![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
mc776(these thoughts occurred to me after reading this post.)
People will follow this rule implicitly whenever the randomizers involves more than numbers: it may be a table with a hard list of possible outcomes, or a random name generator that prevents an unsyllabifiable mishmash of consonants and multiple consecutive apostrophes, or a random map generator that avoids spawning actors inside walls.
Three generally recognized sets of things that fall outside of desirable outcome range: absurdity, whiff factor and loss of agency.
Absurdity is simple enough: the range of possible results on various settings that must work together must be limited so they can work together - your character's name should not be nothing but a string of stops, and should spawn in a room rather than fused into solid rock.
Whiff factor is not really a thing when you're in a fairly matched game where the acting character's player truly stands a chance of losing: miss means one extra opportunity for the opposition to hit. However, in most long RPG campaigns, under a paradigm where the player has a strong story-related investment in their character, and absent any will to implement a mechanism to explicitly immunize characters from elimination from the long game if they fall in any given combat, the usual express or implicit policy is that all the fights will be weighted so the players win without any (permanent) casualties. This means that a player missing = the players win later rather than sooner.
Loss of agency is inherent to a randomizer: the whole point is that it picks something so you don't have to. In a more impartial setting than the "my guy" RPG above described, a difference in outcome between winning and not winning - i.e., losing - will go to the heart of the competitive appeal of the game, so no randomization that directly contributes to a victory/loss condition is desirable except where you specifically want something to be subject to risk.
Not all limitation on player agency is bad, however. At minimum a player should not have to dictate all the terms of the adversity they wish to overcome; often randomness is added to throw someone off their plan and force some risk management where the opposing player cannot be trusted to do the same. Then there's the plausibility/suspension of disbelief factor since we never see that kind of deterministic behaviour in the messy complicated clusterfuck we call real life. And then there may be a desire (more often in roguelikes, less often in bullet-hell shmups, specifically avoided in coin-op arcades to encourage repetition) to foreclose on the option of winning by memorizing any specific combination of actions, but instead force them to adapt to a new situation (albeit with generally the same rules) every time they fail, so that they must always be trying to dynamically apply rules rather than go by rote and trial-and-error.
And then, of course, there's "let the player win sometimes despite their incompetence so as not to hurt their feelings" argument. Granted, if someone wants to beat the game just to see how the singleplayer campaign ends without putting up with some nerdy voice doing inane commentary watching an LP, built-in cheat codes are vastly preferable for that purpose.
But I see another aim, or at least an extension of one of the previous ones: to continue generating more new situational content even after initial map generation and unit placement. Wildlife or bandits might wander into the field looking for easy pickings; a particularly devastating spell may alter the terrain in ways beyond the caster's direct control; a desperate caster might let loose some unknown magic they only know how to unleash but not control.
So some thoughts about how to make every move count without making something randomly not count:
And that's about all I can think of at the moment. Thoughts?
People will follow this rule implicitly whenever the randomizers involves more than numbers: it may be a table with a hard list of possible outcomes, or a random name generator that prevents an unsyllabifiable mishmash of consonants and multiple consecutive apostrophes, or a random map generator that avoids spawning actors inside walls.
Three generally recognized sets of things that fall outside of desirable outcome range: absurdity, whiff factor and loss of agency.
Absurdity is simple enough: the range of possible results on various settings that must work together must be limited so they can work together - your character's name should not be nothing but a string of stops, and should spawn in a room rather than fused into solid rock.
Whiff factor is not really a thing when you're in a fairly matched game where the acting character's player truly stands a chance of losing: miss means one extra opportunity for the opposition to hit. However, in most long RPG campaigns, under a paradigm where the player has a strong story-related investment in their character, and absent any will to implement a mechanism to explicitly immunize characters from elimination from the long game if they fall in any given combat, the usual express or implicit policy is that all the fights will be weighted so the players win without any (permanent) casualties. This means that a player missing = the players win later rather than sooner.
Loss of agency is inherent to a randomizer: the whole point is that it picks something so you don't have to. In a more impartial setting than the "my guy" RPG above described, a difference in outcome between winning and not winning - i.e., losing - will go to the heart of the competitive appeal of the game, so no randomization that directly contributes to a victory/loss condition is desirable except where you specifically want something to be subject to risk.
Not all limitation on player agency is bad, however. At minimum a player should not have to dictate all the terms of the adversity they wish to overcome; often randomness is added to throw someone off their plan and force some risk management where the opposing player cannot be trusted to do the same. Then there's the plausibility/suspension of disbelief factor since we never see that kind of deterministic behaviour in the messy complicated clusterfuck we call real life. And then there may be a desire (more often in roguelikes, less often in bullet-hell shmups, specifically avoided in coin-op arcades to encourage repetition) to foreclose on the option of winning by memorizing any specific combination of actions, but instead force them to adapt to a new situation (albeit with generally the same rules) every time they fail, so that they must always be trying to dynamically apply rules rather than go by rote and trial-and-error.
And then, of course, there's "let the player win sometimes despite their incompetence so as not to hurt their feelings" argument. Granted, if someone wants to beat the game just to see how the singleplayer campaign ends without putting up with some nerdy voice doing inane commentary watching an LP, built-in cheat codes are vastly preferable for that purpose.
But I see another aim, or at least an extension of one of the previous ones: to continue generating more new situational content even after initial map generation and unit placement. Wildlife or bandits might wander into the field looking for easy pickings; a particularly devastating spell may alter the terrain in ways beyond the caster's direct control; a desperate caster might let loose some unknown magic they only know how to unleash but not control.
So some thoughts about how to make every move count without making something randomly not count:
- Your hit always does X damage, subject to circumstantial modifiers, but a random portion of it damages MP instead of HP.
- Your hit always does X damage, subject to circumstantial modifiers, but a random portion of it debuffs a stat instead of damaging HP.
- Your hit always does X damage, subject to circumstantial modifiers, but a random portion of it is added to one's own HP or MP instead of damaging the target's HP. (I should note at this point that this is the opposite of what I tend to find myself writing, generating a random damage value and then deliberately or deterministically apportioning that among the possible effects.)
- An action may have a non-random number of results, picked from a random table, with multiple pickings of a single result having cumulative effect if possible, and if that's not possible (e.g. yes/no status effects that last the entire map or until specifically removed) you just roll again unless there are fewer available outcomes than effects to be picked, in which case that additional effect is lost.
- There's still a chance to miss, but the shot then changes one or more terrain tiles on, adjacent to, or somewhere past the target's tile. This should only work for especially powerful weapons (fireball spell good, .22LR bad) as the terrain alteration should be pretty significant for it to mean as much as a hit.
- AoE has a slightly different shape around the edges each time.
And that's about all I can think of at the moment. Thoughts?