Further dynamic tweaks: "you can instamerge this time" edition (#11715)
* Starting a replacement of how threat works. * no, we do it this way * Added threat levels to jobs * Added threat to... a lot. * Updated for traitor classes. * Fixed errors, except for one. It's consistently giving me "maximum number of internal arrays exceeded (65535)". I have no idea what could be causing this. * Added type annotation to GetJob. * This one I should change though * wow how'd that happen * spammable means low threat * Made story threat have initial threat level on average * Made somet rulesets force if they won the vote * ) * Gave EVERY job threat, added a config for it. * Rebalanced some numbers * Update code/game/gamemodes/dynamic/dynamic_storytellers.dm Co-Authored-By: Ghom <42542238+Ghommie@users.noreply.github.com> * Removes mush threat * Makes devil threat scale with form * reviewing reviewer's review of reviewer * Gutlunches can be friendly spawned, so no * Also made forced-friendly mobs not count * null checks better * Made antag threats in config, too * various fixes * Another couple dynamic fixes * Made an admin message chunk all one line. * Make roundstarts ignore current threat It's not even calculated yet, so this is probably better. * Minimum pop for chaotic/teamwork. * More conveyance issues, removed superfluous threat costs * More conveyance and tweaks * Makes storyteller min players use all players instead of ready * Lowered chaos weight with chaotic * Blob now has correct cost * Makes phylactery count for threat * Makes random storyteller have random threat level * Made starting rulesets scale up with threat LEVEL * Made "minor rulesets" never have lower weight * Makes chaotic not forced. * Made story about 25% less chaotic Co-authored-by: Ghom <42542238+Ghommie@users.noreply.github.com>
This commit is contained in:
@@ -511,7 +511,7 @@ GLOBAL_VAR_INIT(dynamic_storyteller_type, /datum/dynamic_storyteller/classic)
|
||||
drafted_rules -= starting_rule
|
||||
|
||||
starting_rule.trim_candidates()
|
||||
starting_rule.scale_up(extra_rulesets_amount, threat)
|
||||
starting_rule.scale_up(extra_rulesets_amount, threat_level)
|
||||
if (starting_rule.pre_execute())
|
||||
log_threat("[starting_rule.ruletype] - <b>[starting_rule.name]</b> [starting_rule.cost + starting_rule.scaled_times * starting_rule.scaling_cost] threat", verbose = TRUE)
|
||||
if(starting_rule.flags & HIGHLANDER_RULESET)
|
||||
|
||||
@@ -241,5 +241,5 @@
|
||||
requirements = list(10,10,10,10,10,10,10,10,10,10)
|
||||
high_population_requirement = 10
|
||||
repeatable = TRUE
|
||||
flags = TRAITOR_RULESET
|
||||
flags = TRAITOR_RULESET | MINOR_RULESET
|
||||
property_weights = list("story_potential" = 2, "trust" = -1, "extended" = 2)
|
||||
|
||||
@@ -53,6 +53,8 @@ Property weights are:
|
||||
var/turf/T = get_turf(H)
|
||||
if(H.stat != DEAD && is_station_level(T.z) && !("Station" in H.faction))
|
||||
threat += H.threat()
|
||||
for(var/obj/item/phylactery/P in GLOB.poi_list)
|
||||
threat += 25 // can't be giving them too much of a break
|
||||
for (var/mob/M in mode.current_players[CURRENT_LIVING_PLAYERS])
|
||||
if (M?.mind?.assigned_role && M.stat != DEAD)
|
||||
var/datum/job/J = SSjob.GetJob(M.mind.assigned_role)
|
||||
@@ -135,17 +137,19 @@ Property weights are:
|
||||
if (GLOB.dynamic_classic_secret && !((rule.flags & TRAITOR_RULESET) || (rule.flags & MINOR_RULESET)))
|
||||
continue
|
||||
rule.trim_candidates()
|
||||
var/cost_difference = abs(rule.cost-(mode.threat_level-mode.threat))
|
||||
/* Basically, the closer the cost is to the current threat-level-away-from-threat, the more likely it is to
|
||||
pick this particular ruleset.
|
||||
Let's use a toy example: there's 60 threat level and 10 threat spent.
|
||||
We want to pick a ruleset that's close to that, so we run the below equation, on two rulesets.
|
||||
Ruleset 1 has 30 cost, ruleset 2 has 5 cost.
|
||||
When we do the math, ruleset 1's threat_weight is 0.538, and ruleset 2's is 0.238, meaning ruleset 1
|
||||
is 2.26 times as likely to be picked, all other things considered.
|
||||
Of course, we don't want it to GUARANTEE the closest, that's no fun, so it's just a weight.
|
||||
*/
|
||||
var/threat_weight = 1-abs(1-LOGISTIC_FUNCTION(2,0.05,cost_difference,0))
|
||||
var/threat_weight = 1
|
||||
if(!(rule.flags & MINOR_RULESET)) // makes the traitor rulesets always possible anyway
|
||||
var/cost_difference = abs(rule.cost-(mode.threat_level-mode.threat))
|
||||
/* Basically, the closer the cost is to the current threat-level-away-from-threat, the more likely it is to
|
||||
pick this particular ruleset.
|
||||
Let's use a toy example: there's 60 threat level and 10 threat spent.
|
||||
We want to pick a ruleset that's close to that, so we run the below equation, on two rulesets.
|
||||
Ruleset 1 has 30 cost, ruleset 2 has 5 cost.
|
||||
When we do the math, ruleset 1's threat_weight is 0.538, and ruleset 2's is 0.238, meaning ruleset 1
|
||||
is 2.26 times as likely to be picked, all other things considered.
|
||||
Of course, we don't want it to GUARANTEE the closest, that's no fun, so it's just a weight.
|
||||
*/
|
||||
threat_weight = 1-abs(1-LOGISTIC_FUNCTION(2,0.05,cost_difference,0))
|
||||
if (rule.ready())
|
||||
var/property_weight = 0
|
||||
for(var/property in property_weights)
|
||||
@@ -168,8 +172,10 @@ Property weights are:
|
||||
|
||||
rule.candidates = list(newPlayer)
|
||||
rule.trim_candidates()
|
||||
var/cost_difference = abs(rule.cost-(mode.threat_level-mode.threat))
|
||||
var/threat_weight = 1-abs(1-(LOGISTIC_FUNCTION(2,0.05,cost_difference,0)))
|
||||
var/threat_weight = 1
|
||||
if(!(rule.flags & MINOR_RULESET))
|
||||
var/cost_difference = abs(rule.cost-(mode.threat_level-mode.threat))
|
||||
threat_weight = 1-abs(1-(LOGISTIC_FUNCTION(2,0.05,cost_difference,0)))
|
||||
if (rule.ready())
|
||||
var/property_weight = 0
|
||||
for(var/property in property_weights)
|
||||
@@ -200,7 +206,7 @@ Property weights are:
|
||||
weight = 1
|
||||
event_frequency_lower = 2 MINUTES
|
||||
event_frequency_upper = 10 MINUTES
|
||||
flags = WAROPS_ALWAYS_ALLOWED | FORCE_IF_WON
|
||||
flags = WAROPS_ALWAYS_ALLOWED
|
||||
min_players = 40
|
||||
var/refund_cooldown = 0
|
||||
|
||||
@@ -245,7 +251,10 @@ Property weights are:
|
||||
config_tag = "random"
|
||||
weight = 1
|
||||
desc = "No weighting at all; every ruleset has the same chance of happening. Cooldowns vary wildly. As random as it gets."
|
||||
forced_threat_level = 100
|
||||
|
||||
/datum/dynamic_storyteller/random/on_start()
|
||||
..()
|
||||
GLOB.dynamic_forced_threat_level = rand(0,100)
|
||||
|
||||
/datum/dynamic_storyteller/random/get_midround_cooldown()
|
||||
return rand(GLOB.dynamic_midround_delay_min/2, GLOB.dynamic_midround_delay_max*2)
|
||||
@@ -318,9 +327,10 @@ Property weights are:
|
||||
property_weights = list("story_potential" = 2)
|
||||
|
||||
|
||||
/datum/dynamic_storyteller/story/do_process()
|
||||
/datum/dynamic_storyteller/story/calculate_threat()
|
||||
var/current_time = (world.time / SSautotransfer.targettime)*180
|
||||
mode.threat_level = round(mode.initial_threat_level*(sin(current_time)+0.5),0.1)
|
||||
mode.threat_level = round(mode.initial_threat_level*(sin(current_time)+0.25),0.1)
|
||||
..()
|
||||
|
||||
/datum/dynamic_storyteller/classic
|
||||
name = "Classic"
|
||||
|
||||
Reference in New Issue
Block a user