Chaosforge Forum

  • October 07, 2024, 23:42
  • Welcome, Guest
Please login or register.



Login with username, password and session length
Pages: [1]

Author Topic: Theory: AoMC vs luck  (Read 2733 times)

Nameless

  • First Sergeant
  • *
  • Offline Offline
  • Posts: 115
  • Proud owner of 22/22 bronzes: "It's a start!"
    • View Profile
Theory: AoMC vs luck
« on: March 28, 2011, 00:44 »

People have often said that the Angel of Max Carnage requires more skill and game mastery than other challenge modes. The in-game description itself opens with "You hate chance,". However, in my experience an AoMC run involves a greater element of luck than standard play.

Why? Because AoMC eliminates only a single - and perhaps the least important - element of chance from DoomRL. In a standard run, thousands of shots can be fired. These will range from misses to max damage, but due to the very large number of rolls it is incredibly unlikely that on any given run you will give or take notably more or less damage than usual. In small situations a single roll can be important (eg. under fire from a mancubus), but these situations are uncommon enough that a player can reasonably predict and either avoid or recover from them.

By comparison, the mere existence of an invulnerability globe on a given level will make a much larger difference to the weapon damage given or taken. As these are rolled for only when a level is generated, there is also a much higher chance of abnormally many or few globes during a given run. In the extreme case, whether or not Hell's Armory appears can make a significant difference to a player's firepower - and this roll is made exactly once per run.

Fixing weapon damage in itself might shift the luck/skill balance slightly towards skill. However, by fixing the damage at maximum several other luck elements become much more important than normal. Let's look at the mancubus example again. In a standard game, the player can run behind a corner and kill it without risking much more than a large med pack. In AoMC, the same player with the same tactics risks instant death; the random chance being whether the mancubus moves or fires on the turn the player spots it. That one random roll now becomes much more important than the entire succession of rolls in a standard game!

Playing with high skill will reduce this. So would playing with high skill in a standard game, ergo the same player will be most influence by luck on AoMC regardless. The reasoning is fairly simple: Setting the damage high inevitably raises the consequences of other rolls, and sufficiently bad rolls can not average out over the game because, of course, the game ends when the consequences hit 'death'.

Conclusion: Sudden death is awesome, but there's a reason pro gaming tournaments don't start with it. :)
Logged

MaiZure

  • Marketing/PR Ops Lead
  • Greater Elder
  • Major General
  • *
  • *
  • Offline Offline
  • Posts: 827
  • Retired 1-Star General (.993)
    • View Profile
    • Project repository
Re: Theory: AoMC vs luck
« Reply #1 on: March 28, 2011, 02:06 »

Without getting in to rigorous analysis, I can attest that my DoomRL growth as a player has always suggested AoMC to be easier than a standard game.

(Ignoring AoPc)

My first HMP victory was AoMC followed by a standard game
My first UV Win and Full Win was AoMC before winning a standard game
and in N!, I've made it in to level 21 on AoMC but only level 14 in standard (not counting my N! AoPc win)

Pre-9.9.2, many have said that Destroyer Diamond was the easiest of the lot.
Fanta Hege: http://forum.chaosforge.org/index.php/topic,3026.msg28349.html#msg28349
Game Hunter: http://forum.chaosforge.org/index.php/topic,3354.msg29324.html#msg29324


My personal play style has always favored tactical defense, which is the only way to play AoMC.

EDIT: The factor that matters most in my AoMC is the 98% accuracy with all weapons - being able to radar shoot with the chaingun starting at dlevel 2 or 3 seems to hit enemies out of range 50% of the time, which is great for clearing out those long, thin hallways.
« Last Edit: March 28, 2011, 02:39 by MaiZure »
Logged
Hell Knight 1st Lt. (.997)
[22/13/9/2/0]

Tuomari Jim

  • Sergeant
  • *
  • Offline Offline
  • Posts: 75
  • Just lost.
    • View Profile
Re: Theory: AoMC vs luck
« Reply #2 on: March 28, 2011, 02:31 »

People have often said that the Angel of Max Carnage requires more skill and game mastery than other challenge modes.

Who, exactly?

The in-game description itself opens with "You hate chance,". However, in my experience an AoMC run involves a greater element of luck than standard play.

I'll grant you that. Arachnotron caves are all the deadlier.

Why? Because AoMC eliminates only a single - and perhaps the least important - element of chance from DoomRL.

Yeah, it's not like being able to one-shot mancubi and archviles is a big deal or anything.

In a standard run, thousands of shots can be fired. These will range from misses to max damage, but due to the very large number of rolls it is incredibly unlikely that on any given run you will give or take notably more or less damage than usual. In small situations a single roll can be important (eg. under fire from a mancubus), but these situations are uncommon enough that a player can reasonably predict and either avoid or recover from them.

True enough.

By comparison, the mere existence of an invulnerability globe on a given level will make a much larger difference to the weapon damage given or taken. As these are rolled for only when a level is generated, there is also a much higher chance of abnormally many or few globes during a given run. In the extreme case, whether or not Hell's Armory appears can make a significant difference to a player's firepower - and this roll is made exactly once per run.

Most of the times I encounter an invincibility globe, it happens to be on the wrong side of the map or it's the last thing I discover on the map.

Fixing weapon damage in itself might shift the luck/skill balance slightly towards skill. However, by fixing the damage at maximum several other luck elements become much more important than normal. Let's look at the mancubus example again. In a standard game, the player can run behind a corner and kill it without risking much more than a large med pack. In AoMC, the same player with the same tactics risks instant death; the random chance being whether the mancubus moves or fires on the turn the player spots it. That one random roll now becomes much more important than the entire succession of rolls in a standard game!

True enough, but on AoMC you need to change tactics from running to shooting, unless you're already next to a corner. With a sufficiently modded plasma gun(high power will do) you can invariably one-shot incoming mancubi. You can hear that there's a mancubus behind the corner but dare not peek? No problem, shoot through the frigging wall. Peek-a-boo you fat bastard mancubus you! On AoMC, going through walls is a national sport.

Playing with high skill will reduce this. So would playing with high skill in a standard game, ergo the same player will be most influence by luck on AoMC regardless. The reasoning is fairly simple: Setting the damage high inevitably raises the consequences of other rolls, and sufficiently bad rolls can not average out over the game because, of course, the game ends when the consequences hit 'death'.

Hmm.

Without getting in to rigorous analysis, I can attest that my DoomRL growth as a player has always suggested AoMC to be easier than a standard game.
--
My first UV Win and Full Win was AoMC before winning a standard game
--
My personal play style has always favored tactical defense, which is the only way to play AoMC.

I will second these. AoMC is all about finding new ways to scout with a shotgun, not to mention corner shooting. In fact, going back to standard game is a bit of a shock: "What do you mean my chaingun is useless? Since when were demons this persistent? How come I'm always running out of ammo? I SWEAR MY ROCKETS USED TO HIT" and all that jazz.
« Last Edit: March 28, 2011, 02:37 by Tuomari Jim »
Logged

MaiZure

  • Marketing/PR Ops Lead
  • Greater Elder
  • Major General
  • *
  • *
  • Offline Offline
  • Posts: 827
  • Retired 1-Star General (.993)
    • View Profile
    • Project repository
Re: Theory: AoMC vs luck
« Reply #3 on: March 28, 2011, 02:35 »

In fact, going back to standard game is a bit of a shock: "What do you mean my chaingun is useless? Since when were demons this persistent? How come I'm always running out of ammo? I SWEAR MY ROCKETS USED TO HIT" and all that jazz.

I would also add: "What!? These Barons and HKs can actually enter my LOS when I'm radar shooting with the combat shotty?" In AoMC it's pretty much auto-knockback even outside visual range. Power modded, we're talking double knockback!
Logged
Hell Knight 1st Lt. (.997)
[22/13/9/2/0]

Game Hunter

  • Programmer
  • Elder Chaos Guard
  • Lieutenant General
  • *
  • *
  • Offline Offline
  • Posts: 1044
  • Looks like game to me.
    • View Profile
    • Channel, the Roguelike
Re: Theory: AoMC vs luck
« Reply #4 on: March 28, 2011, 12:26 »

We could talk forever about which lucky variables are more important in a game:

  • weapon availability (as simple as getting that combat shotty on dlvl2, or the greater complexities of rocket launchers and plasma rifles)
  • floor generation (hellish caves and arenas are often the worst offenders here, although maze of rooms in AoRA is a big pain in the ass) and special level generation
  • spawning point (did you start in a relative center with enemies on all sides? Congratdulations, you're probably screwed)
  • Powerup generation and placement (Berserk Pack? Awesome. Wrong side of the room? Far less awesome.)
  • Enemy generation and placement (see above on spawning point, although the same is true when a big stack of enemies is also right by the stairs or impeding your movement)
  • Ammo generation, or ammo carrier generation (too many VMR or B/HK packs? Sucks to be you!)
And the list goes on and on. And it's true, only two probabilistic functions has been collapsed in AoMC: damage and accuracy. But the fact that SOMETHING has been made to basically be a constant in an ever-changing game is already a huge advantage. We can finally start abusing the limitations of AI, for instance, or readily calculating out the exact number of shots necessary to kill a given enemy. It allows you a better preparation against other bad-luck situations. The nice thing about AoMC is that the difficulty modifier, too, is less important. HMP/UV/N! all get accuracy boosts, but these don't matter when everything has a +12 to-hit. Health concerns, are more similar to AoHu, since you'll be taking a crapload of damage whenever an enemy gets the chance to attack you: this means you're forced to commit to a more tactical offense (which, because of maxed damage, is very doable with a shotgun).  I think I speak of it best in my guide when I say that it's the kind of challenge that is hard to learn but easy to master.

Now I can ramble on some more, but I think you'll find that, statistically, more people complete Max Carnage runs than they do standard runs, specifically on UV/N!. Although I believe part of this is because more damage means you'll probably be finishing levels faster on average, which means it's less of a slog in general.
Logged
I'm just a dude playing a dude disguised as another dude.

Latest LPs: Angband, Delver
Pages: [1]