stim and pkill stats are big simplifications that cause numerous problems:
They are capped per-application, meaning that aspiring+heroin works better than heroin+aspirin
stim is single axis: you can take mind-wrecking depressants and counter them with coffee to end up sober
They time out separately from effects: meth will time out before meth stim does, meaning that you stay stimulated after the crash
They motivate hacks like coffee "curing" tiredness. This may be considered a good thing, though - being able to stay permanently awake may be a good gameplay mechanic.
They motivate hacks like meth curing fatigue and then causing it back, resulting in it being a sleep aid
Alcohol metabolism is hacked in rather than being more "natural" alcohol->nutrition transfer
Alcohols, unlike painkillers, act immediately
Addiction to painkillers decreases pkill effect from all sources, such as pain dulling by a bionic
Vomiting clears stim and pkill in bloodstream and from bionic pain dulling
19986
Bonus (separate issue):
They are separate from morale effects of the relevant substances
Idea of a fix:
stim and pkill become effect properties
To get current stim and pkill, we'd accumulate those effects somehow. I see two options: sum all and apply some non-linear function, or sum the strongest positive and the strongest negative (say, [-1, 2, 3, -2] becomes 3 + (-2)). Pure sum would be risky, as it would most likely result in cherrypicked "designer drug combos" which could be tedious to apply and make little sense due to per-effect caps.
To maintain slow release pkill, effects gain ability to cause other effects. For example, "heroin in bloodstream" would apply (increase duration of) "hard painkiller", "bleeding" would apply "blood loss", "alcohol consumed" would apply "drunk".
(Optional) Per-effect morale bonus. So that morale bonus from vodka comes from being drunk and not the taste and wears off with the drunkenness rather than time itself
Results:
Aspirin+heroin both time out independently, order of application doesn't matter
Coffee+vodka wear off separately, possibly at different rates
Vomiting/blood filter can clear only the relevant effects (you can't vomit heroin out of your veins)
All drugs for which it makes sense can become slow release. This one is risky for gameplay - the "drugs as potions" thing is somewhat protective from the "one slip and YASD" gameplay.
More natural formulation of exhaustion after drugs. Not an issue for meth where we want it to restart, but with adrenaline you want to wait out the whole effect (including comedown) before re-applying.
Non-hardcoded formulation of effects that progress rather than regressing. ie. bite wound->infection
Potential problems:
The stim/pkill sum formula. This needs a good idea, otherwise aspirin will still work when on heroin
Too liberal application of slow release effects could be one of "those" "features". Not really an issue with the proposal itself, but side effects have to be considered too.
The effect-causing-effects thing needs a syntax that is flexible enough to allow inheriting body parts (bite wound->infection) and changing them (bleeding arm->global blood loss), but is readable enough for typical case
stim
andpkill
stats are big simplifications that cause numerous problems:stim
is single axis: you can take mind-wrecking depressants and counter them with coffee to end up sober19986
Bonus (separate issue):
Idea of a fix:
stim
andpkill
become effect propertiesstim
andpkill
, we'd accumulate those effects somehow. I see two options: sum all and apply some non-linear function, or sum the strongest positive and the strongest negative (say,[-1, 2, 3, -2]
becomes3 + (-2)
). Pure sum would be risky, as it would most likely result in cherrypicked "designer drug combos" which could be tedious to apply and make little sense due to per-effect caps.Results:
Potential problems: