenging part for that build.

The main argument that I understand made that the game is 'won' by Lair or that the early game is the most difficult part comes from the fact that situations where the game is 'unwinnable' are only really seen very early (inescapable adders with bad rolls, autoexplore into banishment / multiple orc priests / Grinder paralysis / gnolls with branded polearms), whereas later in the game, if you die its almost always very clearly your fault for mishandling the situation. This ends up getting extrapolated far beyond the scope of that argument into a bizzare hypothetical scenario where every player has perfect knowledge of every situation and makes 'optimal' decisions at all times, when that can only apply to perhaps a handful of players at best, and even they make mistakes, misread the situation, or encounter unexpected situations.

Late game offers players more ways to circumvent trouble, but it throws far, far more trouble at players than the early game does. Pretending that past Lair the game doesn't pose any challenge to me more like a seductive, self-congratulatory elitist claim than a fair evaluation of the game's difficulty.

Feel free to claim that I only hold this opinion because I'm bad at the game. I won't deny that I am!

For this message the author Aethrus has received thanks: 2
JFunk, ololoev
User avatar

Dis Charger

Posts: 1949

Joined: Monday, 18th February 2013, 07:59

Location: France

Post Saturday, 12th September 2015, 11:40

Re: Early Game Myth

Sequell wrote:2400 games for mopl (!won): 594x D:1, 424x D:2, 341x D:3, 235x D:4, 134x D:5, 89x D:6, 59x D:7, 46x D:8, 41x D:10, 37x Abyss:1, 36x D:9, 26x Lair:1, 23x Orc:4, 18x D:$, 17x Ossuary, 17x IceCv, 14x D:11, 13x Lair:3, 12x Orc:3, 12x Sewer, 11x Elf:3, 11x Lair:2, 11x Bailey, 10x D:12, 10x Lab, 9x Abyss:3, 8x Abyss:2, 7x Lair:7, 6x Vaults:5, 6x Snake:5, 6x Lair:5, 5x Lair:6, 5x Orc:2, 5x Lair:8, 5x Orc:1, 5x D:13, 5x Lair:4, 5x Vaults:3, 4x Shoals:5, 4x Swamp:2, 4x Swamp:5, 4x D:15, 4x Volcano, 3x Depths:1, 3x Elf:1, 3x Slime:6, 3x Vaults:4, 3x Snake:2, 3x Dis:7, 3x Swamp:4, 2x Vaults:2, 2x Crypt:1, 2x Depths:3, 2x Spider:5, 2x D:14, 2x Temple, 2x Snake:3, 2x Zig:12, 2x Shoals:3, Tomb:1, Tar:2, Zig:19, Abyss:4, D:17, Snake:1, Geh:7, Vaults:1, Depths:2, Zot:4, Elf:2, Zig:24, Spider:1, WizLab, Slime:3, Tomb:3, Zig:5, Spider:2, Zot:5, Swamp:3, Slime:5, Abyss, Swamp:1
1922 out of 2400 death...
Online stats
Fastest Hell runes (enter Hell branch -> get the rune)
Icy : 56 / Iron : 126 / Obsidian : 215 / Bone : 125

For this message the author mopl has received thanks: 2
greedo, Sar

Swamp Slogger

Posts: 147

Joined: Sunday, 14th June 2015, 07:19

Post Saturday, 12th September 2015, 11:47

Re: Early Game Myth

mopl wrote:
Sequell wrote:2400 games for mopl (!won): 594x D:1, 424x D:2, 341x D:3, 235x D:4, 134x D:5, 89x D:6, 59x D:7, 46x D:8, 41x D:10, 37x Abyss:1, 36x D:9, 26x Lair:1, 23x Orc:4, 18x D:$, 17x Ossuary, 17x IceCv, 14x D:11, 13x Lair:3, 12x Orc:3, 12x Sewer, 11x Elf:3, 11x Lair:2, 11x Bailey, 10x D:12, 10x Lab, 9x Abyss:3, 8x Abyss:2, 7x Lair:7, 6x Vaults:5, 6x Snake:5, 6x Lair:5, 5x Lair:6, 5x Orc:2, 5x Lair:8, 5x Orc:1, 5x D:13, 5x Lair:4, 5x Vaults:3, 4x Shoals:5, 4x Swamp:2, 4x Swamp:5, 4x D:15, 4x Volcano, 3x Depths:1, 3x Elf:1, 3x Slime:6, 3x Vaults:4, 3x Snake:2, 3x Dis:7, 3x Swamp:4, 2x Vaults:2, 2x Crypt:1, 2x Depths:3, 2x Spider:5, 2x D:14, 2x Temple, 2x Snake:3, 2x Zig:12, 2x Shoals:3, Tomb:1, Tar:2, Zig:19, Abyss:4, D:17, Snake:1, Geh:7, Vaults:1, Depths:2, Zot:4, Elf:2, Zig:24, Spider:1, WizLab, Slime:3, Tomb:3, Zig:5, Spider:2, Zot:5, Swamp:3, Slime:5, Abyss, Swamp:1
1922 out of 2400 death...


Absolutely true! Not actually relevant to my argument, which is that the fact that you're more likely to die early on statistically doesn't mean that the late game is a cakewalk or less challenging. But yes, I fully acknowledge that most players have a spread of deaths similar to that.

Crypt Cleanser

Posts: 747

Joined: Friday, 6th January 2012, 12:30

Post Saturday, 12th September 2015, 12:57

Re: Early Game Myth

Aethrus wrote:The main argument that I understand made that the game is 'won' by Lair or that the early game is the most difficult part comes from the fact that situations where the game is 'unwinnable' are only really seen very early (inescapable adders with bad rolls, autoexplore into banishment / multiple orc priests / Grinder paralysis / gnolls with branded polearms), whereas later in the game, if you die its almost always very clearly your fault for mishandling the situation. This ends up getting extrapolated far beyond the scope of that argument into a bizzare hypothetical scenario where every player has perfect knowledge of every situation and makes 'optimal' decisions at all times
No, it's really not that hard to see a bad situation coming and use a tp scroll or whatever. What you described there with early game vs late game is exactly how it is. Unless you're playing a weak character, in which case you still see the danger coming but have a harder time dealing with it, making the game harder.

I often die in the midgame, but basically every time my mistake is obvious in retrospect and was made due to carelessness.

People who say that the game after Lair doesn't pose any challenge are not correct, but I'm not sure if anyone says that.
The Lair is genuinely easy most of the time though. The sometimes hard part is killing the first one or two spiny frogs and black mambas, then you have enough xp to do the rest.

Sar

User avatar

Ziggurat Zagger

Posts: 6418

Joined: Friday, 6th July 2012, 12:48

Post Saturday, 12th September 2015, 13:16

Re: Early Game Myth

Aethrus wrote:often I'm going to be facing packs of Blink Frogs, Death Yaks and Elephants that I can't turn into a series of one-on-one fights anyways

It's actually pretty easy to turn a fight with a pack of death yaks into a series of 1v1 fights, that's what I usually do when I encounter them.
Aethrus wrote:But pick a Minotaur, a Berserker, or Zot forbid, both, and substantiate the claim D1-D9 is the most challenging part for that build.

While playing something like a MiBe does make earlygame significantly easier, it doesn't really make the rest of the game any harder.

For this message the author Sar has received thanks:
tedric
User avatar

Slime Squisher

Posts: 419

Joined: Wednesday, 21st September 2011, 09:45

Post Saturday, 12th September 2015, 13:42

Re: Early Game Myth

If I just throw in an orc priest for obvious an example.

There is NOTHING in the midgame/lategame even remotely as dangerous as an orc priest for a 30-50 HP character. Well A-liches may qualify but they don't have irresistible and unavoidable attacks afaik.

When orc priest threat (among other earlygame dangers) is gone with higher damage and more HP, there are usually MR problems and possible disto weapons left. Once I've overcome those, game is significantly easier. Sometimes RNG doesn't give you good resistances, absolutely deadly weapon, top wands etc etc. But I have not played a game yet where I could not win a 3 rune game because floor god is just not giving me good enough items.

Spider Stomper

Posts: 220

Joined: Sunday, 26th July 2015, 15:38

Post Saturday, 12th September 2015, 16:03

Re: Early Game Myth

Aliches can also be silenced, turning them into little more than skeletons.

But take what mopl wrote, ti's actually even worse than it looks at first glance. You shouldn't be looking at danger across the whole game so much as by room. How dangerous is Vault:1?

Well his statistics says it's 594 times easier than d:1. This is what people mean when they say the game is over by lair for experienced players. You CAN die, but the danger and excitement is generally long gone.
Last edited by greedo on Saturday, 12th September 2015, 16:08, edited 1 time in total.

Shoals Surfer

Posts: 297

Joined: Wednesday, 9th July 2014, 08:20

Post Saturday, 12th September 2015, 16:07

Re: Early Game Myth

The idea, I think, is that Crawl is an easy game and under ideal conditions you would virtually never die except in situations that kill you before you have an answer for them, and these situations occur almost entirely in the early game.

I disagree, though. Although it is true that very few encounters would kill a decently prepared player if they are willing to run away from the slightest hint of danger, it is not Crawl's fault that it fails to be a difficult and engaging game for lily-livered cowards like that. It fills me with disgust that a player, for example, would stoop so low as to avoid meleeing hydras with a bladed weapon instead of trying to outdamage their head increase, or to "ninja" (a reprehensible practice in itself) late game runes from their guardians without slaying them in honorable combat. The game may be made much more entertaining by simply avoiding such degenerate behavior.

I think steps need to be taken to remove this kind of gameplay. I recommend changing all sources of teleport to the new yanks-you-towards-enemies ring teleportation (with or without the "fails when it wouldn't yank you towards enemies" bit), altering ?blink into ?blinkclose so that it can be used to murder your enemies sooner but not to avoid danger, and adding an obsidian axe-like mesmerization effect to any ability that increases your speed. Upstairs should have a flat 75% chance to crumble if you're trying to use them under tension. Elf should be forced after Orc, Crypt and Tomb should be forced while in Vaults.

For this message the author Croases has received thanks:
Sar

Spider Stomper

Posts: 220

Joined: Sunday, 26th July 2015, 15:38

Post Saturday, 12th September 2015, 16:15

Re: Early Game Myth

Ninjaing is actually more likely to result in death btw (without cTele), especially the orb. People do it because it's more fun :)

Tomb Titivator

Posts: 909

Joined: Thursday, 3rd January 2013, 20:32

Post Saturday, 12th September 2015, 17:06

Re: Early Game Myth

The claim "if you've made it to lair, your chance of an unavoidable death is infinitesimally small" is not the same as the claim "nothing past the Lair staircase requires any thought or careful play to survive."

I'm pretty sure there are some players who agree with the second statement (or at least have voiced it in a snarky, hyperbolic way) -- but to the extent that anyone says it seriously around here, it's pretty clearly a product of experience bias.

The first statement is hard to prove empirically, but I think any player who assumes it as a precept and then analyzes their deaths from that perspective will soon have enough anecdotal data to support it as fact.
Wins (Does not include my GrEE^Veh 15-runer...stupid experimental branch)
User avatar

Dis Charger

Posts: 1949

Joined: Monday, 18th February 2013, 07:59

Location: France

Post Saturday, 12th September 2015, 17:07

Re: Early Game Myth

greedo wrote:Ninjaing is actually more likely to result in death btw (without cTele), especially the orb. People do it because it's more fun :)
And it can get you a better score...
Online stats
Fastest Hell runes (enter Hell branch -> get the rune)
Icy : 56 / Iron : 126 / Obsidian : 215 / Bone : 125

Blades Runner

Posts: 552

Joined: Tuesday, 10th April 2012, 21:11

Post Saturday, 12th September 2015, 17:19

Re: Early Game Myth

Aethrus wrote:So a lot of crawl 'wisdom' I see propogated here has a tendency to get hyperbolized to memetic proportion, and rules of thumb get exaggerated to statements of absolute fact about the game, and it seems like people have started actually taking them seriously and not as the tongue-in-cheek arguments that they are...


I don't understand why people come here with this attitude. The rules of thumb for the most part are quite practical and instructive, and you should be happy that they are stated at all. If you want only "absolute facts" and "serious arguments", go ahead and do endless trial and error or start diving code. And if you dislike using the experiences and insights of people who have played thousands of games, then don't pay attention to them. Meanwhile, people who take what is useful instead of complaining will improve and win games.
User avatar

Ziggurat Zagger

Posts: 4478

Joined: Wednesday, 23rd October 2013, 07:56

Post Saturday, 12th September 2015, 17:27

Re: Early Game Myth

Croases wrote:It fills me with disgust that a player, for example, would stoop so low as to avoid meleeing hydras with a bladed weapon instead of trying to outdamage their head increase

This is easy to fix. Just make hydras grow more heads whenever they take damage, no matter how. Also make them speed 20 on land so you can't just run away.
DCSS: 97:...MfCj}SpNeBaEEGrFE{HaAKTrCK}DsFESpHu{FoArNaBe}
FeEE{HOIEMiAE}GrGlHuWrGnWrNaAKBaFi{MiDeMfDe}{DrAKTrAMGhEnGnWz}
{PaBeDjFi}OgAKPaCAGnCjOgCKMfAEAtCKSpCjDEEE{HOSu
Bloat: 17: RaRoPrPh{GuStGnCa}{ArEtZoNb}KiPaAnDrBXDBQOApDaMeAGBiOCNKAsFnFlUs{RoBoNeWi

Spider Stomper

Posts: 220

Joined: Sunday, 26th July 2015, 15:38

Post Saturday, 12th September 2015, 17:30

Re: Early Game Myth

Also give them hellfire, torment, dimensional anchor, death's door, torpor snail aura, cloud dispersal, and digging.

Zot Zealot

Posts: 1031

Joined: Friday, 26th April 2013, 19:52

Location: AZ, USA

Post Saturday, 12th September 2015, 18:40

Re: Early Game Myth

For whatever it's worth, here's some stats. I looked only at recent games (cv >= 0.15). Details are in spoilers:

After I reach Lair, my chances of winning have increased 31% since starting the game (54% vs. 23%).
Spoiler: show
I have a 22.6% win rate over this time (not too shabby, I would say - I have 172 all-time wins):
  Code:
WalkerBoh   !lg walkerboh recent / won
Sequell   55/243 games for walkerboh (recent): N=55/243 (22.63%)

I make it to Lair 41% of the time:
  Code:
WalkerBoh   !lairratio walkerboh recent
Sequell   walkerboh (recent) has reached Lair in 100 of 243 attempts: 41%

And win 54% of my games after reaching Lair:
  Code:
WalkerBoh   !lm walkerboh recent br.enter=lair / ktyp=winning
Sequell   54/100 milestones for walkerboh (recent br.enter=lair): N=54/100 (54.00%)


For a few other pretty good players, there's a pretty consistent 25-30% increase in win chance just by virtue of reaching Lair.
Spoiler: show
Lasty (112 all-time wins):
  Code:
94/337 games for lasty (recent): N=94/337 (27.89%)
lasty (recent) has reached Lair in 159 of 337 attempts: 47%
93/160 milestones for lasty (recent br.enter=lair): N=93/160 (58.12%)

+30% chance of winning.

Berder (104 all-time wins)
  Code:
103/477 games for berder (recent): N=103/477 (21.59%)
berder (recent) has reached Lair in 203 of 477 attempts: 43%
102/203 milestones for berder (recent br.enter=lair): N=102/203 (50.25%)

+29% chance of winning.

MorganLeah (173 all-time wins)
  Code:
82/679 games for morganleah (recent): N=82/679 (12.08%)
morganleah (recent) has reached Lair in 206 of 679 attempts: 30%
79/208 milestones for morganleah (recent br.enter=lair): N=79/208 (37.98%)

+26% chance of winning.

Zooty (164 all-time wins)
  Code:
130/243 games for zooty (recent): N=130/243 (53.50%)
zooty (recent) has reached Lair in 154 of 243 attempts: 63%
128/155 milestones for zooty (recent br.enter=lair): N=128/155 (82.58%)

+29% chance of winning.


Our OP has a 14% increased chance of winning after reaching Lair.
Spoiler: show
  Code:
3/217 games for aethrus: N=3/217 (1.38%)
aethrus has reached Lair in 18 of 217 attempts: 8%
3/19 milestones for aethrus (recent br.enter=lair): N=3/19 (15.79%)


------------------------------------------------------------------------------------------------

Finally, here's a general breakdown of my win % throughout the game. I broke this down by year played to show my progress over time:
Spoiler: show
2012
  Code:
Win % at game start: 2.03% [16/787]
Enter Lair:          9.94% [16/161] +8%
Collect one rune:    27.6% [16/58] +19%
Collect three runes: 41.5% [16/41] +14%
Enter Zot5:          57.1% [16/28] +16%


2013
  Code:
Win % at game start: 6.78% [56/826]
Enter Lair:          27.7% [56/202] +20%
Collect one rune:    53.9% [56/104] +26%
Collect three runes: 73.7% [56/76] +20%
Enter Zot5:          87.5% [56/64] +14%


2014
  Code:
Win % at game start: 22.0% [67/305]
Enter Lair:          53.6% [67/125] +32%
Collect one rune:    70.7% [65/92] +17%
Collect three runes: 84.4% [65/77] +14%
Enter Zot5:          93.1% [67/72] +9%

(Not sure why Sequell listed 65 instead of 67 wins for the rune milestones...)

2015
  Code:
Win % at game start: 22.5% [33/147]
Enter Lair:          50.0% [33/66] +28%
Collect one rune:    78.6% [33/42] +29%
Collect three runes: 94.3% [33/35] +16%
Enter Zot5:          97.1% [33/34] +3%


Some final thoughts based on this and my own experience:

- Players generally have a significantly increased chance of winning once they reach Lair. This is regardless of skill level.
- This does NOT mean that the game after Lair is trivial (even quite good players can have <60% win rate after reaching Lair).
- If you break down the game into the 5 segments I chose above, my win % spiked the most when I was newer if I collected one rune.
- As a more experienced player, my win % spikes the most when I reach Lair.

For this message the author WalkerBoh has received thanks: 4
all before, Crauler, dpeg, Lasty

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Saturday, 12th September 2015, 19:13

Re: Early Game Myth

WalkerBoh wrote:For whatever it's worth, here's some stats. I looked only at recent games (cv >= 0.15). Details are in spoilers:

After I reach Lair, my chances of winning have increased 31% since starting the game (54% vs. 23%).
Spoiler: show
I have a 22.6% win rate over this time (not too shabby, I would say - I have 172 all-time wins):
  Code:
WalkerBoh   !lg walkerboh recent / won
Sequell   55/243 games for walkerboh (recent): N=55/243 (22.63%)

I make it to Lair 41% of the time:
  Code:
WalkerBoh   !lairratio walkerboh recent
Sequell   walkerboh (recent) has reached Lair in 100 of 243 attempts: 41%

And win 54% of my games after reaching Lair:
  Code:
WalkerBoh   !lm walkerboh recent br.enter=lair / ktyp=winning
Sequell   54/100 milestones for walkerboh (recent br.enter=lair): N=54/100 (54.00%)


For a few other pretty good players, there's a pretty consistent 25-30% increase in win chance just by virtue of reaching Lair.
Spoiler: show
Lasty (112 all-time wins):
  Code:
94/337 games for lasty (recent): N=94/337 (27.89%)
lasty (recent) has reached Lair in 159 of 337 attempts: 47%
93/160 milestones for lasty (recent br.enter=lair): N=93/160 (58.12%)

+30% chance of winning.

Berder (104 all-time wins)
  Code:
103/477 games for berder (recent): N=103/477 (21.59%)
berder (recent) has reached Lair in 203 of 477 attempts: 43%
102/203 milestones for berder (recent br.enter=lair): N=102/203 (50.25%)

+29% chance of winning.

MorganLeah (173 all-time wins)
  Code:
82/679 games for morganleah (recent): N=82/679 (12.08%)
morganleah (recent) has reached Lair in 206 of 679 attempts: 30%
79/208 milestones for morganleah (recent br.enter=lair): N=79/208 (37.98%)

+26% chance of winning.

Zooty (164 all-time wins)
  Code:
130/243 games for zooty (recent): N=130/243 (53.50%)
zooty (recent) has reached Lair in 154 of 243 attempts: 63%
128/155 milestones for zooty (recent br.enter=lair): N=128/155 (82.58%)

+29% chance of winning.


Our OP has a 14% increased chance of winning after reaching Lair.
Spoiler: show
  Code:
3/217 games for aethrus: N=3/217 (1.38%)
aethrus has reached Lair in 18 of 217 attempts: 8%
3/19 milestones for aethrus (recent br.enter=lair): N=3/19 (15.79%)


------------------------------------------------------------------------------------------------

Finally, here's a general breakdown of my win % throughout the game. I broke this down by year played to show my progress over time:
Spoiler: show
2012
  Code:
Win % at game start: 2.03% [16/787]
Enter Lair:          9.94% [16/161] +8%
Collect one rune:    27.6% [16/58] +19%
Collect three runes: 41.5% [16/41] +14%
Enter Zot5:          57.1% [16/28] +16%


2013
  Code:
Win % at game start: 6.78% [56/826]
Enter Lair:          27.7% [56/202] +20%
Collect one rune:    53.9% [56/104] +26%
Collect three runes: 73.7% [56/76] +20%
Enter Zot5:          87.5% [56/64] +14%


2014
  Code:
Win % at game start: 22.0% [67/305]
Enter Lair:          53.6% [67/125] +32%
Collect one rune:    70.7% [65/92] +17%
Collect three runes: 84.4% [65/77] +14%
Enter Zot5:          93.1% [67/72] +9%

(Not sure why Sequell listed 65 instead of 67 wins for the rune milestones...)

2015
  Code:
Win % at game start: 22.5% [33/147]
Enter Lair:          50.0% [33/66] +28%
Collect one rune:    78.6% [33/42] +29%
Collect three runes: 94.3% [33/35] +16%
Enter Zot5:          97.1% [33/34] +3%


Some final thoughts based on this and my own experience:

- Players generally have a significantly increased chance of winning once they reach Lair. This is regardless of skill level.
- This does NOT mean that the game after Lair is trivial (even quite good players can have <60% win rate after reaching Lair).
- If you break down the game into the 5 segments I chose above, my win % spiked the most when I was newer if I collected one rune.
- As a more experienced player, my win % spikes the most when I reach Lair.

Note that this isn't quite the right way of looking at these statistics, that you've gone from 23% to 54% doesn't mean the game is 31% easier from that point forward, only that you've *completed* 31% of the average game difficulty by that point, if you want to compare game difficulty you need to compare *overall success rates* to each other:

So 41% of your games make it as far as the lair, and 54% of your games go on to win from there, which means that it's 54% of the remaing not-dead 59% (31.86%) vs your overall 23% which means that for you the area after the lair is about 9% easier than the area before it.
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!

For this message the author Siegurt has received thanks: 3
Arrhythmia, dpeg, Lasty

Zot Zealot

Posts: 1031

Joined: Friday, 26th April 2013, 19:52

Location: AZ, USA

Post Saturday, 12th September 2015, 19:54

Re: Early Game Myth

So 41% of your games make it as far as the lair, and 54% of your games go on to win from there, which means that it's 54% of the remaing not-dead 59% (31.86%) vs your overall 23% which means that for you the area after the lair is about 9% easier than the area before it.

That doesn't make much sense to me. If anything you would say that 59% of my characters die before Lair, and 46% die after reaching Lair. Which would be a 13% difference.

Another way of thinking about it is this: If I were to start 100 games, you would expect 23 to win. Of the 77 who die, 59 of them would die before making it to Lair (41% lair rate). The other 18 would die somewhere after Lair. From that measure, if I am going to die, I would expect it to happen before Lair about 59/77 = 77% of the time. :)

I don't really care which way you want to twist around the numbers. You are probably right that looking at death rate is more useful than win rate. But I think the general point/trend is valid either way.

For this message the author WalkerBoh has received thanks:
Lasty

bel

Cocytus Succeeder

Posts: 2184

Joined: Tuesday, 3rd February 2015, 22:05

Post Saturday, 12th September 2015, 20:02

Re: Early Game Myth

Looking at overall success rate doesn't make sense here.

Let A = "you win" and B = "you get to lair".

To see why Siegurt's calculation doesn't make sense: suppose Prob (A) = 60% And Prob (A given B) = 100%. Then literally, "if you get to Lair you win" - thus B is the same as A, and Prob (B) is also 60%. But by Siegurt's calculation, you would be looking at the remaining "non-dead", so 100% of 60% (or 100% of 40% - it's unclear what exactly he is doing) - that would imply that difficulty of game after Lair is the same or worse as the game before Lair!

Math follows under spoiler:
Spoiler: show
You want to compare Prob (A given B), with Prob (A) or Prob (B) (unconditional). Which is what WalkerBoh's calculation did. Siegurt is looking at Prob (B) * Prob(A given B) which is same as Prob (A and B), which is not what the OP is talking about.


(Edit:mixed up Siegurt and Sar)
Last edited by bel on Saturday, 12th September 2015, 20:50, edited 3 times in total.

For this message the author bel has received thanks:
WalkerBoh

Dungeon Master

Posts: 3618

Joined: Thursday, 23rd December 2010, 12:43

Post Saturday, 12th September 2015, 20:18

Re: Early Game Myth

It is absolutely reasonable to expect a game of any length to show decreasing mortality with progress. This is why the early game can be as hard as it wants, but the late game cannot.

This fact is separate from the question whether the late game is too easy.

And still another (but this time related) issue is that certain tactics (luring monsters one by one, for example), help a lot with survival, but might be considered undesirable.

For this message the author dpeg has received thanks: 2
Rast, WalkerBoh

Zot Zealot

Posts: 1031

Joined: Friday, 26th April 2013, 19:52

Location: AZ, USA

Post Saturday, 12th September 2015, 20:21

Re: Early Game Myth

Thanks for the explanation bel, you explained very nicely what I was trying to do in the original post.

But per Siegurt's point, here is the redone stats for my games by year if you just look at death rate for each section of the game. The % is telling you how likely it is that a given character will die before reaching the next milestone, assuming they made it to the current one. So to take an example, my characters in 2012 had a 79.5% chance of dying before Lair. If they made it to Lair, they had a 64% chance of dying before reaching a rune. If they got rune, they had a 29.5% chance of dying before getting three runes, and so on.

Data:
Spoiler: show
2012
  Code:
% chance of dying before Lair: 79.5% [626/787]
Between Lair and 1 rune:       64.0% [103/161]
One and three runes:           29.3% [17/58]
Three runes and Zot5:          31.7% [13/41]
After reaching Zot5:           42.9% [12/28]

Overall death percentage: 98.0%



2013
  Code:
% chance of dying before Lair: 75.5% [624/826]
Between Lair and 1 rune:       48.5% [98/202]
One and three runes:           26.9% [28/104]
Three runes and Zot5:          15.7% [12/76]
After reaching Zot5:           12.5% [8/64]

Overall death percentage: 93.2%


2014
  Code:
% chance of dying before Lair: 59.0% [180/305]
Between Lair and 1 rune:       26.4% [33/125]
One and three runes:           16.3% [15/92]
Three runes and Zot5:          6.49% [5/77]
After reaching Zot5:           6.94% [5/72]

Overall death percentage: 78.0%


2015
  Code:
% chance of dying before Lair: 55.1% [81/147]
Between Lair and 1 rune:       36.4% [24/66]
One and three runes:           16.7% [7/42]
Three runes and Zot5:          2.86% [1/35]
After reaching Zot5:           2.94% [1/34]

Overall death percentage: 77.5%


PS: I really don't want to derail the thread with a statistics discussion, I just wanted to give a sense of some of the numbers involved. I'd be more than happy to start a new thread (or have this moved), if people are interested in looking into the statistics more.

bel

Cocytus Succeeder

Posts: 2184

Joined: Tuesday, 3rd February 2015, 22:05

Post Saturday, 12th September 2015, 20:43

Re: Early Game Myth

Aethrus wrote:And seriously, 'the early game is the hardest part of the game?' For races and backgrounds that lack useful tools at the start, sure. But pick a Minotaur, a Berserker, or Zot forbid, both, and substantiate the claim D1-D9 is the most challenging part for that build.

  Code:
!lg * mibe recent / won
<Sequell> 452/28560 games for * (mibe recent): N=452/28560 (1.58%)
!lairratio * mibe recent
<Sequell> * (mibe recent) has reached Lair in 5897 of 28560 attempts: 21%
!lm * mibe recent br.enter=lair / ktyp=winning
<Sequell> 471/6062 milestones for * (mibe recent br.enter=lair):
N=471/6062 (7.77%)

Looks like MiBe has higher chance to reach Lair, than to win after that. It is perhaps largely spurious though.

Possible reasons (except all the usual caveats about Sequell data)
a) Lots of newbies play MiBe, and not too many experienced players do (I couldn't find WalkerBoh, Berder, Lasty or elliptic there)
b) People don't play like "Hypothetical Optimal Player" - many are probably speedruns etc.

(I checked a few of my own accounts, the deaths after lair were due to mostly fooling around, quits or when I was really new)

Ziggurat Zagger

Posts: 6454

Joined: Tuesday, 30th October 2012, 19:06

Post Saturday, 12th September 2015, 23:36

Re: Early Game Myth

bel wrote:Looking at overall success rate doesn't make sense here.

Let A = "you win" and B = "you get to lair".

To see why Siegurt's calculation doesn't make sense: suppose Prob (A) = 60% And Prob (A given B) = 100%. Then literally, "if you get to Lair you win" - thus B is the same as A, and Prob (B) is also 60%. But by Siegurt's calculation, you would be looking at the remaining "non-dead", so 100% of 60% (or 100% of 40% - it's unclear what exactly he is doing) - that would imply that difficulty of game after Lair is the same or worse as the game before Lair!

Math follows under spoiler:
Spoiler: show
You want to compare Prob (A given B), with Prob (A) or Prob (B) (unconditional). Which is what WalkerBoh&#39;s calculation did. Siegurt is looking at Prob (B) * Prob(A given B) which is same as Prob (A and B), which is not what the OP is talking about.


(Edit:mixed up Siegurt and Sar)

I inverted win percentage and survival percentage, sorry about that.
Spoiler: show
This high quality signature has been hidden for your protection. To unlock it's secret, send 3 easy payments of $9.99 to me, by way of your nearest theta band or ley line. Complete your transmission by midnight tonight for a special free gift!
User avatar

Ziggurat Zagger

Posts: 4478

Joined: Wednesday, 23rd October 2013, 07:56

Post Sunday, 13th September 2015, 00:26

Re: Early Game Myth

I've played 578 games total, won 31, so 5.4%. From DCSS 0.13 on, I've played 122 games and won 18, so 14.75%. Some improvement has happened.

Of the 104 games I've lost since 0.13, 20 were lost after reaching Lair and 84 before Lair. So I've won 47% of the games where I've reached Lair.
27 games were lost on D:1 or D:2, so I've won 19% of the games where I've reached D:3.

What does this prove? That half the time after reaching Lair, I make a stupid mistake and die...
DCSS: 97:...MfCj}SpNeBaEEGrFE{HaAKTrCK}DsFESpHu{FoArNaBe}
FeEE{HOIEMiAE}GrGlHuWrGnWrNaAKBaFi{MiDeMfDe}{DrAKTrAMGhEnGnWz}
{PaBeDjFi}OgAKPaCAGnCjOgCKMfAEAtCKSpCjDEEE{HOSu
Bloat: 17: RaRoPrPh{GuStGnCa}{ArEtZoNb}KiPaAnDrBXDBQOApDaMeAGBiOCNKAsFnFlUs{RoBoNeWi

For this message the author Sprucery has received thanks:
Sar

Spider Stomper

Posts: 234

Joined: Thursday, 30th April 2015, 08:29

Post Sunday, 13th September 2015, 03:05

Re: Early Game Myth

If you fight with multiple gnoll in D1, you always die.

But of late game characters can survive multiple OOF/hell sentinel/greater mummy.

For this message the author onget has received thanks:
radinms
User avatar

Vaults Vanquisher

Posts: 422

Joined: Wednesday, 13th August 2014, 08:34

Post Sunday, 13th September 2015, 03:33

Re: Early Game Myth

Bot says "early game is dangerous. smart players avoid it".
If you do early game, you are not smart player.

For this message the author radinms has received thanks:
onget

Shoals Surfer

Posts: 297

Joined: Wednesday, 9th July 2014, 08:20

Post Sunday, 13th September 2015, 04:52

Re: Early Game Myth

FR: Game mode where you start in the middle of the orb chamber and must travel to D:1, grab the orb and return it back there.

Since we've reversed the order of early, middle and late game, clearly this must be easier, at least until you clear 99% of the game and run into the dreaded hobgoblin with stones.

Slime Squisher

Posts: 387

Joined: Monday, 15th August 2011, 16:31

Location: Frankfurt

Post Sunday, 13th September 2015, 09:25

Re: Early Game Myth

I do think "you've won the game once you reached Lair" is hyperbole, otherwise nobody would regard "Death's Door" being worth the spell slots. However, I don't remember one single death after that point that wasn't my own fault, even if that fault was just "lack of knowledge" or "bad build decisions".

To be fair, the vast, vast, vast majority of my deaths in early D are my own fault, too. Unavoidable deaths do occur, but in my experience they are very, very rare. I even blame at least half of my D1 deaths on myself (though I'm not going to adopt the sort of gameplay that would help me avoid them.)

Early D is the most interesting and challenging part of the game, because it just so happens that it is the most frequently played and thus best tested and thus most balanced part of the game.

For some reasons, I'm personally finding the first two levels of Lair branches the most difficult part. At least currently.
"... while we / Unburden'd crawl toward death." -- King Lear I,1
User avatar

Slime Squisher

Posts: 338

Joined: Wednesday, 20th November 2013, 11:37

Post Monday, 14th September 2015, 07:18

Re: Early Game Myth

mopl wrote:
Sequell wrote:2400 games for mopl (!won): 594x D:1, 424x D:2, 341x D:3, 235x D:4, 134x D:5, 89x D:6, 59x D:7, 46x D:8, 41x D:10, 37x Abyss:1, 36x D:9, 26x Lair:1, 23x Orc:4, 18x D:$, 17x Ossuary, 17x IceCv, 14x D:11, 13x Lair:3, 12x Orc:3, 12x Sewer, 11x Elf:3, 11x Lair:2, 11x Bailey, 10x D:12, 10x Lab, 9x Abyss:3, 8x Abyss:2, 7x Lair:7, 6x Vaults:5, 6x Snake:5, 6x Lair:5, 5x Lair:6, 5x Orc:2, 5x Lair:8, 5x Orc:1, 5x D:13, 5x Lair:4, 5x Vaults:3, 4x Shoals:5, 4x Swamp:2, 4x Swamp:5, 4x D:15, 4x Volcano, 3x Depths:1, 3x Elf:1, 3x Slime:6, 3x Vaults:4, 3x Snake:2, 3x Dis:7, 3x Swamp:4, 2x Vaults:2, 2x Crypt:1, 2x Depths:3, 2x Spider:5, 2x D:14, 2x Temple, 2x Snake:3, 2x Zig:12, 2x Shoals:3, Tomb:1, Tar:2, Zig:19, Abyss:4, D:17, Snake:1, Geh:7, Vaults:1, Depths:2, Zot:4, Elf:2, Zig:24, Spider:1, WizLab, Slime:3, Tomb:3, Zig:5, Spider:2, Zot:5, Swamp:3, Slime:5, Abyss, Swamp:1
1922 out of 2400 death...

Lol! Really!
I checked sequell today, your D1-D9 deaths are 1958 of 2400 games. That means you have a (1 - 1958/2400) * 100% = 18,4% chance to survive during D1-D9 run.
You have 78 succesful ascensions, that mean tha from that succesful D1-D9 runs in 78 / (2400 - 1958) * 100% = 17,6% runs you survived after D:9.
Statistics says that for you D1-D9 is less dangerous than D:9-escape with an orb.

Slime Squisher

Posts: 377

Joined: Friday, 1st February 2013, 21:08

Post Monday, 14th September 2015, 15:00

Re: Early Game Myth

Early game rarely provides opportunities to make strategically important decisions - esp. before the temple. In my case, there's a lot of automatization going on. We play early game so often, that eventually tactics apply themselves without involving a brain. If anything, patience and calmness are tested to the fullest, as one careless keypress (aka tab) is more deadly than later on.

There is also another mechanism which applies to 95% of my non-streak games and I think it applies to many of you, too. I become cautious about my character only after I get attached - sometimes around temple, sometimes in lair, but never on d:1. It's pretty safe to say that the more we invest, the more we care.

As long as deadly=hard, your numbers do not lie... if that's the definition.

For this message the author Bart has received thanks:
onget

Return to Crazy Yiuf's Corner

Who is online

Users browsing this forum: No registered users and 22 guests

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