Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
Welcome to rpgcodex.net, a site dedicated to discussing computer based role-playing games in a free and open fashion. We're less strict than other forums, but please refer to the rules.
"This message is awaiting moderator approval": All new users must pass through our moderation queue before they will be able to post normally. Until your account has "passed" your posts will only be visible to yourself (and moderators) until they are approved. Give us a week to get around to approving / deleting / ignoring your mundane opinion on crap before hassling us about it. Once you have passed the moderation period (think of it as a test), you will be able to post normally, just like all the other retards.
There are always exceptions, of course, but the first tabletop games I played all used dice. I'm not talking about only D&D. The various alternative games TSR and other publishers were putting out during the early PnP craze, used dice as well. I can only comment based on experience I've had. And to me, a RPG includes dice. The 20-sider is iconic.
The fact that you think a shift away from a PnP framework is good, and statements such as less random is better, means we are approaching the subject from very different places. For the record, I haven't read any reason why less randomization/unpredictability/luck is better. The premise of removing even some of it from a classic RPG strikes me as bizarre. The counter arguement could be interpreted as change merely for the sake of change, implementing new systems just to be different, which is equally invalid.
Not changing for the sake of changing. That'd be the worst possible reason. I also do not advocate to completely remove randomness from cRPGs, it's very much needed for some parts. I just would want some random factors, connected to fighting mechanics mostly, to be governed by an "intelligence" that the designer would have put in-game, to give them a "purpose".
To give you a simple example, let's consider a room inside a dungeon :
1/ game without AI : the RNG decides if 0, 1 or 2 orcs spawn, and what spot they spawn at. 100% Random.
2/ game with AI : the Orcs do exist, they live, breathe and act somewhere in the dungeon. The AI decides if the Orcs happened to hear the incoming player, if one of the Orcs was sleeping instead, if the Orcs are clever/motivated enough to work together against the player threat, from which side(s) they will enter the room, if one of the Orcs goes to call for reinforcements while the other Orc fights, etc.
To the player, the result might seem exactly the same from the outside (fighting some Orcs, in a context built upon some "random" factors), but in the second case the "randomness" adds to the game. It has a hidden purpose, that makes sense, that the player could even discover and counter, if he was smart enough to gather some intel beforehand, or talented enough to sneak in undetected, etc.
If the player gets killed, because he sucks at fighting Orcs, and reloads :
in game 1/, he might just win out of sheer luck, even though context remains the SAME.
in game 2/, there is no way he should be able to win if he didn't care to further improve his skills, or revise his tactics, because the AI will recreate the same conditions that lead to his demise.
Maybe is it because I like Strategy/War games as much as I enjoy RPGs that I look to remove the RNG in cRPGs as much as possible in situations that do not warrant it.
To replace it with something that would rely more on the the tactical skills of the player, or his ability to learn, and take advantage of the game mechanics (spells, feats, geography, weapon and armor characteristics, etc.) rather than on sheer luck.
IMO, dices work for table top games, they do not work that well for video-games, some parts of them anyway.
Personally, I don't see the point of dice rolls in the situation of a fight in a cRPG.
If the ATK (or whatever stat) of your char, modulated by some other factors (geography, lighting, position, morale, stamina, etc.), is high enough to overcome the DEF/EVD (or whatever stat) of his enemy, he should connect everytime.
I am not petioning for glancing blows here either. I just think that the outcome of a combat can be changed by so many other different factors (avalaibility and use of feats, skills, spells, support tactics, formations, flanking/surrounding manoeuvres, moving onto higher ground, etc.), depending on player skill rather than on the RNG, that there simply isn't the need anymore for dice rolls to keep battles interesting. More interesting than watching a collection of "hit, miss, hit, miss, hit, hit, miss..." messages streaming on the bottom of my screen anyway.
Precisely. That's also what contributed to kill dice rolls, or the validity of the paradigm that uses them so extensively in cRPGs. Plus the fact that one can potentially have hundreds of consecutive fights in cRPGs within an hour, when one has one or two (if they're correctly DM'ed out) in a table top game.
Also, if all the game mechanics are built around lowering the variance of dice rolls to accomodate the player (ignoring critical misses on a natural "1", negating critical hits by enemies), artificially inflating his chances at connecting based on some behind-the-scene workings around levels (at level N-1, impossible to hit an enemy, at level N, "automagically" it becomes possible almost every time), one has to wonder for how much longer these keep making any sense.
Would it be a stretch to suggest that some people who advocate less randomization in cRPGs, also are less inclined to support turn-based combat? They might even see the shift toward real time as a good thing.
Would it be a stretch to suggest that some people who advocate less randomization in cRPGs, also are less inclined to support turn-based combat? They might even see the shift toward real time as a good thing.
The vast majority of RPGs, including real-time "console popamole RPGs", use dice rolls in one form or another. "Derandomization" isn't a big thing, at all. So there's not enough data to support the existence of this correlation you're suggesting.
That's more or less where I'm standing too. I vastly prefer TB over RT whenever possible, except when it leads to games that make a poor use of it, forcing the player to mindlessly repeat actions in between 10-seconds long animations. Worst offenders are games that feature an "AUTO" button = obvious and belated excuse of the game designer for not being able to come up with an interesting TB combat system.
I fail to see the connection between RT/TB and randomization in cRPGs however.
The vast majority of RPGs, including real-time "console popamole RPGs", use dice rolls in one form or another. "Derandomization" isn't a big thing, at all. So there's not enough data to support the existence of this correlation you're suggesting.
Isn't a bit thing at all ? What is the point of dice rolls/"randomization" in the case of your average video-game/cRPG ?
Randomization makes sense for :
loot tables
encounter tables
geographies generated on the fly (landscapes, dungeons, etc.)
behavioral patterns of some NPCs (consider it as an "unpredictable" reaction to an outside/player stimuli, hence "interesting")
etc.
Randomization is NOT needed, and not even desirable, in the context of checks opposing (sometimes at high frequency) a skill against a challenge factor :
"picking locks", "persuasion", etc.
"attacks", "evade", "resists", etc.
You might want to keep randomization to calculate damage, especially criticals, to keep it somewhat "dynamic", but even then I'd rather much prefer to see a system that computes it based on :
the weapon/spell itself (duh)
the location that the hit/spell connects to
what kind of armor is protecting the location (electricity based spells may do more damage to enemies wearing metallic armor)
the strength and current stamina/fatigue or morale of the opponents
other situational factors, like hitting from the back of a horse, or geography (using 2handed weapons in narrow tunnels, or thick forests, isn't quite optimal to deliver deadly blows, for example)
I could elaborate, but I'd like to see what people find so compelling in a high degree of random in their fights and skill checks first.
I don't really want to get into this discussion, but I'd like to clarify something I said earlier in this thread and I think you may have misunderstood.
It is not the position of Rampant Coyote's post in OP that this type of randomization is an "uninteresting" type of randomization. He absolutely supports this randomization, as long as its consequences are not too extreme.
Randomization is good. Wildly disparate and irreversible consequences based on that randomization are bad.
Thanks for the clarification, and that's also how I understood it, but I see what you're hinting at.
I guess I should stop posting in this thread, since you seem to think (and probably rightfully so) that any replies should pertain to the OP.
I was too lazy to start a proper thread It was just too tempting to discuss antiquated dice (D&D) systems for skill checks/fights in cRPGs here and now.
Thanks for the clarification, and that's also how I understood it, but I see what you're hinting at.
I guess I should stop posting in this thread, since you seem to think (and probably rightfully so) that any replies should pertain to the OP.
I was too lazy to start a proper thread It was just too tempting to discuss antiquated dice (D&D) systems for skill checks/fights in cRPGs here and now.
Q: Can a weak goblin (NPC) hit a trained knight (PC) ?
A:
Case of a random (dice-based) battle system :
it's very unlikely to happen
if it happens, the player will feel frustrated/resentful, because there is probably nothing he could do about it
if it does not happen, the player will feel eventually bored with meaningless opponents
Case of a non-random battle system (with a clever AI) :
it has a good chance to happen
if it happens, it's because of poor judgement/tactical skills of the player confronted to the AI (which used reinforcements, surprise attacks, formations, any other dirty tricks it was designed to use, etc.)
-> as a result, the player feels encouraged to improve at the game, and to keep paying attention all through it
Q: Can a weak knight (PC) hit a powerful dragon (NPC) ?
A:
Case of a random (dice-based) battle system :
it's very unlikely to happen
if it does not happen, the player may feel frustrated/resentful
if it does happen, it's probably because the player cheated, not a very satisfying long-lasting experience to most
Case of a non-random battle system (with a clever AI) :
it has a good chance to happen
if it does not happen, the player knows it's just not meant to happen (the game designer is a dragon freak)
if it does happen, it's because the player used game mechanics to their fullest (dragon slayer sword, special feats, advantageous geographical features, etc.)
-> as a result, the player feels encouraged to improve at the game, always looking to use ALL the game mechanics to his advantage
Q: Are games with dice-based (random) battle systems more fun/rewarding ?
A:
Case of a random battle system :
the game typically produces a great deal of dice rolls in a very short time
-> the player will ignore them after a while
some hidden/obscure mechanics will tamper with the randomness of the rolls
-> a random system with tampered randomness isn't so random anymore, so why bother in the first place ?
-> since these mechanics will remain hidden most of the time, the player can remain confused
players that save-scum, or grind against countless of weak opponents, will eventually get the best results
-> neither a fun, nor a rewarding battle system (to me)
Case of a non-random battle system (with clever AI) :
if the system is well designed, the player will be forced to pay attention to what is happening on the battle field
random events, whether painful or favorable, whether in battle or outside of it, will remain rare, hence more unexpected and more "rewarding" to experience
it's meaningful in-game mechanics, their proper use, at a proper time, that makes fights feel "dynamic", "heroic", "epic", whatever, not some dice (EDIT : contrary to P&P where dice rolls happen less often, are much more dramatized, and thus can grant that feeling of achieving an "epic" task thanks to a successful roll)
players that learn how to take advantage of all the mechanics/features of the game will get the best results
-> it's actually a rewarding battle system, and fun as well (to me)
EDIT : The same observations apply more or less to skill checks too.
Are you advocating a system where it's immediately known whether you'll be successful in a particular action during combat (ex. generic-combat-skill 6 always hits against-generic-combat-skill 4), Karma? How do you have "chance" without randomization? Why does the use of randomization preclude the use of "clever" AI?
There's a chance my understanding of your posts is flawed, but that's what forum access at 1 AM is about.
Well, I can't answer that, but I sure can rant - and perhaps that'll do
For starters, you'll notice that D&D (older versions) pretty specifically and intentionally makes it so that: 1) even with ungodly armor class (-10 or -32768, who cares?) you'll still get hit if the attacker rolls a natural 20 2) even with ungodly saves, you'll still fail the save on a natural 1. The consequence being: you could be playing the most tweaked-out munchkin dragon slayer, the lowliest of the low kobolds with a poisoned dagger still has a 1:400 chance of insta-killing you. That's without even going into countless other insta-kill monsters and threats such as catoplebas, rot grubs, ear seekers, yellow mold, medusas, ...
Sure, you'd need a roguelike to properly implement that in a cRPG. Also, you'd need a much wider spectrum of supported player actions, because you don't get around a catoplebas by reloading until you pass your save, you get around it by blindfolding yourself and fighting it (it's pretty wimpy once you can't trigger the instant death); you don't prevent an ear seeker form nesting in your brain by enchanting a ring to give you +5 on Ear Membrane Thickness rolls, you use a listening horn with a mesh, etc. So, the ruleset pretty much works only when player ingenuity is expected and required (as opposed to just min-maxing your char and just letting the dice do the work for you). Since the dice can always screw you up, you're also encouraged to come up with ways that will eliminate dice rolling from situations that could have disastrous consequences. And you're encouraged to avoid combat and get XP from acquired treasure, since every combat could have a very unpleasant outcome (and randomly rolling an encounter with a hundred-ish orcs when you're level 1 is an important part of that general "style"). And of course, if your DM is a dick you're screwed (like any ruleset can help you there). In (off-topic) summary: bring back instakill poison in 5th edition ffs.
Back to the topic, if you blindly use the same rules in a non-roguelike cRPG, you'll just end up with reloading tedium. But hey, remember that Wizardry didn't save dungeon progress but only character progress, and had random encounters involving random combat rolls etc.? So obviously, some significant adjustments are needed. However, randomness still plays an important role.
If the lowliest opponent doesn't have even a miniscule chance of hurting your demi-god power gratification avatar, the game loses something important. I'd say that randomness is the closest we've found to injecting soul into computer code; but who's to say randomness is not what accounts for our souls as well? And if there is a correct sequence of moves that always gets you the same outcome in the same situation, the game is worse than "deterministic", it is "memorizeable". Combat should always carry risk, even when executed by the numbers (in a manner that would make military tacticians shed manly tears of joy). And that is exactly what separates good cRPG combat from chess and other strawmen. That it results in a reload when the dice don't go in the player's favor, when the same tactics result in a success when they do is obviously... suboptimal. There are different directions when handling this, but I'd say that letting the player reduce the influence of randomness on his outcome by smart play and planning is superior to letting the game reduce the influence of randomness across the board because the players can't be arsed to do it (learn to play?).
Skill checks? I don't know, I haven't made any games or played any that would randomize skill thresholds like we mentioned before - but I'm still sure I would like to at least try a game like that before saying it's definitely pointless. Also, consider human nature: if a game is good, players will want to learn it. We want the learning to be about the ruleset, instead of about the "map" (just like we want strategy games to be about, well, strategy and tactics instead of memorizing that Map B always has Y units of Resouce Z at location (24, 13)). And a game that is perfectly "memorizeable" will inevitably invite contempt. Sure, now we're talking personal character and willpower, but consider this - if Drog didn't have every goddamn skill/attribute check in Arcanum memorized, how less likely would he be of learning to hate it so much? (not 100% less likely, because there are still the bugs and the blatantly unfinished parts).
Also, I don't like messing with the player characters (with random adjustments like J1M mentioned), because fundamentally, PCs are the player's input and responsibility (and the more control players have over their characters, the better), while the game world is solely the game maker's responsibility. And dunno, but having the game world be more "worldly" (=unpredictable and not-trivially-deterministic, also not trivially memorizeable) sounds like a clear win to me. Just like we use turns and phases to make the player conform to the ruleset (instead of bypassing it at will with rudimentary timing and twitchy clicking skills), it sounds like a good idea to make the player conform to the ruleset instead of bypassing it at will with rudimentary memorization and googling skills. I'm not talking about preventing the player completely from any kind of badwrongfun, I'm talking about elementary focus shifting.
Then, riddle me this: what use is a ruleset in a static experience? Why do you need rules for mechanics, if the "mechanics" in question are merely "if your Skill X is > Y, and you choose Option A that involves it, you win"? If you're bothering to design actual mechanics, added fuzziness within their design parameters is a good thing. You could achieve that fuzziness with statistic modeling and whatnot, but it makes sense to use the simplest mechanic that achieves the desired effect (and in the case of percentile skill rolls, I think a (1d20-10)% adjustment would work as well). Without the "fuzziness" and unpredictability, you might as well not have any mechanics and reduce the whole game to a CYOA "make one of 1d6+1 predefined choices". And where's the "game" or "world" part in that?
And before you get on your high "statistic modeling" and "AI" horse, consider this: what are random tables based on, if not quasy-statistic modeling and creative nonsense (remember Gygax's previous career)? What does good AI operate on, if not tables of data and (quasy)statistic outcome prediction or heuristics? And if you have AI that can predict the outcome of actions much better than the player, should it always do so? Can the predictability of AI choices be used against it by players? I'd say AI that plays smart but with a decent degree of unpredictability would be much more interesting to play against than the alternative. But before a decent AI developer starts making cRPGs (that'll be the day), random tables and a bit of rolling can do the job just fine. Like Gary Archimedes Gygax famously didn't say: give me a large enough set of random tables and I'll simulate the Earth.
If the lowliest opponent doesn't have even a miniscule chance of hurting your demi-god power gratification avatar, the game loses something important. I'd say that randomness is the closest we've found to injecting soul into computer code; but who's to say randomness is not what accounts for our souls as well?
Sure, you'd need a roguelike to properly implement that in a cRPG...
Back to the topic, if you blindly use the same rules in a non-roguelike cRPG, you'll just end up with reloading tedium. But hey, remember that Wizardry didn't save dungeon progress but only character progress, and had random encounters involving random combat rolls etc.? So obviously, some significant adjustments are needed. However, randomness still plays an important role
I don't think it necessarily has to result in reloading tedium, so long as the instant death situations are recognizable as such before they kill you or if unpredictable, like a kobold with a poisoned dagger or the proverbial Gnome with the wand of death-- that its rare enough to not be an everyday worry.
To use NetHack as an example, once you've played enough to recognize that cockatrices will turn you to stone, you know that ranged combat is the way to go. Once you have enough of that knowledge across all monsters, the major everyday dangers of the dungeon become the potential for the RNG to overwhelm you with monsters above your pay grade. Which is easily addressed by tweaking the encounter tables.
I'd say that randomness is the closest we've found to injecting soul into computer code; but who's to say randomness is not what accounts for our souls as well?
For starters, you'll notice that D&D (older versions) pretty specifically and intentionally makes it so that: 1) even with ungodly armor class (-10 or -32768, who cares?) you'll still get hit if the attacker rolls a natural 20 2) even with ungodly saves, you'll still fail the save on a natural 1. The consequence being: you could be playing the most tweaked-out munchkin dragon slayer, the lowliest of the low kobolds with a poisoned dagger still has a 1:400 chance of insta-killing you. That's without even going into countless other insta-kill monsters and threats such as catoplebas, rot grubs, ear seekers, yellow mold, medusas, ...
"Natural 1" critical hits/misses have been eliminated from most cRPGs for a long time. I can't remember a cRPG in which I critically missed for the last time, maybe Fallout. Even then, it happened so rarely that I barely noticed it. Most people dieing to a critical hit will instantly reload anyway, so the significance of such a feature is subjective at best.
I don't believe in unavoidable (=depending on the RNG) insta-kill abilities. On the design sheet, it sounds like a great concept. Once in game, most players will hate it. Quite like dice, it's a concept that works great with the supervision of a human DM. In video-games, it doesn't. Make these abilities 100% insta-kill if the player didn't bother to counter or prepare himself, 0% otherwise.
Sure, you'd need a roguelike to properly implement that in a cRPG.
Also, you'd need a much wider spectrum of supported player actions, because you don't get around a catoplebas by reloading until you pass your save, you get around it by blindfolding yourself and fighting it (it's pretty wimpy once you can't trigger the instant death); you don't prevent an ear seeker form nesting in your brain by enchanting a ring to give you +5 on Ear Membrane Thickness rolls, you use a listening horn with a mesh, etc. So, the ruleset pretty much works only when player ingenuity is expected and required (as opposed to just min-maxing your char and just letting the dice do the work for you).
Yes (to more player actions, opposed to letting the dice do the work for you). A thousand times, yes.
Since the dice can always screw you up, you're also encouraged to come up with ways that will eliminate dice rolling from situations that could have disastrous consequences. And you're encouraged to avoid combat and get XP from acquired treasure, since every combat could have a very unpleasant outcome
Well, here we go (again) : games should have battle mechanics relying on dice/randomness, yet they should have as design focus to encourage players on how to bypass or cheat on them. Why bother with dice in that case ? There is clearly an issue here (or the game designer is schizophrenic).
Back to the topic, if you blindly use the same rules in a non-roguelike cRPG, you'll just end up with reloading tedium.
Agreed. A good game design should make save-scumming unnecessary, or at least undesirable.
But hey, remember that Wizardry didn't save dungeon progress but only character progress, and had random encounters involving random combat rolls etc.? So obviously, some significant adjustments are needed. However, randomness still plays an important role.
Most game designers and players will probably heavily object to constantly mimicking the restrictions that Wizardy games managed to impose. Barring some exceptions, I don't like unlimited respawning, for example. It makes sense for the local wildlife, merchants or wandering Orc tribes, but looks very artificial in a dungeon that I've just painfully finished to clear minutes ago. (Limited) random encounters are fine, ideally bundled with a way to avoid them, if your ranger has a high enough tracking skill, or some such. Random encounters that have invisible/moving spawn points and drop onto your party like paratroopers are the evidence of game design of another century.
If the lowliest opponent doesn't have even a miniscule chance of hurting your demi-god power gratification avatar, the game loses something important.
I agree, but I've already answered to that in one of my previous messages. One doesn't need dice to see level 1 kobolds (or whatever) killing a high level knight.
I'd say that randomness is the closest we've found to injecting soul into computer code; but who's to say randomness is not what accounts for our souls as well?
The RNG is no longer needed to find a soul in computer code. It's time to switch to something more appropriate for our times : proficient coders, and AIs.
And if there is a correct sequence of moves that always gets you the same outcome in the same situation, the game is worse than "deterministic", it is "memorizeable". Combat should always carry risk, even when executed by the numbers (in a manner that would make military tacticians shed manly tears of joy). And that is exactly what separates good cRPG combat from chess and other strawmen.
Take a black belted kung-fu champion fighting your average Joe in a bar brawl. There will be absolutely *nothing* random in the outcome of the battle (as I currently present it). There is absolutely *nothing* random in how Joe will get his teeth kicked out one by one, if he persists in attacking mindlessly round after round. What might change the outcome is if Joe somehow manages to surprise (flanking, charging) the champion with a devastating blow, if Joe switches to another weapon (a gun, or something that he can throw) that invalidates the short range skills of the champion, if Joe manipulated the champion into intoxication (alcohol, drugs) beforehand, or if Joe calls in buddies to surround the champion to slowly wear his stamina down over time.
Call it predictable if you like, it's what the game should ultimately replicate.
Now, you make a good point about meta-gaming. This one had me doubting for a pretty long time as well, before seeing the light.
Without going full derp on the LARPing stuff, if you've ever watched two people duking it out medieval style, and if these people were actually savvy enough to reproduce medieval fighting styles based on various historical literature that happened to reach us, you'll probably have noticed their simulated battle to be a fantastic and manly succession of "attack - counter - fall - counter/counter - stand up - push - feint - pull - attack" moves. Upon closer inspection, you may have associated this style of fighting to a competition ruled by R/P/S combinations. Counter rock with paper, paper with scissor, etc. These premises are trivial, hence easily memorizable, and easily transposable to cRPGs.
You may wonder how translating the randomness of an attack roll to the randomness of picking a premise (rock, paper, scissor) would result in a better game ? Well, here's why :
1/ the dynamic concatenation of the premises leading to successful outcomes is nothing trivially memorizable, or even achievable by your average Joe without long and proper training. Anyone can learn how to play Chess, but not everyone will become good at it, despite near infinite literature describing, annotating and analyzing games that have been played for centuries. Meta-gaming is therefore not an issue if the game is supported by extensive, robust game mechanisms, and an AI that knows how to handle them. The risk of the player failing will still be present, even more so that the AI is competent.
2/ 50% of what previously was decided by the RNG is now decided by the player :
If Joe = lazy player, or kung-fu champion = skilled player
-> the kung-fu champion wins
-> 50% of the outcomes that do not (ever) change
If Joe = skilled player, or kung-fu champion = lazy player
-> the kung-fu champion may have troubles
-> 50% of the outcomes that *may* change
This is a much more interesting paradigm than what Hollywood most "modern" cRPGs offer us nowadays : click on enemy, stand there while exchanging blows, maybe animate a couple of side-steps or extravagant combos here and there, and count on the RNG to fake somewhat "dynamic" battles.
3/ Keep in mind that part of the randomness remains within the damage rolls. Here is where you get the equivalent to your "oh shit" or "fuckin' damn right, brofist!" table top moments.
That it results in a reload when the dice don't go in the player's favor, when the same tactics result in a success when they do is obviously... suboptimal. There are different directions when handling this, but I'd say that letting the player reduce the influence of randomness on his outcome by smart play and planning is superior to letting the game reduce the influence of randomness across the board because the players can't be arsed to do it (learn to play?).
There's no such issue with a battle system that doesn't rely on dice. Reloads will remain as a commodity to the player that just got his ass handed to him, and would like to retreat to a safer haven, but that's it. No more save-scumming, to get "optimal" results, or bypass game mechanics. And everyone rejoiced in unisson.
Skill checks? I don't know, I haven't made any games or played any that would randomize skill thresholds like we mentioned before - but I'm still sure I would like to at least try a game like that before saying it's definitely pointless.
Trying it is adopting it. No one will make me believe that persuading someone is a matter of chaos/dice. Nor is it a matter of fate. It's a matter of training your social skills in the appropriate environment beforehand. If someone sucks at socializing, they will keep sucking if nothing else changes, would they keep trying till the end of time. If they're good at picking locks, they should succeed at their first try. I don't give a flying fuck about any "pseudo-realistic" chance of 5,5476476% that they might fail. Insignificant, hence pointless to have or keep.
Also, I don't like messing with the player characters (with random adjustments like J1M mentioned), because fundamentally, PCs are the player's input and responsibility (and the more control players have over their characters, the better), while the game world is solely the game maker's responsibility.
Please keep this in mind while reading what follows below, your own words : "fundamentally, PCs are the player's input and responsibility (and the more control players have over their characters, the better), while the game world is solely the game maker's responsibility".
And dunno, but having the game world be more "worldly" (=unpredictable and not-trivially-deterministic, also not trivially memorizeable) sounds like a clear win to me.
Just a sane reminder before proceeding. As I (among others) have said before, some randomness in cRPGs is good : encounter tables, loot tables, NPC behavioral patterns, weather, geography, etc. What I contest is the validity of the randomness in skill checks (whether during or outside battles). Populating the game world thanks to dice (or random seeds) is fine, and desirable for re-playability alone. Leaving battles to AIs built upon statistics, ideally using heuristics to adapt to the player style or wants (easy/medium/hard modes), is equally desirable. They do not lead to more predictable or trivialized encounters than what the current cRPG game mechanics offer.
See Firkraag in BG2 : fight him once, you will be surprised and probably crushed. Fight him twice or more, and you'll start to quickly see patterns, learn how to adapt to the spells he casts, or feats he uses, and manage to counter them eventually. The same fight conducted by an AI (worth of its name) may actually be less predictable because Firkraag, as a dragon, could reasonably resort to other tactics befitting such a prestigious beast : fight you in another, more favorable spot ? Fight you in a city, with human shields ? Call in minions, depending on how you proceeded in the dungeon beneath his lair (sneaking through, leaving most of the minions alive ? destroying everything in your path ? did you manage to bribe some minions to join your side instead ?).
Note that this only becomes possible with an AI. NPCs would no longer be reduced to protozoa randomly leaping out of their static spawning points. They would become true entities (mini-AIs) on their own, able to cooperate and tactically act at a higher level (a bit like ants).
Then, riddle me this: what use is a ruleset in a static experience? Why do you need rules for mechanics, if the "mechanics" in question are merely "if your Skill X is > Y, and you choose Option A that involves it, you win"?
(Captain Obvious responding here for a moment) I'd say that the use of a ruleset is the crux of it, like any game. How could you otherwise tell that a hit should connect if "X > Y", and miss in other cases ? Also, you're making it sound like as though using the option to "win" would be the unreasonable choice. If the odds are in favor of the player, he would be a fool not to exploit that advantage. Why, oh why, should the glorious beauty of such a simple concept be altered by some randomness, or worse obscure behind-the-scenes mechanics ?
(Back to serious) I can see what you're implying. However, it's anything but a static experience. You seem to believe that cRPGs are stale as soon as they no longer rely on fuzziness to solve contests, which isn't true. cRPGs aren't supposed to be played like some kind of Blackjack or Poker games. In those games, the input is necessarily random, otherwise there isn't any "game" left. Table top RPGs can be played without any kind of dice, and still be as exciting/fun (to most). Why not cRPGs ?
Let me prove it to you :
A is an armored knight, B is a weak kobold
Traditional concept :
A attacks B, dice roll, probable success
B attacks A, dice roll, probable miss
A attacks B with feat F1 (which has no consequence other than improving the odds of the attack), dice roll, probable success
B attacks A with feat F2 (which has no consequence other than improving the odds of a critical attack), dice roll, probable miss
...
My concept :
action : A attacks B at the torso (if B doesn't react, the attack will always succeed)
reaction : B defends against A with feat F3 (B spends stamina)
action : F3 allows B to counter-attack A with skill S1 (B spends stamina)
reaction : A cannot oppose S1, so it's A attacking, but B actually scoring a hit!
B manages to push A backwards, and finds an opening... at the armpit! Where an attack by B against A would have had no chance before, it suddenly becomes possible (with 100% success).
action : B attacks A at the armpit
reaction : A uses feat F4 to disarm B (A spends stamina)
A disarming B would not have previously succeeded because B still had 100% stamina left. Now it's possible (with 100% success).
action : B has no more feats/skills left to use at the moment, so A disarms B
reaction : B recoils and moves backwards, to grab another weapon (dagger) instead
...
As the player, the whole point of the "game", or rather the deterministic approach of it, should be to influence the odds of a contest as much in your favor as possible : through your intelligence, through your active/reactive input, through making sound decisions at the right time. Basically, making the best use of the resources at hand : stats, feats, skills, items, intel', geography, henchmen, etc.
Your skill at the game will be essential, crucial, dare I say indispensable to win. No matter how many times you reload, no matter how many dice could be rolled, if you suck at the game, you will keep losing. This makes a game fun to me. This is the clear opposite of games in which there is no point in managing your resources so carefully, since random rolls could very well destroy your best planned out tactics, or come to help you in dire situations that you shouldn't have been able to overcome.
If you're bothering to design actual mechanics, added fuzziness within their design parameters is a good thing. You could achieve that fuzziness with statistic modeling and whatnot, but it makes sense to use the simplest mechanic that achieves the desired effect (and in the case of percentile skill rolls, I think a (1d20-10)% adjustment would work as well). Without the "fuzziness" and unpredictability, you might as well not have any mechanics and reduce the whole game to a CYOA "make one of 1d6+1 predefined choices". And where's the "game" or "world" part in that?
To me, and within the boundaries of what we're currently discussing, it seems pointless. Pointless because most cRPGs will heavily tamper with that fuzziness. Pointless because you take out of the player's hands the control of a part of the game which shouldn't leave them. Pointless because non-random battle mechanics don't equate to "guaranteed" predictability, to more meta-gaming or any less fun, as shown above.
There is maybe one exception in the genre that is dear to some of us (me included) : rogue-likes. Why is not pointless in rogue-likes ? Because it's a genre in itself, that is played and enjoyed on its own merits, and especially because it doesn't tamper with randomness. There is no such thing as 50%, 30% or 10% randomness in them. It's brutal, it's harsh, it's 110%, it's *everywhere*. It's just not for every gamer.
For that reason, I pretend that cRPGs can't all be rogue-likes. They don't have to be rogue-likes, with fuzzy game mechanics, before being good or fun to play.
And before you get on your high "statistic modeling" and "AI" horse, consider this: what are random tables based on, if not quasy-statistic modeling and creative nonsense (remember Gygax's previous career)?
Statistics != RNG. The RNG doesn't create the tables, it operates them. Ideally, they should also be managed by the AI, but that's probably too much to ask for at the moment.
What does good AI operate on, if not tables of data and (quasy)statistic outcome prediction or heuristics? And if you have AI that can predict the outcome of actions much better than the player, should it always do so? Can the predictability of AI choices be used against it by players? I'd say AI that plays smart but with a decent degree of unpredictability would be much more interesting to play against than the alternative.
I know, there is some controversy about this, but I will leave it to quantum physicists to determine who has their preference among the two above. Or to quote yourself, for less controversy : "fundamentally, PCs are the player's input and responsibility (and the more control players have over their characters, the better), while the game world is solely the game maker's responsibility".
To make cRPGs fun/challenging/rewarding, make the player decisive, not dice rolls.
Are you advocating a system where it's immediately known whether you'll be successful in a particular action during combat (ex. generic-combat-skill 6 always hits against-generic-combat-skill 4), Karma? How do you have "chance" without randomization? Why does the use of randomization preclude the use of "clever" AI?
There's a chance my understanding of your posts is flawed, but that's what forum access at 1 AM is about.
