Open Lasercar opened 8 months ago
Sorry it's taken me so long to reply!
I always love to see more info about this game. The bit on pet damage is particularly intriguing, though I assume it's just damage on quests? Per one of the loading screen tips, "Bring pets on quests to give Dwellers a damage boost!" Also, while that shows how damage of the pets compare to each other, how does it compare to weapon damage? I'd assume it's 1, 2, or 3 damage using the same scale of weapon damage that we see, making it a small boost, but I could also see it being a different scale. The part on Mr. Handy makes me question if it's the same scale, as I feel like he doesn't do much damage at all, not 10 damage worth.
On everything else, it's definitely common for unused code and assets to be left in game files. Maybe it was used just for debugging, maybe it was cancelled because they didn't have the time or felt like they couldn't make it balanced or fun, maybe they thought it was adding more complexity than they wanted, etc.
While I'll do my best to update the FAQ with new info, I don't feel like digging through code, so I'll pass there.
Yeah, the pet damage is for rooms too, and I checked the prefabs of the parameters and mr handy is set to deal way less damage than the default ones (here's the normal emergency parameters): Though pets deal way more??? That, and there's some other nice values, like how long each moving incident does battle in a room. Also, the room damage multiplier values.
There aren't that many changes with the survival parameters, other than with incidents and such (it all fits into one image too) But geez, fires dealing 4x the damage? no wonder why they hurt
Max damage per cycle is plugged into this average level adjusted damage function to get the incident's damage:
Damn, there's a lot of useful info there, thanks for finding all this!
On pets, it is showing that dweller are getting damage from their weapon, GetBonusForEffect, and GetFollowersBonusDamage. I'm not sure that alone is confirmation that pets of any kind give a damage boost in the vault though. The followers damage bonus could be the part that checks if they specifically have a damage pet, while the bonus effect is something else (or vice versa). I'm skeptical because the 2, 4, and 6 damage for common, rare, and legendary lines up with the bonus the damage specific pets provide. I guess that would mean damage pets do 4, 8, or 12 damage, but that seems like far too much. I've grinded out a number of "defeat an enemy without weapons" objectives, by rush spamming a level 1, single size room with damage pets. It can still take a while to kill things, which I feel just wouldn't be the case if they were getting 8 or 12 damage from the pets.
I'll have to make some updates with a lot of this incident info. Seeing the actual chances, confirmation that opening the door increases the chances of deathclaws, the population thresholds, Mr. Handy damage, etc. Also yeah, that definitely explains why fires seem like the worst incident in survival. Might be a bit before I update the FAQ though.
No, I'm quite certain that the pets do the extra damage specified. I even did a spontaneous test with a random quest - a sawed-off shotgun did way less damage than fists + alien drone. Also, I've found the wasteland logic stuff... The junk cycle uses the wasteland encounter script (everything in it does). That means, you're basically going to get crappy junk whether you like it or not, as your dweller's level and weapon damage are used to test the encounter's success: If a pet with a damage bonus is equipped, it adds the numbered pet damage bonus 2/4/6 to the min and max damage, so if you want to (nearly) guarantee all (2/3) legendary junk encounters, add a +6 damage pet. Also, the SumOfRandomInts is a Random.Next, meaning, having 1 in a stat will have zero effect - it can only roll a zero twice. The weapon is similar with the Unity.Random.Range it uses, but that's been accounted for by adding one extra.
More things, The item cycle has code to drop pets (if they were included in the GetRandomItemType)! So, they were originally going to drop like outfits and weapons (though only common/rare ones), before someone got the grand idea of putting them into their own, rarer lootbox... (item luck is a DateTime.UtcNow.Millisecond, so no luck influence) Also, despite the code allowing it, the localizer keys for the weapon-outfit/pets and rarity only have one in each... lol. Guess no one could be bothered typing out a few more lines.
And yes, item and junk cycles are confirmed to 60 and 180 minutes, respectively with slight variation every drop aka cycle. That, and also, no legendary threshold, period (it's set to zero). However, it does mean that rares were "buffed" (referring to when those changes happened) to drop twice as often in item cycles (rare threshold = 0.2). hooray, I guess.
The junk cycle encounters. Special stat, difficulty value, min, max time (in minutes), failure damage, failure exp, item, success damage, success exp:
The localization for the pet drops (if we had them): I rescued a {0} from a Raider trap and now it's following me. A {0} just appeared and saved me from walking into a pit! Better take it with me to watch my back.
There's also no stat influences on the item chances either (that means perception's 'Huge contribution to finding weapons/outfits/recipes' is false). For recipes, if the weapon/outfit factory is unlocked, then it can run the recipe chance function - which creates a float between 0.0 and 1.0. If this value is greater or equal to the recipe chance below (divided by 100 float), then it grants a recipe rather than a weapon/outfit. 25% of common raider drops are converted to rare recipes, while 25% of rare drops are converted to legendary recipes. 5% of wasteland drops are converted to rare recipes, while 25% of rare drops are converted to legendary ones. Raider rares are a 5-25% chance depending on the average level (5% increase per 1-10 level bracket), so while raiders can drop more rares than the wasteland, you need your average level to be 41-50. For legendary recipes to drop, you need the rare upgrade of the weapon/outfit room first.
Anyway, I'd be interested to see some math for the legendary junk. Which outfit, with or without the 2/4/6 pet damage bonus effect, is the best-best now how we know how works? As in, which one would give the lowest chance of failure for all three legendary junk events.
Here's the wasteland creatures. There are 27 of them (x2, for the different progress descriptions and 50% chance of zero exp on fail, with some giving different amounts of exp). Brackets are survival mode values. Special stat, difficulty value, min, max time (in minutes), failure damage, failure exp, success damage, success exp - variant fail/win experience:
Locations Same as the creatures, just that they're one-time events as they're locations.
NPCs
Oh, one more thing I found (though not by looking through the asset ripper generated unity project) - missile launchers fire faster than the fat man (that, and the projectile is way faster and targets their enemy, not the room), meaning they deal more damage. They have trouble continuously firing at 15A. Yeah. I even tested 3 miss launchers and it basically wasn't that much different to having 3 vengeances/ 2 vengeances and a MIRV.
A bit of an offtopic comment to this, but do you think you could link this forum post of some suggestions I've written up to Bethesda through the support center on their website? (I don't have an account) https://steamcommunity.com/app/588430/discussions/0/4352240417146399507/?tscn=1712368771
I noticed that you updated the wasteland guide referring to this reddit post: https://en.reddit.com/r/foshelter/comments/1akw2dc/comprehensive_wasteland_outfit_testing_on_max/
So, I got a little curious myself and checked the 1.15 wasteland logic. It's total chaos.
There's tons of things out of wack for some reason. It's like someone (like an intern) or something (like unity) f'd it up. For example Creature 27 (Super Mutant Behemoth) has the following different values: P, 14, 90, 605, with the fail/win unchanged. Meanwhile Creature4 (Radroach) has values of S, 16, 110, 740, and Creature4B, unlike Creature27B, also has changed values - I, 22, 200, 1370. WTH?
Special stat, difficulty value, min, max time (in minutes), failure damage, failure exp, success damage, success exp B = alternative/variant:
While the locations and NPCs have been spared, the junk cycle has not. Every single junk cycle bar one normal that starts at 600 minutes, now start at 1200+ with equally high difficulty values:
Also, everything now counts as a creature too (and I mean everything - locations, junk cycle locations, npcs), because why not??? Meaning, they count for the kill wasteland creatures objective.
The above changes make it very easy to exclude most of the normal junk drops, that and since there's no strength check, aside from Hubris Comics headquarters, you can basically max that one out. Basically, max strength, endurance, charisma, luck, minimum perception and only equip a 15 damage weapon. The best normal mode junk outfit with these stats is Famine's vestment - only 3 of the 16 possible intelligence rolls will make the dweller succeed in the intelligence junk drops. That said though, is it worth excluding the normal and rare junk drops or gear your dweller to succeed in nearly every legendary junk event they get?
Survival mode has been spared somewhat from these changes, as the difficulty and min/max times get replaced, but the special stat changes still pass through.
The objective manager prefab is different too (larger size), but I can't be bothered thoroughly check for any differences.
Wow, thanks for all the work here! There's a lot to read and unpack. I really do appreciate all the work you've put in here, but it could be a while before I get around to updating the FAQ with all of this.
With all of the encounters, I don't understand what the difficulty even means. If there's a SPECIAL stat associated with it, then I assume that goes into it. You also mention weapon damage affecting it. But like, the Savage Dog has a difficulty of 1(2) compared to the 52(68) of the Super Mutant Behemoth. How does the SPECIAL stat and weapon damage translate into beating the difficulty levels? You mention elsewhere that a difficulty of 83 is almost impossible, but I have no frame of reference for why that is.
I'm confused about this sentence from you:
"There's also no influences on the item chances or re either (that means perception's 'Huge contribution to finding weapons/outfits/recipes' is false)."
I'm guessing the "or re either" is a typo, but I can't make out what you mean. Also, what specifically do you mean by no influence? You also say that "item and junk cycles are confirmed to 60 and 180 minutes, with slight variation every drop aka cycle". Just to be clear, the understanding before this was that dwellers got guaranteed weapon/outfit/recipe drops every 60 minutes, while there were other events sprinkled throughout that checked against a SPECIAL stat. Then for junk, that there were events every 160 minutes plus events sprinkled throughout, and both checked against a SPECIAL stat.
I see the events you've listed for weapons, outfits, recipes, and junk. Are these part of these 60 and 180 minute cycles (with the game picking one if it's within the listed min and max time), or are those additional events? Is any loot guaranteed, or do all have some sort of difficulty check? Either way I can see from the events you listed there are barely any perception ones, so I do see the belief we had that it affected weapon/outfit/recipe drops a lot is false.
That's crazy on the mobile differences with creatures. Just to be clear then, the rest of your data is from game version 1.13? As we've uncovered a good bit of differences between 1.13 and 1.15, I want to keep in mind what version all of this is as I make updates. Also, on the posts that found out about ECL and legendary junk in normal, they also found experimentally that agility and luck made the biggest difference for survivability. But looking at the creature data here, I'm not seeing that either stat particularly dominates. I'm tempted to trust the experimental data over what the code supposedly says though, so I'm not sure what to make of things.
Okay, I missed the encounter function. I'm rusty on floats, and I'm especially not sure why they take the dweller level, convert it to a float, convert it back to an int to assign it to num2
, then convert it back to a float to add it to num
. From what I can gather, it adds their level divided by 2, a random number between their min damage and max+1 damage, and whatever the SumOfRandomInts
function does. I assume the effectiveStat
is the value of the required SPECIAL, so say the event is for E and they have 15E, then the function has the inputs (0, 15, 2)
, and I don't know what that means. I can find documentation for RandomHelper
, but not that specific function.
60 and 180 minutes, respectively.
Yes, the rest of the info aside from that 1.15 wasteland logic comment is using 1.13. The only difference between 1.13 and 1.15 is the new quests (and locations/rooms for them, quest enemies, new weapons/outfits/dwellers and alien raiders, and of course, the wacky wasteland logic edit.
As I stated up there, every creature has a variant - meaning those 6 agility enemies is technically 12 (out of 54 possible creatures), that, and every creature has a time that they start appearing and when they stop, so not every creature is able to be in play at any given time, plus, those A (well, 3/6) enemies and the L deathclaws are both difficult and hurt hard, so they have an outsized influence on survivability, that is, before 1.15 screwed up them, as you can see.
The respectively isn't what I was asking for clarification on, I got that. Our previous understanding was that you ultimately had four different kinds of events. Repeated weapon/outfit/recipe events, fixed ones, repeated junk events, and fixed junk events. The repeated events were the 60 and 180 minute timers respectively (though we thought junk was 160). But then the fixed events were on top of the events you were getting every 60 and 180 minutes, like the National Guard Depot being 60 hours.
So to clarify, the only events are an item event roughly every 60 minutes, a junk event roughly every 180 minutes, and then enemy and NPC encounters?
So to clarify, the only events are an item event roughly every 60 minutes, a junk event roughly every 180 minutes, and then enemy and NPC encounters?
Yep, and the location encounters (the non-junk cycle ones) Sorry about the confusion.
And the locations? Are the 17 locations not events for the 60 minute cycle, while the 12 junk events are for the 180 minute cycle?
On creatures I guess a good bit of the top ones are A or L, and the post that found about it was in survival, where I guess Overlords and Behemoths aren't really encountered either looking at the minimum times. I booted my 1.13 survival vault to check, and all of my returned explorers had fought Masters, but I didn't see a single Overlord or Behemoth in any of their logs.
Locations are just NPCs, but you know, locations
So what events trigger on the 60 minute cycle? Or did you not list them here?
The 60 minute cycle is only for the weapon/outfit drops (and recipes ofc, but they're just converted weapon/outfit drops). Locations and NPCs are basically one time creatures (with rewards/objective completion scripts added to the success outcome), if that sense.
Okay, I think I get it now. Again, to clarify:
Yes, you got it now.
Awesome, thank you!
Also, on all the encounter difficulties. What does the SumOfRandomInts
function do? I can't find documentation for it. I assume it gets passed in the relevant stat for the event, so say a dweller has 15E and it's an E event, it gets passed in (0, 15, 2)
, but I can't discern the output.
Ah. That.
Also, the SumOfRandomInts is a Random.Next, meaning, having 1 in a stat will have zero effect - it can only roll a zero twice. The weapon is similar with the Unity.Random.Range it uses, but that's been accounted for by adding one extra.
I typed a bit explaining it.
I assume it gets passed in the relevant stat for the event, so say a dweller has 15E and it's an E event, it gets passed in (0, 15, 2), but I can't discern the output.
Yep.
Ah, you literally explained it right below, and I'm an idiot and missed it, haha.
The range must be exclusive then if having a stat at 1 makes it so it rolls 2 zeroes then?
Yes. For example, having a stat of 17 allows for a 17x17 table of the possible total rolls (between 0-16) While it could roll a double zero, that's only 1 of the 289 options with such a stat - 0.003, 0.3% out of 100% Having a stat of 2 allows for 2x2 table of possible total rolls (between 0-1) - 0,0/0,1/1,0/1,1
That makes sense then that a difficulty of 83 is near impossible. +25 from being level 50. Dragon's Maw with a +6 damage pet gives 34-42, putting you at 59-67. You'd need 24-16 more, which especially if you don't have 17 in the stat would require lucky rolls.
Yep, though granted, no dweller is going to reach that far due to all the drops.
Yeah. Would be worth analyzing how much damage pets are worth it then for legendary junk, assuming level 50 explorers with a Dragon's Maw. One of the thing I'll likely do whenever I get around to updating the FAQ.
It's nice to see this does confirm the findings of the old ECL post too. They found that C past 10 seemed to not boost the chance of getting legendary junk from its event, while boosting E and L did from theirs. Looking at the difficulties, the C event is lowest at 50, E at 55, and L and 60. So that makes sense. Plus as they found that SPECIAL seemed to not matter in survival, that also makes sense considering the difficulties are far lower there at 24, 26, and 29. That's guaranteed with a level 50 explorer and any weapon that isn't crap.
Yep. Also, I was going to make a post detailing some information about the quests, though once I got to the point of looking for the wasteland quests so I could document when each one can happen and at what levels and their chance, well, it seems like the version of AssetRipper I currently have doesn't get the variables for the quests themselves for whatever reason, as sip.
So, I noticed that the wasteland quest stuff is handled alongside the rest of the game parameters, so here's the info on that. First, the conditions for the surprise quest (yes that's what the code calls it) check to run in the first place - you can't be in an emergency, the tutorial must be finished, you must have an overseer room, and you have to be online for 30 minutes before it's allowed to start surprising you with quests (running the check function below). It picks the first valid team for a surprise quest when it passes the chance A valid team is one that's either exploring or surprisingly, currently going into one, though they can't be dead of course, that'd be weird. Surprise quest chance: About what I expected, more (exploring) wasteland dwellers, higher quest chance. After the minimum 30 minutes, you're able to get further surprise quests after every 20 minutes past that.
Looking back at the ECL post, they say damage pets are not worth it, seemingly cause it's causing earlier events to succeed more often before the explorer can get to the legendary junk events.
On quests, it'd be worth checking out this. robot9706 (who made the popular FSSE) has a list of game data here, the quest section being filled with info. It's how we've gotten a lot of our current understanding of quests.
https://robot9706.github.io/FSSE/game_data.html
On being online for 30 minutes, that just seems blatantly wrong. Almost all the time I find you can load into the game with active explorers, and get a random quest within the first minute of being there if it's been a while since you last played. At a glance it looks like no SPECIAL affects the ability to find them though, which is so wonderful to finally know. People spread everywhere that perception helps you find them, and I've tried to fight it saying we don't know, it may help, it may not.
Ah, yeah, the minimum 30 minutes is after you close and open the game (and a vault) again.
So you could load into the game, do a random quest, quit, and then load back in ~30 minutes later for another random quest?
Also I don't understand your graph. Why the two separate lines and no X-axis? Is that saying that for example at 6 dwellers exploring there's an 11% chance for a random quest? 11% how often?
So you could load into the game, do a random quest, quit, and then load back in ~30 minutes later for another random quest?
Yep, though every 20+ minutes.
Also I don't understand your graph. Why the two separate lines and no X-axis? Is that saying that for example at 6 dwellers exploring there's an 11% chance for a random quest? 11% how often?
Every 60 seconds.
Okay, so I am reading the graph right then. I see this also further confirms what we learned from robot9706, that explorer level determines the difficulty of the quest they get. Though we were fairly confident in that already through all our testing.
Ah yeah, looks like I'm wrong on the 30 minute delay - it's a 30 minute delay into a new save before you can start getting quests, though that seems a bit weird, as you can't exactly get 18 dwellers and enough caps for an overseer room in 30 minutes, but whatever. Perhaps they had the room at a lower dweller requirement? (like, idk, 6?) You could test this with a new save and the save editor.
Also, I've updated the junk cycle locations with their names.
And, I got this screenshot of my 40 min difficulty guy (stats: 1+2 / 1 / 10+2 / 10 / 2 / 10 / 10+2) failing the 45 difficulty, intelligence check of the sealed room on 1.15 (aka mobile). He also only has 7 normal junk (98% double), despite literally doubling every one he succeeds - My other explorers with vengeances/dragon maws have the same amount or more at this same time (and, that 7 junk hasn't changed since the 7 hours I've taken the screenshot). So yeah, this is basically the strat, if you can find (or just procreate) a normal dweller with in 1 perception and intelligence. Then max out ECL and SA for wasteland survivability quest attack rate. Then send them into the wasteland with robot armor (or any other one that doesn't add perception/intelligence, but rather S/E/C/A/L) and a 15(-15) damage weapon. This'll ensure that they never succeed in 5 of the 12 junk events. You could drop the agility, and their weapon by 5, to also ignore the Children of Atom camp, but I don't think it's worth it considering the difficulty of the 3 legendary junk events. That and they to reach those and hit the inventory cap before dying.
Oh, one more random thing (and also edit to this comment) - if you have multiple mobile devices, each of them can do the 4 hour adreward, meaning, the more devices of the same mobile OS you have, the more rewards you can pile into a single save every four hours using cloud saves (though always close fallout shelter afterward - it doesn't time down if put the device to sleep with fallout shelter open).
Anyway, I haven't gotten any more ideas of what else I could/should check yet. Though, maybe you have ideas on what I could/should look at?
Anyway, I haven't gotten any more ideas of what else I could/should check yet. Though, maybe you have ideas on what I could/should look at?
SPECIAL stat details
Anyway, I haven't gotten any more ideas of what else I could/should check yet. Though, maybe you have ideas on what I could/should look at?
SPECIAL stat details
Uh, not very helpful... There's only a few things SPECIALs do - reducing production room cycles (and the 'right room' happiness bonus), wasteland encounters and quest requirements - perception for crit speed reduction, endurance for health, charisma for speedy dweller breeding, agility for quest attack rate and luck for caps and faster crit chances. (There's also an unused in quest SPECIAL checks system aka, quest challenges. Would be interesting to see how they'd play in a custom quest, but this exported project is all out of wack - the menu's and the menu's camera is broken, and it's missing all the quest data, for example.)
Anyway, I found some more things: The uh.... function for calculating the production room luck caps. I can't be bothered figuring out how this works right now. Dweller compatibility is a dice roll with the Charmisa/last partner factoring in to make the check easier, though if it fails you can just cycle them from the room to try again.
The dweller related game parameters:
The deathclaw incident chance is checked every time the door is opened or the radio calls a dweller, so there's that. Also, check out this: Yeah, these were the default parameters for the radio room. And it looks like, the code for spawning legendary/rare dwellers wasn't edited or commented out, so technically, those reports of legendary/rare dwellers from radios might be a very rare bit change occurrence. I can remember getting a legendary dweller I already had by radio once, though maybe it was just a rescue quest? idk.
Anyway, here's the radio room success and additive deathclaw rates (0.05 + this) Also, table as I've just learned that github has them. | Radio room | Level 1 | Level 2 | Level 3 |
---|---|---|---|---|
Size 1 | 0.8, 0.02 | 0.1, 0.04 | 0.13, 0.06 | |
Size 2 | 0.18, 0.03 | 0.1, 0.05 | 0.13, 0.07 | |
Size 3 | 0.3, 0.08 | 0.1, 0.11 | 0.13, 0.15 |
That, and 20/40/60 total charisma halves the call time, but that's probably common knowledge. And yes, the legendary/rare dweller chances on all 9 are set to zero. Though, I wouldn't be surprised if in older versions that there was one or two that still had chances for them.
Surprise Quest Chance: Exploring Dwellers | Trigger Chance (out of 100) |
---|---|
0 | 0 |
1 | 2 |
2 | 4 |
3 | 6 |
4 | 8 |
5 | 10 |
6 | 11 |
7 | 12 |
8 | 13 |
9 | 14 |
10 | 15 |
11-15 | 16 |
16-20 | 18 |
21 | 20 |
Quest Revive Cost (difficulty range) Min Range | Max Range | CAPS Cost |
---|---|---|
0 | 10 | 50 |
11 | 20 | 100 |
21 | 30 | 200 |
31 | 40 | 350 |
41 | 50 | 650 |
51 | 60 | 1150 |
61 | 150 | 2000 |
Uh, not very helpful...
It is. To put final nail in the coffin about "SPECIAL hidden bonuses" discussions like: Str def/atk, Int bonus exp and etc.
By the way, are you dissecting latest (1.15.15) version or pc version? There are quite a few small differences. Like twins shared SPECIAL progresion, for example.
PC Version. Besides the weird wasteland logic changes, 1.15 is just a few new quests (quest locations, and quest enemies), weapons, outfits and dwellers and retextured raiders (also that drone and the two themes). Alright then, here's all the SPECIAL uses in a single image. As I said, there really aren't that many. That said though this image contains the UI and unused uses of this function, so I've closed the scripts where they're the only thing. For example, here's an unused (and the only unused) use of the luck stat:
That makes sense then that a difficulty of 83 is near impossible. +25 from being level 50. Dragon's Maw with a +6 damage pet gives 34-42, putting you at 59-67. You'd need 24-16 more, which especially if you don't have 17 in the stat would require lucky rolls.
Oh, I just realized I didn't make the bit about the damage pets in the wasteland encounters clear enough - If they have a damage pet, the damage 2/4/6 damage is included to the min and maximum of their weapon. The pet rarity damage bonus is not included, see the image below: Kinda a shame too, because it'd make for an easy way to ensure that your normal mode explorers succeed in nearly all the legendary junk events they get. I guess I'll add it to my suggestions I wish to see acted upon.
Woah, I did not expect a new update. Too bad it's more of the same of 1.15 - new quests, weapons, outfits and dwellers, but no QOL changes/bugfixes.
The wasteland is unchanged. But, there's two new surprise quests! And they're not just 2-3 rooms, either... Edit: They're not marked as repeatable, so you can only complete them 5 times (one for each difficulty variant)
There's some new outfits and weapons: Outfits list: 164 -> 181 - 9 new outfits ingame? Weapons list: 187 -> 229 - 11 new weapons ingame? Perhaps there's more than a few new ones that are lunchbox only? Or, are used in the quests for NPCs?
And there's an extra 722 lines in the languages, though that's probably because of the quests and the mr handy lines being duplicated for the new mr handy.
Oh, and I've sorta figured out how to get ingame. Something's preventing buttons from registering clicks, so this was about as far as I could get. Besides that, and the shaders of course, things seem to be ok, other than a few one-time errors (and lots of warnings from the code).
So, maybe the PC version of shelter could continue to receive updates - unofficially.
You're doing amazing work here. I 100% will update the FAQ, just not sure when, gonna need to edit a lot of sections.
Oh and, https://github.com/AssetRipper/AssetRipper/issues/1287#issuecomment-2053971911
I left unity idle in play mode for a while (enough to make the game do it's time load process) and that seems to have fixed the issue???
https://github.com/AssetRipper/AssetRipper/assets/64717068/f2e01359-9147-4c43-a953-e7ed12bdc899
First some random thing I noticed. A legendary 6 damage pet does actually deal that whopping +12 damage - One of my explorers in a max diff food fight with one was one shotting every raider in the 3 wide room. Maybe the quest strategy is to have one on hand that you can swap around to multiply the effects of a big crit? Anyway...
Luck Critical Hit Meter Increase Luck Value | Increase Rate |
---|---|
1 | 1.2 |
2 | 1.26 |
3 | 1.32 |
4 | 1.39 |
5 | 1.46 |
6 | 1.53 |
7 | 1.61 |
8 | 1.69 |
9 | 1.77 |
10 | 1.86 |
11 | 1.95 |
12 | 2.05 |
13 | 2.15 |
14 | 2.26 |
15 | 2.37 |
16 | 2.49 |
17 | 2.61 |
The critical hit meter increases at the above constant rate every second: For example, at 17 luck your dweller gets crits 40% faster over 10, and 110% faster than 1 (55% faster at 10 over 1). Once the dweller's m_criticalHitMeter is higher than 100 (and targeting/fighting an enemy) you can trigger the minigame: Though I'll also add that after a 'successful' minigame (you didn't time it out), the dweller's m_criticalHitMeter gets randomised a bit: Failing the minigame sets the m_criticalHitMeter to 25 which sounds fine, until you realize that it also results in a damage factor of zero, aka, zero damage (so just don't let it timeout).
Perception Reduction Ratio List Perception Value | Speed Ratio |
---|---|
1 | 1 |
2 | 0.9683 |
3 | 0.9166 |
4 | 0.8749 |
5 | 0.8333 |
6 | 0.7916 |
7 | 0.7499 |
8 | 0.7082 |
9 | 0.6665 |
10 | 0.6248 |
11 | 0.5831 |
12 | 0.5414 |
13 | 0.4998 |
14 | 0.4581 |
15 | 0.4164 |
16 | 0.3747 |
17 | 0.333 |
Note: The critical hit limited/timed to 10 seconds - so using too high of a perception stat - like the max can make it harder to hit perfectly, in that it's jarring that it's half the speed of 10 perception and makes you wait longer for the next cycle.
Dweller Attack Rate Agility Value | Attack Rate | AR in seconds |
---|---|---|
1 | 0.08325 | 12.01201 |
2 | 0.09375 | 10.66666 |
3 | 0.104 | 9.61538 |
4 | 0.1145 | 8.73362 |
5 | 0.125 | 8 |
6 | 0.13525 | 7.39371 |
7 | 0.14575 | 6.86106 |
8 | 0.156 | 6.41025 |
9 | 0.16675 | 5.997 |
10 | 0.177 | 5.64971 |
11 | 0.1875 | 5.33333 |
12 | 0.198 | 5.0505 |
13 | 0.20825 | 4.80192 |
14 | 0.21875 | 4.57142 |
15 | 0.22925 | 4.36205 |
16 | 0.2395 | 4.17536 |
17 | 0.25 | 4 |
Unlike the other two stats, the attack rate tapers off quite quickly. This is how the AR in seconds is determined:
Also, the AR in seconds is just the trigger for the attack, not the attack itself:
For weapons with "long" animations, like the fat man, that's an extra ~2/3 seconds, hence way the missile launcher is just better than it despite the slightly lower damage - it shoots (and reloads) faster.
Also, this is just the dweller attack rate. Most enemies attack much faster, especially at high difficulties, but I can graph those curves later.
Caps | ||
---|---|---|
Chance | Min amount | Max amount |
0.5 | 300 | 500 |
Scaling Multiplier | ||
Min rooms | Max rooms | Multiplier |
0 | 10 | 1 |
11 | 20 | 1.5 |
21 | 30 | 2 |
31 | 40 | 2.5 |
41 | 50 | 3 |
51 | 60 | 3.5 |
61 | 70 | 4 |
71 | 10000 | 5 |
Nuka Quantum | ||
Chance | Min amount | Max amount |
0.4 | 5 | 15 |
Lunchboxes | ||
Chance | Min amount | Max amount |
0.05 | 1 | 1 |
Pet Carriers | ||
Chance | Min amount | Max amount |
0.05 | 1 | 1 |
Cooldown | Required Dwellers | |
480* | 13 |
*This would imply that it's an 8hr cooldown, but ingame it's now on a 2hr cooldown (previously 4hrs) in 1.16. Maybe the code was changed to use a 120/240 cooldown instead? A bit misleading to have an image of all four things (without drop %s), thereby implying that they each have an equal chance of dropping, when they don't. Still, 50% chance at not caps (with 10% of that being either a lunchbox or pet carrier), every two hours, per cloud connected device? I'll take it.
Confirmation on when incidents start occurring: It's after getting more dwellers than the minimum requirement for that incident. Though interestingly, the deathclaw chance can increase at the minimum. That would explain why one can trigger them so easily after going above 35/60.
Level Number | Minimum Experience | CAPs Reward* | Experience Difference | Experience Difference by Num. of Super Mutant Behemoths | Total Num. of Super Mutant Behemoths |
---|---|---|---|---|---|
1 | 0 | 2 | 0 | 0.0 | 0.0 |
2 | 3600 | 3 | 3600 | 0.9 | 0.9 |
3 | 16200 | 4 | 12600 | 3.2 | 4.1 |
4 | 28350 | 5 | 12150 | 3.0 | 7.1 |
5 | 40500 | 6 | 12150 | 3.0 | 10.1 |
6 | 55080 | 7 | 14580 | 3.6 | 13.8 |
7 | 69660 | 8 | 14580 | 3.6 | 17.4 |
8 | 84240 | 9 | 14580 | 3.6 | 21.1 |
9 | 98820 | 10 | 14580 | 3.6 | 24.7 |
10 | 113400 | 11 | 14580 | 3.6 | 28.4 |
11 | 139320 | 12 | 25920 | 6.5 | 34.8 |
12 | 165240 | 13 | 25920 | 6.5 | 41.3 |
13 | 191160 | 14 | 25920 | 6.5 | 47.8 |
14 | 217080 | 15 | 25920 | 6.5 | 54.3 |
15 | 243000 | 16 | 25920 | 6.5 | 60.8 |
16 | 267300 | 17 | 24300 | 6.1 | 66.8 |
17 | 291600 | 18 | 24300 | 6.1 | 72.9 |
18 | 315900 | 19 | 24300 | 6.1 | 79.0 |
19 | 340200 | 20 | 24300 | 6.1 | 85.1 |
20 | 364500 | 21 | 24300 | 6.1 | 91.1 |
21 | 388800 | 22 | 24300 | 6.1 | 97.2 |
22 | 413100 | 23 | 24300 | 6.1 | 103.3 |
23 | 437400 | 24 | 24300 | 6.1 | 109.4 |
24 | 461700 | 25 | 24300 | 6.1 | 115.4 |
25 | 486000 | 26 | 24300 | 6.1 | 121.5 |
26 | 534600 | 27 | 48600 | 12.2 | 133.7 |
27 | 583200 | 28 | 48600 | 12.2 | 145.8 |
28 | 631800 | 29 | 48600 | 12.2 | 158.0 |
29 | 680400 | 30 | 48600 | 12.2 | 170.1 |
30 | 729000 | 31 | 48600 | 12.2 | 182.3 |
31 | 810000 | 32 | 81000 | 20.3 | 202.5 |
32 | 891000 | 33 | 81000 | 20.3 | 222.8 |
33 | 972000 | 34 | 81000 | 20.3 | 243.0 |
34 | 1053000 | 35 | 81000 | 20.3 | 263.3 |
35 | 1134000 | 36 | 81000 | 20.3 | 283.5 |
36 | 1215000 | 37 | 81000 | 20.3 | 303.8 |
37 | 1296000 | 38 | 81000 | 20.3 | 324.0 |
38 | 1377000 | 39 | 81000 | 20.3 | 344.3 |
39 | 1458000 | 40 | 81000 | 20.3 | 364.5 |
40 | 1539000 | 41 | 81000 | 20.3 | 384.8 |
41 | 1684800 | 42 | 145800 | 36.5 | 421.2 |
42 | 1830600 | 43 | 145800 | 36.5 | 457.7 |
43 | 1976400 | 44 | 145800 | 36.5 | 494.1 |
44 | 2122200 | 45 | 145800 | 36.5 | 530.6 |
45 | 2268000 | 46 | 145800 | 36.5 | 567.0 |
46 | 2397600 | 47 | 129600 | 32.4 | 599.4 |
47 | 2527200 | 48 | 129600 | 32.4 | 631.8 |
48 | 2656800 | 49 | 129600 | 32.4 | 664.2 |
49 | 2786400 | 50 | 129600 | 32.4 | 696.6 |
50 | 2916000 | 50 | 129600 | 32.4 | 729.0 |
*In 1.16, leveling up in the vault rewards CAPs by a new unknown formula: 2 luck, 49->50 - 100? (before: 469017 after: 469117) 12 Luck, 49->50 - 11200 (before: 984371 after: 995571) That's a lotta behemoths.
Special | Minimum Experience* | CAPs Reward** |
---|---|---|
1 | 300 | 1 |
2 | 3600 | 2 |
3 | 14400 | 3 |
4 | 36000 | 4 |
5 | 72000 | 5 |
6 | 126000 | 6 |
7 | 201600 | 7 |
8 | 302400 | 8 |
9 | 432000 | 9 |
10 | 594000 | 10 |
*Divide by 100 for the time to level up in minutes **Result from the code
Each level to a training room decreases the time required by 5%, and each dweller is a 2% reduction. This means the maximum possible reduction you can get is 52% - 10% from level 3 room, +12% from full size 3 room, and +30% from an uninterrupted training pet.
So, this started by me wanting to put a number on the damage bonus that pets give and the damage mr handy does to incidents. And since Fallout Shelter is a unity game and those can be converted into broken unity projects with AssetRipper...
I found answers to both, but now I even more questions. For example, why they went through the trouble of making the script for wasteland dwellers a team (of dwellers) and even call the script WastelandTeam, and then hardcode it to only allow a single dweller on said team. Seriously, just look at the encounter success test function in the script (though after some further thought, since this adds up all the dwellers in the wasteland team, having enough multiple low level, low special, low weapon damage dwellers in the same team would basically guarantee that they pass every single encounter check, no matter how high it is, that and, even with a 25 team limit, most of the vault could be sent out to explore.).
Or, why room decorations were cut. Judging by the code, they were basically going to be room bonuses (you know, like pet bonuses, but for the rooms themselves). With the unused bonus effects seen here probably being some of the ones they used: Maybe they were too overpowered as they'd been gotten similar to legendary weapons (the full recipe icon you get for the party themes is actually the decoration recipe icon) - find the recipe and then simply craft it.
Since I've just kept getting carried away these past few days looking at this stuff, I'm going to stop myself. Maybe you could try looking through it.