Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - SubterminalOptimization

Pages: [1] 2
1
"everything can be measured and planned out"

COUGH* I'd like to remind you that opponents do criticals hits too.

You guys are very naughty with your temptations for a challenge run here.  I have a lot of work at the moment!  I specifically beat Dominating at Level 6 (and posted my run breakdown) because of some people complaining about Dominating being too difficult/impossible at Level 25.

Now you're going to claim that critical hits, bigger health pools, etc. make it too hard / inconsistent to do a Deathless (no saveload) Dominating? :p

I hope for Shredded Cheddar to respond...I think it's worth checking for an update from him (and maybe others will be interested in investigating this?).

http://underrail.com/forums/index.php?topic=3409.0

But for now I would remind you that opponents can only critical you when they have a chance to attack you.  If you are playing "optimally" I suspect this should actually be very, very rare in and of itself.

2
I always take Interloper on my stealth builds though though I do not think you can find or buy Tabis before you are done with Depot A.

Ah I think you're right other than random. I forgot that the Black Eels Quest line leading to Blaine (and his store/Blueprints) is actually right *after* this quest.  You definitely do not need to complete Depot A though or get Armadillo parts.  I guess you'd be limited to crummy +4 stealth +1 Agility boots or something like that in the general case.  I actually know for a fact that you can *possibly* get at least non-stealth +12% movement tabis or so from a mutie near the south entrance (I distinctly remember looting once) but that's likely random.

there are a lot of PSI beetles near the shrooms on the left side which you need to stand in to avoid the Bilocation proc

You do not need to visit each mushroom though.  The "middle" one on the West/left side is almost always either completely clear or you can move around up/down as needed to keep the stacks cleared.

I do not see a reliable way to get through it on a regular basis without reloading and/or without fighting.

It's definitely not 100% reliable to do without reloading, but I think it's something like ~80%, probably a bit higher with intensive practice.  You definitely have to very actively click in response to enemy movement, you cannot just click-click-click three times to three waypoints and be done.

In a nutshell "Autistic" as I call this difficulty is an explosion of effective HP and a significant increase of damage.

Even so, it can be completed at Level 6 with strategy and some patient reloading, so there's definitely a decent margin :).  I haven't tried a Hardcore/Iron Man/Deathless run as another poster is doing, but I strongly believe that such a run could be done reliably (probably in 90%+ range).

Of course this game does have flaws, like any other, but I am skeptical that Dominating should be something that can be done without saveloading unless you have both *extreme* planning and deep game knowledge.  It's a strange criticism to me because that is exactly what I expected from the mode based on its description (and based on the fact that the game has virtually no real-time "action"...everything can be measured and planned out).

3
Tchort Fight Strategy Summary

- Descend ladder in full stealth except for Gas Mask swapout when needed (Gas Mask is NOT optional).
- Place two TNT charges on *the left/West side* Mutagen Tanks and blow them.  Be careful to hug the left wall for the NorthWest tank path or Tchort will see you (I always do this one first).
- Place two TNT charges on the *right/East side* Mutagen Tanks and blow them - saving the NorthEast-most tank (the one on the right edge of the screen) for last.
- At this point, *wait for parade* then dodge back to the left side.  I say "wait for parade" because there can be 12 (12 is what I screenshotted and counted last run) Tchortling/Spitters/Devourers that all grouped up on the right side platform in a nice square ball to investigate the TNT.  My tests revealed this is 100% consistent behavior.  In addition, with good RNG, these Tchortlings will stay relatively-densely packed for enough time to kill the remaining Tentacles AND Tchort without ever straying far enough to the center to aggro you.  This is a foundational component of my strategy.

- I don't think it's 100% necessary, but I used combat-stealthing (enter combat, move, exit combat) to easily maneuver past the enemies to get to Western shack as quickly as possible and keep Tchort tremor damage low.  The amount of time you have to wait for openings through the "parade" varies with RNG.

- Trap laying is neither necessary nor relevant.  Just rely on RNG for the enemies to never stray toward the center and everything becomes easier and better.  In my experience this happens almost 20% of the time if not more.

- Mutagen Puzzle was Completed, so now Three tentacles to fight: the tentacle types remaining should not actually matter - if you execute strategy properly they will never attack. (My Exitus 1 "seed" was: CB V1 KU NX TM HG XC MH EV JC).  I've been stealthed 100% of the time until now so Tentacles have not left the Eye area until now.

- The left-side (West side) shack is crucial because you can go back there, exit combat and get a save snapshot, switch to full combat gear and then PICK OFF TENTACLES WHILE DODGING BEHIND SHACK DOOR after each burst.  The Tentacles appear in a perfect formation such that none of them can hit you inside shack but all can easily be hit from right outside shack door.  The Tchort-adds on the other side of the map DO NOT aggro or come over (I think they are too far away for sound).
- I killed all three tentacles this way (plenty of focus stims for all these separate shots).  As mentioned, even if several squares away you can easily deal hundreds of damage per burst at Level 6.  After killing all 3 Tentacles, exit combat and get a save/snapshot in.

- with Regenerative Overcoat and optimized movement my health barely ticked down. Tremors were still at the 26-30 damage range.

- Key moment: This is where we rely on RNG and not silly Tchortling Spitter :p.  Walk towards Eye of Tchort and you will hopefully see NO Tchortlings at all.  It's that simple.  Just a completely empty open path right to the Eye.  If your RNG is good, you can enter combat on the walkway, run right up to the Eye and unload nonstop Focus-Stimmed-bursts for 7-10+ turns (based on Tentacle respawn and how long you can survive that particular tentacle focusing you with all buffs up).  I had enormous margin here and probaby with only a little bit more optimization could have dealt closer to 5000 damage rather than the 3000 HP needed to kill Tchort (based on my turn margin).

- Even after Tchort is dead, if you're lucky, the Tchortlings will be nowhere near you so you can literally immediately save, stretch, relax, fully rebuff, and then stroll to the Ladder (and if unlucky, maybe pick off one stray minion...although even with madness debuff they won't fight each other at all if they were never in combat so there may be a lot of "survivors").


Tchort Fight Notes

- I did Mutagen Puzzle because tentacles are very hard to kill now and my HP is just too low (and my current fight approach requires tentacles completely out of the fight for a while).  I can no longer reliably one-burst tentacles.  It is important to kill them as quickly as possible but there is absolutely no way to fight them directly (without shack dodging).
- Fighting Tchortling adds is unnecessary and extremely dangerous.  You cannot reliably win a 1:1 fight vs. any type and in fact will probably lose.



Arke IRIS Fight

- Definitely want to clean up the initial bots before talking to Arke and de-power all four power switches in the room.

- This fight is a lot harder at Level 6. I went all out with 12 EMP mines and 15 bear traps, alternating layers of Bear Traps and EMP Mines so that the bots would be constantly stunned.  I had to leave a sneaky path for me to get back inside the bottom left console area (my holdout) of course.

- When the south entrance bots initially spawn they will be in a group and you get a good hit/free bursts there, so there should always be an EMP mine there

- I used W2C bullets mandatory for almost *everything* and not just Industrial Bots, unless I KNEW I could get away with a non-W2C burst for a weak/stunned enemy.

- I actually did need both guns, my first gun got yellow during the fight from how much bursting was needed.

- The four center turrets are easy to clean up at the end but I was also able to exit combat and get a save in before killing them
- Focus stims were definitely massively helpful here.

- It was not needed for this run, but I observed that it is possible to semi-reliably get the bots into a "stuck" pattern in front of the narrow door where two bots are both trying to get in and end up walling off all other enemies so you can get almost unlimited free hits if you aim past them.  This could be abused even at Level 1 to potentially complete the fight, though I'm not sure if this would count as a bug/exploit (it's definitely behavior I've seen elsewhere).


Arke Basement

- I no longer reliably one-burst Greater Coil Spiders.  This is a pain.  I need to flashbang and molotov much more now.
- I killed the initial scorpions with bear traps and killed the spiders guarding the first key.  I then planted one TNT and lured almost ALL spiders out of big electricity room and into hallway. I was still able to sneak by without fighting (again except for earlier, the first small room with two spiders for door key to open the big electricity room).
- Similar to Caerus Residential, there are two key strategies for Electro Spider killing:
  Need to kill one spider and you are near door?  Burst spider then close door.
  Need to kill two spiders (Wasi Abdul keycard, Arke basement key)?  Burst one spider, adrenaline, flashbang, burst second spider.  Update for Level 6 run: molotov and run back to a door multiple times.
  Need to kill three spiders?  You're doing something wrong, there is never a time this is needed.  However, molotov and door-running theoretically scales to any number of spiders since they do not heal. 
- You can always easily just go right to the ladder in the last room of Arke Basement, saveload if the scorpions bump you.  You can obviously trap if you really need to but it's not even necessary.


Mushroom Forest

- Even with CAU Armor and Biohazard boots I take damage here.  CAU Armor really seems to be mandatory.  Add in Irongut+Aegis though and only the chompers can do meaningful damage.
- The room before the final room is still a joke with some simple Adrenal+Jumping Bean + Irongut + running to the transition.

- New with Level 6: the last room is no longer so simple.  I have to do partial kiting instead of waking everything at once.  Now I kill the left-most mushroom sentry using hit and run even though it will respawn very quickly.  I need to do this so that I can safely inch forward and aggro ONLY the first spitter/chomper pair and then kite them back (so I only have to fight two enemies at once).

- Now I can inch further forward north and aggro the next one and the next one (ignore the sentries for now).  Keep chugging irongut and thinning out the non-respawning enemies.

- After the non-respawners are gone, mostly return to old strategy.

- Taking the "northern" path to the heart and hiding behind the rock or whatever allows me to focus shooting the heart from relative safety.
- I don't bother killing the sentry, because with enough buffs + energy shield+ biohazard boots+cau armor it doesn't do enough damage plus killing it would take so much ammo and turns that it would not actually give me more effective damage time on the heart itself (before my gun durability or HP run out).

- I am glad I crafted a bunch of Iron Gut but with Biohazard boots and very good RNG I can still imagine a world where sub-Level 6 is possible without it.


Tchort Institute
- mostly the same as Level 10
- Again I sneak into Eidein's office, kill his guards to trigger "tremors", go down elevator, go BACK UP elevator, and he is there.
- A very nice side effect for this run in particular is that I am able to place multiple layers of Bear Traps around where Eidein WILL be but before he's actually there, so there's zero chance for problems in that fight (I basically can't hit Tchortists unless they are immobilized and even then it's much more likely that I'll need multiple bursts to kill them).
- I just assassinate Eidein and go back down and done with Institute (within minutes of joining the Institute if I wanted to, this time around I had to do a lot of shopping and shop refreshes).
- I do not know if this is a "bug" but it is confusing that some zones "autohostile" you even when stealthed (rest of the Institute) and others don't (elevator area) even when they both have an alert "Eye" to indicate "Controlled Zone"
- This time I carried 229 Weight Units of stuff (Lightly Encumbered) into Deep Caverns with me and I used a lot more of what I brought this time.

Lunatic Mall (Jumping Bean Mall)

-This Time: I just use Jumping Bean+Adrenal+Cloaker and tight clicking to get past first floor (no killing).

-I use Toxic Gas Grenades to kill the 2 guys on the left-hand-side, then a couple traps + burst on the patrolling guy.
-The person inside the bottom left store I just used one bear trap to keep out of my line of sight, didn't need to kill but timing was annoying.
  I can still use Adrenaline + Jumping Bean + my stealth gear.  Those buffs all stack and it's still insane.  Jumping Bean + Adrenaline + 22% boots + Rathound Regalia + Interloper.
 
  I did get barely-orange eyes on the guys around the statue table but I moved so fast that I easily outran all of the investigation movement and they turned back to yellow before I even got halfway toward the blue floor transition.

 
First Protectorate Epione Lab Quest to kill Upper Underrail Free Drone Base with Shredder

- You can toss several grenades at the front door like a doorbell, so actually you don't need to do the riskier back-door area with no retreat space.
- Toxic Gas Grenades + Molatovs + Bear Traps were all prepped outside the door ahead of time.  I needed 1 save reload. I literally killed them all more easily this time at Level 6 than my first Dominating run at level ~20?.   Insane.
 
 
JKK Questline

-The JKK Quest Line is WAY easier with stealth, although the last quest with the Spiders was annoying. 
- For the last JKK quest, the Ghoul/Mutants are also a real pain to fight or sneak by without aggroing the whole floor so I painstakingly gas grenaded the southern room with 3 mutants without entering combat.  I had to use a different throwing angle and separate grenades for each one I gassed.  But it was definitely rewarding in the end.
-For the last JKK quest, I used molotov + running to burn that stupid patrolling spider in the utility room down over the course of three throw molotov/run/door close cycles.  This was especially necessary at Level 6.
-For the last JKK quest, for the patrolling bots, a couple EMP mines and/or bear traps makes the job a lot easier.
-For the last JKK quest, I lured the Industrial Bot into the neighboring room from the pump controller and then closed the doors on him because, again, screw him.

4
Hey guys, I'm back with the first half of what will likely be my final update (it's taking me longer to write than expected).  As a result, I'm going to try to add a lot of information here in case anybody ever wants to try to push further (which I  believe is possible).  Perhaps that will be me in the expansion, but not for now I think.  Anyway, inspired by the discussion here, I performed a lot of Assault Rifle-based testing and obtained a lot of data since my previous post on this thread.  I've completed my most recent run fairly quickly, largely as a result of the unexpectedly promising testing data.

--------
TESTING
--------
As already stated, I believe the Tchort Fight is the sole gatekeeper of the run.  There are other difficult areas but I think they are strictly easier than the Tchort Fight in terms of virtually every requirement: raw health, raw damage, stealthing/sustained survivability without saveload checkpoints.

I mainly tested by repeatedly reloading my reference save files at Tchort Fight, using Cheat Engine to modify my skill numbers, and attempting to complete the fight.  This is very easy to do if you have a Level-up pending, as you can freely manipulate the value in the Skills screen up or down and iteratively search for those exact numbers in memory.  You can set the numbers to anything you want (you can have "impossible" totals that don't sum up to what you should be able to have at your Level).

GUNS
Through my testing I was able to confirm that, among other things, it is possible to kill Tchort Tentacles with far less gun skill than I thought.  I only tested down to the range of Guns 55-59 because I initially believed that that would be the lowest relevant skill level, but at this point I'm not so sure.  In any event, my testing revealed that this range is/was sufficient to kill tentacles in a "reasonable" amount of time with comparable-skill-level crafted guns.  I am defining "reasonable" as fast enough to kill 3 tentacles (with tanks blown + mutagen puzzle) and have them all simultaneously dead for a minimum of 7 additional turns (it seems that Tchort Tentacles take 20 turns to respawn!!).  Dealing 600-700 damage (maybe more) against tentacles in a single burst (with various relevant buffs) is no longer probable but it is still absolutely possible and moreover that is actually far more damage than what is needed.  You do not even need to be point blank range for this much damage (I was several squares away) to achieve this.

HEALTH

I now believe that TheAverageGortsby is correct that a Regenerative Vest with Black Cloth is the optimal chest armor for lowest-possible-level Tchort Fight.  Even at extremely low-skill crafting there is almost no downside because your Maximum HP will be so low that the absolute healing number is still a huge proportion of your health pool...the most relevant part is that you receive sustained healing (10 increments) without having to break stealth *and* those incremenets are guaranteed to not be wasted (since they only trigger belopw 70% HP).

I didn't thoroughly test whether low levels will run into trouble with respect to stealth instant-detection-threshold (since they can only make low-stealth Regenerative Vests) but I am fairly sure that Rathound Regalia will not be sufficient for characters with ~90 HP or lower and I am not sure low levels can afford to put points into Constitution.

STEALTH
I confirmed that with Stealth in the range of ~120 (not counting cloaking device) you can still pretty easily stay undiscovered during the entire tank-blowing-phase of the Tchort Fight and that is really not a high number at all considering that gear can trivially provide more than half of that number.  At 120 stealth though, I think a cloaker is necessary to avoid heading towards extreme RNG (I didn't bother to wait for perfect RNG) and, as mentioned, I neglected to verify the exact instant-detection-threshold so that is something to test for lower level runs.

TACTICS
Although Cheat Engine was extremely helpful to confirm certain projected skill numbers, I also found it equally helpful to load and reload my unmodified Level 10 Tchort Fight save files and complete the fight multiple times.  What I found is that not only was my Level 10 Tchort Fight strategy repeatable but it can actually be further optimized and leverage saveloads for relative-easy-to-obtain RNG that is extremely beneficial.  Crucially, it is possible to get RNG such that the two following things are true:

1. *ALL* Tchortlings travel to the North-East-most Mutagen Tank (if it is the last one blown up) and leave the West side of the map completely clear for Tentacle killing by using the shack (and are not close enough to aggro even from 9 round AR bursts)
2. *ALL* Tchortlings *remain* grouped in that North-East-most Mutagen Tank even *after* killing the three Tentacles, resulting in the situation that you can literally (in combat) walk right up to Tchort and take free shots on him without any interference for 7+ turns (until Tentacles respawn).


-----
Level 6 Run
-----

The target level for this run was largely based on the testing data I had available.  Through testing I determined that Level 6 could obtain Gun Skill 60 without sacrificing access to Chem/Bio Skill 60 or Interloper.  Interloper Feat+optimal gear allows for a successful mutagen-tank-blowing phase of the Tchort fight and I also knew from testing that Gun Skill 60 could *easily* provide enough damage per second to complete the Tentacle/Eye Killing phase of the Tchort Fight.  Levels below 6 lose access to Full Auto (a massive 22% raw damage increase Feat) and I was concerned that I couldn't properly/fully test a Level 5 scenario (I don't know how to remove Feats from a Save File) even though I observed that 55 Gun Skill alone shouldn't be a showstopper (though harder).

From previous runs, I knew that so long as stealthing the mutagen tanks and killing Tentacles in a few turns each is possible, the entire rest of the run should be possible because everything combat-related is essentially strictly easier than these two tasks.

I have now completed Dominating at Level 6 using:

Str: 7
Dex: 3
Agi: 7
Con: 3
Per: 10
Will: 3
Int: 8

Feats:
Echoing Soliloquy (No reason not to get this right?  However, it should be noted that Eye of Tchort stacks extremely fast that there's not much difference...it seems to stack even faster at lower Character levels but I'm not certain)
Gun Nut (It's very possible, though I'm not 100% certain, that Recklessness is better/easier and should be in this slot.  I thought Gun Nut was the least risky choice assuming perfect RNG.)
Opportunist
Suppressive Fire
Interloper
Full-Auto

Skills:
Guns 40 (60)
Stealth 40 (122 with just highest-stealth gear and skills, with cloaking device effective 162...For this run I did not hesitate to use Jumping Bean+Adrenaline for all the "challenging" parts...for all stealthing I used Rathound Regalia because 10% movement speed +25 stealth with Interloper...for Tchort Fight I only switchen to Regenerative Vest for the Killing Tentacle/Eye Phase)
Traps 40 (39) (I used 6 EMP traps during Arke Fight and a couple crawler poison traps for Tchort Fight, but never got too fancy here, used plain old bear traps virtually everywhere else)
Mechanics 40 (53) (60 with House)
Electronics 40 (53) (60 with House)
Chemistry 40 (53) (60 with House)
Biology 40 (53) (60 with House)
Tailoring 40 (53) (60 with House)

I crafted (with House):
- 9mm Smart Muzzled Hornet 14-45, +32% special attacks + second copy with +30% special attack (2 guns are necessary for Arke and Tchort Fight to avoid needing to use a damaged weapon or repairing midfight somehow)...I also absolutely could not find better smart modules (skill required 45 was the best without going over 60, even though I waited 10 shop cycles before taking the elevator and checking EVERY shop)
- Efficient High Frequency Shield Emitter 597 Capacity, Conversion Rate 8.9, 10% dissipation, 5/16/66/98/126
- Regenerative Overcoat 24% / 5 Mechanical, 15% Armor Penalty, Stealth Increased by 22, 29 Health for 5 Energy every turn when below 70% HP, 50 Energy capacity

Noncrafted:

- Biohazard Boots.  I hadn't used these before but they make the final room of Mushroom Forest a lot easier since you don't lose perception or ability to do full bursts while wearing them yet they fully stack with CAU armor.

- Purchased from Oculus store:
 
 - +13% Critical Hit Chance Seeker Night Vision goggles (87 energy capacity, turning on costs 9, consumes 3 energy every 5 seconds), I mostly used these since I think they are better but got the second pair below just in case.
 
 - +27% Special Attacks Smart Night Vision goggles (79 energy capacity, turning on costs 9, consumes 3 energy every 5 seconds)
 
 - +22% movement speed/30 stealth Ninja Tabis.  I probably could buy better if extremely patient or get better by killing all CoreTech Runners but I was pretty confident based on my tests that these would be "good enough".  I wore these Ninja Tabis for the entire Tchort fight and I'm not sure that anything crafted would be better.  The stealth is critical for mutagen-tank-blowing Phase and the movement points are critical for the Killing Tentacle/Eye Phase.
 
 
-Quest Reward from JKK:
 +42 stealth cloaking device, 76 energy, turn on costs 21, consume 7.14 every 5 seconds

- Rathound Regalia (I don't believe it is possible to craft anything better than this for stealthing purposes not counting Regen Vest for Tchort Fight)
- Chemical Assault Unit Armor (I just kill the CAU members outside of the Free Drones base when it's "Go Time" and I don't care about burning Protectorate bridge)
- Standard +17 stealth balaclava (I could not craft better and didn't get the option to buy better)

- Bullet Belt (of course)
- W2C bullets for all the "major" fights in Deep Caverns (You need a fair amount at Level 6.  I took the Faceless +300 and I also brought 350 W2C rounds with me from buying nonstop through the run, I finished Deep Caverns with 200 rounds left so I used ~450 across Mushroom Forest, Arke, Tchort Fight)

This run I used ChemicalAssaultUnit Armor only for MushroomForest.  RathoundRegalia was used for early Tchort fight and almost everywhere else except early (before Core City main quests).  I used Regenerative Overcoat in only one place: the post-mutagen-tanks Tentacle/Eye Kill Phase (you can switch armor inside the shack after first exiting stealth but before fighting tentacles).

Final experience bar (before boarding Train to North Underrail): 185147 / 6000

Notes of interest:

Overall

- I needed to save and load a lot more, even compared to Level 10, but I still didn't need to reload more than ~20-25 times at any given single place.  There were a lot more places that I needed to reload ~10+ times though.  Also, I skipped Carnifex, so these reloads were in mandatory/required spots, such as Hollow Earth TH44 manual stealthing or JKK Final Mission. I completed the Tchort Fight in under 30 minutes, with ~20 reloads and 6 total saves.  I actually reloaded more on the first room of the JKK Final Mission than the Tchort Fight.  Since I had already broken the Tchort Fight down to a science and a reproducible solution during all of my testing, I was pleased but not surprised that this was, relatively speaking, smooth and easy.

- I took significantly more time in Deep Caverns and on Tchort compared to my Level 14 and Level 16 runs but was not too much slower than my Level 10 run.  The gatekeepers (Mushroom Forest, Arke Fight, Tchort Fight) are unchanged.  All of Deep Caverns combined took on the order of ~6 hours.  The Labyrinth is a joke and absurdly easy with Interloper even at Level 6 (and with practice dodging those linebacker worms trying to tackle you).  I only needed to reload once during the *entire* drive shaft segment from starting at Caerus back to Caerus.

- The central theme to the *majority* of this run (everything outside of Mushroom Forest and Tchort) is again: Toxic Gas Grenades, Flashbangs, Molotov Cocktails, Bear Traps, and EMP Mines.  If you can't Bear Trap enemies, flashbang them and/or EMP them as appropriate.  If you can't flashbang or EMP them, burn them with molotovs.  If you can't fight them at all, use precious Toxic Gas Grenades for a guaranteed kill. 

- Toxic Gas Grenades should have their wiki entry updated to reflect the fact that they are in fact the single most broken single weapon in the game.  You can use them to kill virtually any biological enemy in the game regardless of your level or skills, full stop.  I think this was what was alluded to by destroyor when he said:

Emporiom could probably be cheese using toxic gas grenades.

But I did not really fully appreciate/understand that until now.  Not only can I guarantee that Lunatic Mall can be cleared using toxic gas grenades, I can guarantee that literally all non-immune biological/humanoid enemies in the game can be killed by a Level 1 character on Dominating using these grenades.  These grenades, so long as you have enough of them and no time pressures (I was buying Toxic Sludge like candy from every vendor from SGS to Blaine to Junkyard to University), will allow you to kill enemies without entering their line of sight or initiating combat.  They work around corners, and just keep stacking damage regardless of whether enemies have 2000+ HP.  The actual mechanism is very simple: continuously throw toxic gas grenades near your enemies (within 1 or 2 squares) but stay out of line of sight or quickly move away after throwing.  However, I say that the wiki should be updated because the *implications* of this are absurd.

For example, just as a reference, I killed virtually every single member of Jacek's 1000+ HP gang, including Jacek, except the weak Psi guy in the back, while surgically avoiding damaging Hugo for even a single HP (he only has 90).  Except for the last guy, I never entered combat and never fired a round.

After flashbang-stealthing, I also used Toxic Gas Grenades to kill Pulverizer in Arena without re-entering combat or using any traps.  ~2200 HP?  No problem.

Although I didn't test it and am not 100% certain, I'm very confident that (flashbang-stealth step aside) killing Carnifex would be trivial at Level 1 with this strategy.


- Based on this run, and in particular the Tchort Fight experience, I believe I was wrong to say a Level 4 run is not possible.  I still believe that under Level 4 is likely impossible and I believe that Level 4 and Level 5 may not be possible but are *likely* possible as a run moves closer toward Tool-Assisted-level perfect RNG and sticking with an AR-style build.

- Stealth is insane, but stealth with stealth +movement tabis, rathound regalia, plentiful Adrenaline, and also plentiful Jumping Bean is so good that even at Lvl 6 I'm absolutely stunned at effectiveness.  I cannot believe how effective it is even with a weak cloaking device and only 7 Agility.  I didn't even bother with Agility Steak this time around at Lunatic Mall.

- I didn't even craft 100% optimal gear, especially for the Smart Modules I used, and I still didn't feel much gear-related pain.  I possibly got lucky with some of my shop purchases, but that is definitely one of the reasons why I emphasize lower level runs start requiring more time investment and RNG.  As crafting becomes weaker and weaker, the run depends more and more on finding perfect store-bought items for certain slots: balaclava, cloaking device, tabis, and goggles

- I knew I wasn't going for SuperSteel this run so I didn't have to worry about saving Charons so money wasn't an issue at all (as long as I made sure to sell all the guns that I did find).


Tchort Fight *and* Mushroom Forest *and* Arke Fight Consumables

- This run I made a ton more consumables and used them generously.  I rarely waited for a buff to run out before refreshing it (it's rarely optimal Action Point-wise).

- Focus Stims - Crafted 12.  Didn't need to do the old killing Coretech employees routine.  Used 3-4 for each of Arke Fight and Tchort Fight.
- Regenerative Mixture - I still don't know how to get regenerative mixture except by killing Caerus Residential Vents guy "Gunter Vasilica" so I still did that. I only used one and it probably wasn't needed, for the Tchort Fight Phase 2.
- Aegis - I came prepared with a crafted stack of 9.  Perhaps counterintuitively these didn't seem to be make much difference except slightly in Mushroom Forest final room.
- Iron Gut - I came to Deep Caverns with 12 and chugged these nonstop in Mushroom Forest final room encounters.  I ended the run with 3 left.
- Jumping Bean - I crafted 12 and chugged them all the time to make stealthing or move/shooting/ducking easier.  I used 6 of these.  These things are amazing!
- Super Health Hypos - Since I'm friendly with Faceless I have 100% reliable Super Health Hypos to pair with Morphine (although I guess you only could ever really use 1?)
- Morphine - it's really only relevant for final hits on Tchort Eye if a tentacle comes back or if you're unlucky but want to keep going in the Mushroom Forest final room.  The reason is the same for why Aegis isn't so good: the vast majority of tactics should be aimed at completely avoiding ever being in a position where an enemy can damage you.
- For Tchort Fight only I saveloaded for +2 Constitution -2 Dex (or something useless) Junkyard Surprise (this is definitely worth doing instead of just accepting a +1 constitution burger).

- Adrenaline Shot - I didn't end up using these vs. Tchort because with my current strategy there is no point at which it is optimal to wait out the debuff time...you want to burst every single possible turn and there should never be a situation where a few extra AP make the difference between survival/success or not.  I did generously use Adrenaline in Mushroom Forest this time around because I strategically segmented the last room into several fights instead of one continuous (and extremely dangerous) fight.


5
General / Re: Crawlers / Death Stalkers
« on: March 24, 2018, 07:41:11 pm »
I'm surprised it hasn't been mentioned here but if you *know* there are Death Stalkers ahead, setting several bear traps to make a wall (outside of combat) and waiting for them to come to you (or throwing a grenade or using TNT to attract them) is basically a 100% autowin.

This may be considered cheesy but it's definitely not an exploit or bug or anything like that.

In all of my runs I have always used bear traps unless I absolutely had to use a flashbang or something else "dynamic" (but that's very unusual).

Also, if you have the Intimidating "Yell" feat that is almost literally a sonar for scorpions and makes it probably 5x easier to find/kill them consistently.

6
General / Re: Iron Man / Permadeath runs on Dominating difficulty
« on: March 23, 2018, 08:44:47 pm »
Doppelgangers in Tithonus Lab are not that dangerous, they reset when you leave the area. If one of them spots you, just run to the exit and try again.

Ah, didn't know that.  That's unusual (virtually all other screens hostility does not "reset" when you transition off of it).  I'm also used to dying to them in one turn at low level :).  I guess you can still get in trouble if they lock you in/surround you, but it does sound as though Tithonus would generally not be a problem then.


7
General / Re: Iron Man / Permadeath runs on Dominating difficulty
« on: March 23, 2018, 08:35:06 am »
Hello SubterminalOptimization. I like your name very much.

Thanks, I like cheddar so we are even!

Quick tinkering actually does help with doppelgangers, when they step on a bear trap they lose all movement/action points for the turn (unless this has been changed, I haven't used quick tinkering in forever since I think it's OP).

Interesting.  I just did some lazy tests right now using save files I had on hand.  Tithonus Lab Ghosts are definitely immune to immobilization/traps.  Doppelgangers on Silent Isle can spawn, walk over a bear trap (also says Immune), and still hit you in the same turn.  I don't know if this is applicable for Quick Tinkering but I will extrapolate for the moment that it is.

I realize this point is moot from your perspective (you are absolutely free to run with whatever parameters you wish and find fun) but I think it's interesting because I don't generally like to add too many constraints, especially if they are not "necessary."  What I mean by that is that I'm still not convinced that the truest most opimal build for no-death-run success would actually want to invest the Base Ability Points, Skill Points, and Feat Slot all just for this one ability.  It's possible, but my thinking is still focused on the idea that you need to optimize for the highest risk situations and not necessarily for something that would be all-around-helpful but not address the toughest points of the run.

To answer about aimed shot I want the option to have a sniper rifle in slot and delete a target before the battle starts. Since I am on ironman both planning and flexibility are very important

I see.  Again, it only makes sense for you to play the way that is most enjoyable, so I certainly wouldn't raise any objection.  However, from my perspective I don't think that is optimal unless you specifically know that you need it in order to complete the run.  And if I were to do such a run (which I may very well end up doing) I would want to plan out every quest and kill ahead of time to make sure nothing is unnecessary or lower probability than it has to be.  Perhaps there is value in sniping for Mushroom Forest or even Tchort Fight, I'm not certain, but my initial thinking is that surely there must be a better survivability or damage feat that you could use ~100% of the time rather than occassionally.

As you've already pointed out, many gaps can be bridged through the use of crafting.

Yes that plan makes perfect sense and I strongly suspect it is fully optimal.

While I want my playthrough to be as certain as possible that does not mean I am going to avoid fights I can lose. In my opinion that is part of the fun of it

Ah  :).  We definitely have different approaches to fun here which is fine.  In my iteration of Iron Man (if I do it), I would definitely prefer to optimize to get as near to 100% success rate as possible, ideally doing absolutely nothing that is optional.  My perspective is that of a "puzzle solution" rather than an action game, so I think that is why I have a different outlook.

Guns are most likely not optimal since psi is OP and can't miss,

Yes, that sounds like it might be true but humorously I have still never actually played a Psi Character, let alone a fully optimized Level 25 Psi, despite planning to possibly do so and ending up doing an AR run multiple times now.  It's definitely a gap in my knowledge about this game but I am similarly trying to optimize my focus on what is necessary for my current run (sub-Level-10) before learning anything else.

I found your idea for a challenge run to be interesting as well, to minimize the requirements to complete the game. How did you get this idea? Was it from speedrunning? Do you plan on attempting a speedrun at some point?

Although I am a fan of speedruns (not necessarily the grinding but more the routing and strategies) I don't currently have any plans for an Underrail speedrun (though maybe one day I would route for it...that part interests me but I would never have the time to do repeated 6+ hour runs back to back to back).

What actually inspired me for the runs that I am doing would be a combination of several factors (probably no single factor would be enough):
 - I have a huge interest in puzzle games but usually they are too easy or too dishonest (rely on hidden information and brute force) but this challenge (and Iron Man) would be honest custom puzzles.  I did something similar for the classic Fallout games (but for those I already knew speedrun-routing techniques found by others).
 - My wife complained Underrail was "impossible" on Normal (she has beaten the game on Normal now)
 - Although I did die a fair amount, I was a little disappointed by how easy Dominating was
 - I saw a lot of forum complaints on Steam and here saying that Dominating and/or Deep Caverns was too hard
 - So I wanted to post/tell everybody "that's silly, there is a huge margin and that margin is: X Levels" so I figured I better test and find out the value for X

I just want to note that fights that are hard for some builds aren't hard for others.  My sneaky trap-laying crossbow sniper on hard did the mall on the first try, and barely had to actually fight anybody, whereas RH King for example took me many tries. 

I definitely agree with this to an extent, and it makes sense for a generally well-designed balanced game.  However, a perfectly-balanced or even near-perfectly-balanced game is ***extremely*** rare and difficult to achieve.  I don't blame game developers for this because it is just a fact of math and variety.  Specifically for Underrail, which I do consider a well-designed game, I strongly suspect that at almost any given Level there is a single most-optimal build for purposes of a no-death-run.  Similarly, I am certain that only a handful of builds can possibly complete a Dominating run at Level 10 but I'm sure many more variations can succeed at Level 16 or 25.

8
We need a dev to confirm this, but I don't think ThermoD can crit.

Oh I see! If ThermoD cannot crit then I think combining this fact with your earlier data is pretty definitive.  It sounds like AR may actually be a more fruitful investigative route for Level 8 after all.

Re: Muzzled vs. compensated - sorry I'm wrong here as I wasn't really thinking about your specific case. With your effective gun skill so low I do agree w/ you compensated wins for your case.

Ah okay, no problem, understood.  To be clear, you may also absolutely still be right.  It depends on what the actual hit chance turns out to be.  If hit chance is high enough, going for the max damage would still be worth it (because achieving certain max damage breakpoints such as 1 shotting or 2 shotting enemies would be worth the lower consistency and doing some saveloading).   I was simply providing a hypothesis that could justify compensator being better but I am not 100% certain that it is a valid hypothesis for Level 8, for example.

I'm wondering if I have your permission to add this tip and use your image in my Steam guide?

Absolutely you have my permission.  However, you should know that the tentacle wall I showed in my previous screenshot was an amusing/unusual one that I just went ahead and used.  However, for my data gathering (where I had to repeatedly kill the Eye at Level 10 from the same mid-map save) I confirmed that by far the more consistent/reliable positioning is the one where you only need a single square to be blocked.  You actually want to stand right next to the tentacle-hole (even if you have more movement points) and specifically wait there for the sowers to shoot at you.  The outcome will look like this:




9
At 3 WIL level 8 you are only going to have 44 effective Meta skill, meaning ThermD will do 94% Health damage, so with tentacle you will get 940 raw damage, or (217 Heat + 152 Mech) 369 damage after resists + evasion.

Thanks for those specifics, very useful.  It sounds like low Will is not possible then, and therefore any ThermoD build will have to involve low CON.  If ThermoD can Critically hit though, it sounds like it may still be enough damage in combination with high Will? 

Muzzled will always wins out due to the fact the two extra bullets can crit.

Okay whoa, Nelly!  Not sure if I follow you here but it actually seems like you're making a very significant claim.  Are you confirming that critical hit chance provides *additional* raw hit chance above and beyond normal hit chance?  In other words, are you implying that if I have a 10% hit chance and 0% critical hit chance and then equip +10% critical chance goggles that I now have >10% hit chance?  If that is the case then critical hit chance is super overpowered (and focus stims are even more overpowered than I thought). 

However, I don't think this is actually true.  For example, the following situation is highly repeatable for me.



Would I really miss 9 shots so frequently if 27% crit chance meant 27% hit chance minimum?  Maybe you are implying something else, but I'm not certain.  My point (which is why I am surprised by your point) is that 2 additional bullets are almost worthless compared to 15% additional hit chance for the first 7, especially if hitting is required *before* critical chance can apply.

So for example, what is better (and keep in mind at Level 9 and below there will be no concentrated fire stacking to complicate the math)?

Barrel Compensator: 7 bullets with 25% hit chance at X% critical chance (shouldn't matter what the critical chance is) = 1.75 bullets hit on average
Muzzle Brake: 9 bullets with 10% hit chance at X% critical chance = 0.9 bullets hit on average

This math may not accurately reflect the game engine but IF IT DOES, then I believe this trivially indicates that it's better to shoot 7 rounds with 25% hit chance or in other words, Barrel Compensator would be much better.

Focus stim does not help you in mushroom forest as hostiles there are immune to crit.

Ah okay, too bad and also...what a strange developer decision.  Anyway, even though I was whining a bit I think that's still not actually a showstopping problem.  The fact remains that there are trap possibilities, molatov possibilities, and just raw saveloading for good RNG and hits.  For example, I was inspired to collect some data, so I loaded back up my last-stage-of-Tchort-Fight Level 10 save file and repeatedly killed Tchort (pretty consistent actually, ~60% win rate from this save where I am mid-map and Tchort is full HP but all tentacles just killed).

I didn't count killing blows because those don't give me the full damage readout, but the first 2 shots per load I could quickly see how much damage and many hits based on concentrated fire stacks.  I collected all data at POINT BLANK RANGE DIRECTLY IN FRONT OF THE EYE while having some form of TchortSower Tentacle wall protecting me from adds (see screenshot for the situation right before I walk up to Eye and start shooting, I actually did have Focused for these shots even though the screenshot implies I would move and fire next turn).



TrialTotal Burst DamageTotal Hits
113838
211428
38358
411248
515269
69118
710008
88187
915609
1011549
118427
128328
1310066
Average1087.153846153857.92307692307692

Tchort doesn't seem to regen even after 4 turns and the tentacles don't seem to come back even after 6+ turns...does anybody know/have a reference for what those exact numbers are by the way (for mutagen puzzle + tanks blown)?

I suspect that it's not too much to ask Level 8 AR to output half of these damage numbers and if that happens, it means Level 8 should be absolutely doable.  The problem though is that the I am assuming that damage is halving because of loss of Concentrated Fire but that I can still hit with mostly the same weapon and buffs.  If there is a massive accuracy cliff where Level 8 can't even land a shot, then projecting the damage loss from gear/skill multipliers and calculating extra turns needed for net damage don't really matter at all.

So the next step I think is testing AR Lvl8 and ThermoD Lvl8 (high Will) to get exact numbers.  I wonder if AR Level 8 can actually ditch Agility almost completely and go all-in Perception/Int.  The loss of stealth points and interloper won't "really matter" except vs. Tchort because Jumping Bean + Adrenaline can be heavily abused to squeak by everywhere else I think.  Worst case scenario the Lunatic Mall has to be done using TONS of traps and gas grenades (but I think still entirely doable this way, just painful).   This *probably* won't work though because I think you have to destroy mutagen tanks (the healing and tentacle regen is too fast otherwise) and high stealth+ interloper really really helps with that.  But maybe that optimal supersteel gear + optimal tabis + optimal balaclava + optimal cloaker can pull through...

10
Ah, thanks for clarifying and sorry for jumping the gun.  For some reason I did think you meant at Level 8 but Level 25 makes a lot more sense.  The followup numbers you mention are also quite interesting...

You should expect to be able to do 400-500 damage in a bursty turn at level 8

...because I was thinking that Level 8 AR could probably still output something in that range (at least at melee distance) ~500-800 a turn at Lvl8 and Lvl9 (I doubt there's actually THAT much difference between these two Levels).  But I have also still not tested.  If ~800 a turn is possible, then with optimization it would still be possible to kill the tentacles and use the Sower trick to kill the Eye.  Even though the tremors are somewhat of an issue, they're not THAT much of an issue that ~9-12 extra turns (a few extra bursts per tentacle kill and Eye) would outright prevent success.  It should still be fast enough to outpace tentacle respawn and tremor damage with good RNG.

I would also need to test Mushroom Forest because I need to make sure Focus Stim would be enough to overcome the loss of Concentrated Fire.  Mushroom Forest really is not a joke at Level 10 and it's very hard to hit things (though I didn't use traps and I bet that would have made those chompers a lot easier...assuming traps work on them, maybe they don't?).  I also need to move at least 1 point from Intelligence into Perception probably, but that's REALLY going to hurt craft possibilities but I think it's necessary.

It sounds like truly a lot of testing is needed here...so I'll need to learn what the best way to CheatEngine or customize a character/save file is (maybe I'll need to manually find some offsets in memory) unless the SaveFile format is not actually encrypted (haven't actually looked if it's some standard serialization or something).

As mentioned earlier I think anything below Level 8 will be extremely interesting because I will learn something about the game far beyond my imagination...I really don't know what a Level 7 build would look like but if it can be done with Traps then there must be ways of dealing with Tchort threats that I've never even considered.  I'll plan to leave that exercise to others if/so long as I can come up with something for Level 8.  The more I think about it, Level 8 should be possible but just incredibly hard and frustrating because in principle Concentrated Fire is not doing anything that pure luck of 9 consecutive crits would not also accomplish.  The loss of crafting makes little difference I think, the only potential issue is literally not being able to hit anything but if my memory is correct the chance to hit is never 0% and as long as hit chance 40% with goggles can be achieved you just need to hold out for luck approximating (0.4)^9 which is definitely possible.

It's also entirely possible that muzzled << compensated.  I actually don't know how the math works.  Does the increased burst accuracy ALWAYS increase your chance to hit or only if your gun skill is high enough that you're not below the bare minimum accuracy and getting rounded up?  In other words, if an enemy has 3000 dodge/evasion and you have 10 gun skill, I think you have 10% chance to hit because it's bare minimum, does compensator give you 20% minimum or do you stay at 10% because of the skill disparity?

I think the executable's MSIL is obfuscated so it's hard to know without dev confirmation (or very labor intensive debugging/testing).

11
Okay now you've really got my attention.  Thanks for those explicit numbers.  If your spike damage number is correct here, that will be key:

and with max Will and skill, you then deal 3.5k in AoE

If this is true, then it should approximately also one-shot the Eye on Dominating plus or minus a MKIV grenade or trap or something and making healing not even that relevant.  This would need to be tested of course.

If I understand what you are implying, this build would basically be able to walk right up to Tchort (after perhaps blowing tanks for prep and killing two tentacle) and rely on the same Sower lockout trick (or maybe quick tinker although Sower Tentacles last much longer and get created for you) and basically blow away Tchort in a couple turns.  This would eliminate any concern about tremors (it would be just as fast as AR build assuming there is a way to kill a tentacle 1:1...which I assume Psionic can do at Lvl8?) so I would probably never even need to use more than 1 superstim and a regenerative mixture during initial tentacle kill phase (though bringing a regen vest can't hurt).

Am I understanding correctly?  This definitely sounds like an approach with a lot of potential if it can burst out 3.5k damage in a single turn...assuming the tentacle can be killed in 2 or is it 3 turns?  Can you refresh ThermoD and get the damage of its initial application?  If not, how do you burst down a Tentacle in two turns at Lvl8 Psionic?

I don't think luring Adds makes as much sense because the Tchort Tentacles are 100% going to be right next to Tchort already and you have to deal with them anyways...they might as well give you the lethal...although maybe you WILL need to do a 2-enemy Devourer Devourer strat if a single Tentacle cannot actually be killed safely 1:1 at all. But this also sounds highly RNG-dependent so I am less confident about it.

A lot of stuff to test for sure.  As always thanks for the feedback and discussion.  I think sub-10 has to happen and it would be nice to drop all the way to 8 instead of dragging along to 9 first.  I suspect that Level 7 and below, in addition to high probability of not being possible, will also be SUPER hard to execute even if possible, even in comparison to Level8.  If Level 8 is doable, I might take a vacation and try the ironman challenge or something else.

12
General / Re: Iron Man / Permadeath runs on Dominating difficulty
« on: March 21, 2018, 08:52:26 am »
Hi - interesting premise.  I realize that this challenge is very different from a minimum-level challenge, but I do have to agree that this certainly seems like a valid and interesting puzzle.

Specifically, I'm interpreting your challenge as "start a new Dominating character and complete the game by taking train to Northern Underrail after killing Tchort without having ever died or loading your save (other than for taking breaks) and without any cheats/character imports/etc."

This would be a specific challenge case of a more general "minimize number of deaths and resulting loads" problem statement.  I realize this is all a bit obvious but I just want to state my assumptions/understanding up front.

I am not actually certain that your addition of the restriction "No Quick Tinkering" is largely relevant.  It certainly might be, but I am not currently sure if the feat and point investments just to obtain that would actually make a meaningful difference in the majority of cases.  In other words, are the highest risk areas actually areas that Quick Tinker helps with?  If not, Quick Tinkering doesn't actually do much.  For example, does Quick Tinkering protect you from DoppelGangers and Mental Subversion (Tchortlings) or Tentacles or Tithonus Ghosts?  I am pretty sure those are some of the biggest threats.

I also really think you and TheAverageGortsby have already touched on a lot key points here:

Stealth is ridiculously OP, moreso I think than anything else.

 If you pick your fights carefully ....

Avoiding as much combat as humanly possible is going to be key.

Having thought about this for a little bit and factoring in my current low-level experience, I think all of the following will apply at Level25 (and you have definitely mentioned some of these):

- Crafting is an absolute given because it is so absurdly (and consistently) broken, so all planning should account for at least 120 mechanics and likely ~100 electronics at least.
- Moderate Chemistry/Bio is also basically a given because it would be crazy to not have Aegis/Irongut/Jumping Bean/Focus Stims/Bull Head 100% of the time (they are insanely cheap for what they do)...and regen tailored vest probably makes sense (I haven't tested but in theory they do sounds like the single strongest combat armor type on Dominating).
- Lockpicking and hacking max (late game) are definitely partially wasted, especially since mechanics should ensure that you do not need IRIS check anyway, but definitely make sense as priority skills
- I am not actually certain whether Guns is optimal here (I have never played Psionic let alone Level 25 Psionic), but assuming you are doing Guns, obviously SmartMuzzled Hornets with FullAuto/ConcFire are easily optimal.
- I think JKK questline is by far the safest/most easily consistent via simple stealthing.
- I don't think Aimed Shot makes sense as a Feat as you should never be in a situation (I believe) where bursting or moving/consuming are not better choices.  Early on things are so much easier and re-attemptable that choosing a feat just for early game seems like a massive waste.
- Wherever possible, you should skip virtually every mission that involves combat unless you can formulaically win 100% of the time
  - In deep caverns bee-line for the knife quest and NEVER let Eye of Tchort get active except when you have to.
  - In deep caverns, the Tithonus lab is going to be insanely scary unless there are techniques I'm not aware of
  - For the T44 Manual laboratory: this place is VERY scary and random to pure stealth, even assuming you use gas mask and take southern route, but fighting can also go sideways fast since there are so many adds.  I would practice stealthing this area.
  - I would never do the IRIS fight (mechanics is a given)
  - I would never do Mushroom Forest without CAU armor and Irongut,
  - I would also NEVER do the final Free Drone Base Kill mission (waaay to risky and no payoff)
  - Do as few Arena fights as possible (just get your title and recognized by an Oligarch)
  - I would always stealth lunatic mall (a lot on the line / nerve wracking for sure, so practice is important, but since this can be done at lvl10 you should have an ENORMOUS margin in practice)
- Tchort fight is also going to be scary.  There are so many random ways to get killed or screwed by adds/tentacles even with Mutagen puzzle+TNT on tanks, but maybe not as bad as I think if you are fully optimized and have tons of drugs 

I think by far the scariest moments are all going to be in Deep Caverns (by a massive margin).  Have you done any testing or practice with your ideal 25 Character there?  I don't think that anything that comes before DC will be anywhere near as troublesome with planning.  Tithonus Lab in particular strikes me as the single most insane random showstopper with no recourse, so I'm very curious if there are in fact ways to mitigate the risk here.

13
You can gain access to prisoner easily without any skill checks after finishing beast quest line.

Ah, good to know.  I think opening up the Oculus shop has no downside, so that is probably something to do 100% of the time (and lots of easy Charons as well).  Now that I *do* have a formula for doing the Oculus Foundry Quests before Beast, I may or may not end up using this alternative in the future, but I definitely appreciate having another option.  Although your suggestion may be mandatory if I have no STR and cannot open vents (though that also raises other quest-related questions).

Super Steel thread making infused cave hopper leather armor and infused pig leather boots are probably worth contemplation for your play style.

Interesting, yeah, that sounds like it's worth investigating.  I couldn't seem to find on the Wiki a detailed outline of the exact quality+buffs numbers so I suppose this has to be done empirically through the game engine...I think it shouldn't be too bad though because now I've simplified all necessary experiments to lower quality (more easily obtained) possibilities that must be possible at Level 9 and below.

On Dominating, stim healing is gimped badly enough that Regen Vest is basically god-tier.

Interesting, I did consider this item but I didn't do any tests and I'm again not really sure what the math looks like for an optimal Level 9 skill level craft output.  How long can such a vest heal for?  I think it would need to be chugging along for a LONG time (or can you toggle it on and off repeatedly with minimal loss?) to justify using...interesting for sure.  How often will you need to recharge vs. how much larger will the tremors get while recharging?  The exact math matters a lot, because when you are taking 40-75 damage per tremor, even something twice as powerful as stim healing will only buy you marginal time due to the increases every tick and the fact that your max HP is a fixed number.  I did do a waiting test and verified that at least after ~20 minutes the damage never stops increasing and I got as high as ~100 per tremor which, again, is basically a guaranteed death if anything else is remotely in a position to damage you even with morphine and aegis active (and morphine has a long cooldown).  Tremors can also critically hit so even at the 75-80 damage level I can easily get one shot, and I'm not sure if saveloading will fully allow dodging any crits.

EDIT: If you're pretty comfortable with your health pool from that AR run, my previous ThermoD build suggestion might be better with 6 CON and 10 INT

Hard to say because at Level 9 my overall health pool potential would be even lower and I am not sure I was exactly "comfortable" at Level 10.  Obviously 6CON + 2 from Junkyard Surprise will actually mean substantially more health but HP is intimately connected with damage for purposes of calculating success.

I would be more comfortable with 100HP (instead of my Lvl 10 150HP) if I was able to walk into the fight with 200 gun skill and 12 or 14 Feats.  What I was saying is that I am currently hugely skeptical of (though I am always open to evidence) of a build that can have enough damage at Level 8 or below to kill Tchort that could also survive long enough to deal that damage, so the question is whether ThermoD can accomplish this (admittedly I still know NOTHING about this skill other than reading the Wiki entry).  Given that I think ~180-190 HP is the maximum reasonable amount of HP that is possible, I do the following simple mental projection:

Level 10 AR run --> 150 HP --> able to 3 shot Tchort and 1-2 shot Tentacles and got up to ~40 minimum damage tremors, one shot by most tentacle attacks and can never survive being hit by anything (for example, Devourer claws) in combination with tremors
Level 9 and below run --> 190HP --> ~25% more HP so I can afford to 4 shot tchort and maybe 3-4 shot tentacles but still can never be hit by anything in combination with tremors

As you can see, what I am trying to emphasize with this back-of-the-envelope math is that my Level10 AR run did not have a huge amount of time left on the clock even though I'm dealing out potentially 1500 damage a turn with 9 critical concentrated fire rounds, so it's simply not enough to prove that I can stay alive longer at lower levels...I have to be able to stay alive a LOT longer while also dealing out enough damage each interval where I'm not healing to actually kill Tchort.

Tchort has True Sight, which means I will be forced into combat and visible to tchort-adds and tentacles near the eye.  In my Level10 run just now, I didn't need to come and go on the Eye platform in shifts.  If I did, I would have been destroyed by the wave of adds slowly spreading out.  Again, locking them away with traps makes sense, but if I need to lay 20 mines to kill Tchort and each mine placement/destruction cycle takes 4 or so turns, I would have to have placed an absurd number of traps to choke off the adds while avoiding any aggro and staying healed up the whole time.

It may be possible (though I REALLY doubt lvl 5 and below) but I think as a first-order test I need to know how much damage mines can do (for that build) and how much damage ThermoD can do (for your build) to just know mathematically what is possible (damage per turn average). 

14
PART 2 of 2

Tchort Institute

- As stated historically, without the 120 hacking skill I was worried that I would need to spend a lot of time to get to the elevator (either slogging fighting or questing)...
- But the strategy I've found (was this intended?) is extremely helpful here, yet again.  Again I sneak into Eidein's office, kill his guards to trigger "tremors", go down elevator, go BACK UP elevator, and he is there.
- A very nice side effect for this run in particular is that I am able to place multiple layers of Bear Traps around where Eidein WILL be but before he's actually there, so there's zero chance for problems in that fight (I basically can't hit Tchortists unless they are immobilized and even then it's much more likely that I'll need multiple bursts to kill them).
- I just assassinate Eidein and go back down and done with Institute (within minutes of joining the Institute if I wanted to, this time around I had to do a lot of shopping and shop refreshes).
- I do not know if this is a "bug" but it is confusing that some zones "autohostile" you even when stealthed (rest of the Institute) and others don't (elevator area) even when they both have an alert "Eye" to indicate "Controlled Zone"
- This time I carried 227 Weight Units of stuff (Lightly Encumbered) into Deep Caverns with me, although I didn't end up needing most of the stuff causing excess weight.  I did prove to myself that there's no issue getting to the elevator and past the faceless with just a slight movement point debuff.

Lunatic Mall (Jumping Bean Mall)

- Previously: I opened the door at the entrance without combat, then kill the first two Lunatics in a single burst turn (while patrolling guy is not in line of sight).  This allows ending combat and waiting out the "Noise Alert Movement" (is there a term for this?) of the rest of the floor.  After this happens, I could directly stealth to floor 2 without anymore fighting.  This is definitely 100% reproducible.
- This Time: I confirmed that the above is still possible, but I need a flashbang to immobilize them in order to semi-reliably kill both with a single burst.
- This Time: I *also* confirmed that I can *easily* stealth past these two without even killing them if I use adrenal/cloaker and tight clicking (but I prefer killing just because I don't like those two in particular :p )

- Previously:
  On floor 2 I perform the same quick kill/wait out Noise Alert Movement for the 2 guys on the left-hand-side, then the patrolling guy, then the person inside the bottom left store
  I can destealth, adrenaline, stealth from the bottom left nook to the statue without any more fighting (it's fairly tight but with 235 stealth not too bad, could definitely do with some amount less)
  So in total I completed Lunatic Mall with 2 + 4 kills total.

- This Time:
  On floor 2 I still perform the quick kill/wait out routine for the 2 guys on the left and the patroller, but I never really needed to bother with the corner Lunatic inside the shop, you can just wait on the movement cycle for that Lunatic to be all the way in the bottom corner.
  This time I used Adrenaline + Jumping Bean + my stealth gear and I was like a ROCKET.  Those buffs clearly all stack and it's insane.  Jumping Bean + Adrenaline + 23% boots + Rathound Regalia + Interloper made it so that I *easily* (2/3 full yellow eye) could dash in and grab the statue even though I'm only Level 10!
  That's kindof insane.  Jumping Bean must add another ~50 or even 75 effective stealth (whenever speed is relevant which is virtually always).  Completed Lunatic Mall with 2+3 kills (but verified I could have done just the 3 kills).

 
First Protectorate Epione Lab Quest to kill Upper Underrail Free Drone Base with Shredder

- Toxic Gas Grenades + Molatovs + Traps let me attack from the northern hatch and CRUSH them. 
- All previous playthroughs I was lockpicking the front door and luring them out.  This playthrough I was afraid that without all that comfortable space to kite backward through it would be extremely hard but wowee Toxic Gas+Molatov is devastating.
- Coupled with traps and hiding in the door nook, I picked them off very efficiently and only had to saveload ~4 or 5 times (not really that many considering I'm level 10 and how many things can go wrong there).
 
 
JKK Questline

-This time I did JKK Quest Line which is WAY easier with stealth, although the last quest with the Spiders was annoying. 
-For the last JKK quest, I used molatov + running to burn that stupid patrolling spider in the utility room down over the course of three throw molatov/run/door close cycles.
-For the last JKK quest, I lured the Industrial Bot into the neighboring room from the pump controller and then closed the doors on him because screw him.


Junkyard

- As always, I'm now hitting and running, hitting and running, for those dumb photon/laser turrets.  Not sure how to deal with them otherwise.  Very grindy/slow (though the plasma pistol from Scrapper Quest line helps).
- I literally didn't even realize that opening Vents with omnitool required lockpicking until this run.  Luckily this build can just use Rathound Barbeque or Regalia (8 Str total) and Crowbar for Vents everywhere, including Arke.  I didn't plan this and was relieved to confirm it...

OPTIONAL STUFF I DID

The Beast
- I killed The Beast and unlocked SuperSteel but wasn't sure what to do and didn't want to spend time investigating unless I was sure it was needed.
- Whoever mentioned traps + toxic gas grenades (TheAverageGortsby ?) you are SPOT ON.  Man the bladelings do NOT like toxic gas grenades, especially chaining them and stacking them to make a giant thick cloud on their spawn.  Those things are amazing for making bullets do full damage even vs. that armor.
- I also noted that alternately leaving the forge map on the south side transition and then popping in through the caves gives me the ability to easily clean up stragglers and possibly is an alternative kiting strategy in the general case
- I also found that if I stealthed and let all bladelings spawn and stayed out of combat the game would consistently crash towards the last wave...so that's interesting.

Carnifex
- He literally cannot be allowed to have a turn UNTIL you have secured yourself in a corner 2 layers of traps deep.  There is no margin: if he goes first you WILL die (well maybe I survived at 10HP once in ~50 reloads).
- I had to saveload ~20 times for each turn where I had higher initiative (and this is WITH Agi steak).
- Flashbang->stun->stealth->exit combat is still the gold standard 100% Win strategy here (all previous Arena fights are essentially just strictly easier instances of applying this same strategy).
- You have to be completely sealed in with traps, he sees them all instantly and will otherwise walk around them.
- I used Molatov to make him switch sides after getting trapped on one side so he physically walked over 3 bear traps.
- I didn't even need to use W2C ammo, but unless he was trapped I could easily fire 9 rounds and hit nothing but air.
- I think he took a total of 6 or 7 full 9 round bursts which is pretty insane.
- Rathound Regalia is nice to resist the Yell debuff

Access to Oculus
- I had never done this before in a game so partly for fun, partly for Charons in case I needed to melt them down for SuperSteel
- The Foundry Prison quest is actually insanely frustrating without any skills.  I had to do it in a VERY particular way.
- When you're ready to talk to Todor Tallywhacker or whatever, and ONLY then because you only get ONE SHOT and have to do this all at once, you have to:
  - enter the prison
  - stealth to the Control Room with the Supervisor (NEVER let a camera turn red at any point)
  - use traps to kill Supervisor (he will patrol into them)
  - grab his key
  - use his key to get to the vent room
  - open vent with crowbar
  - go talk to the guy in his cell
  - You can never come back (they go to lockdown) so this should be the last step before you go to Chief of Police and complete quest the "good way" by getting Todor released.


LOOKING FORWARD

This run result has convinced me that Level 9 is very likely possible using the same (but optimized Percept/Intelligence with perfect crafts and no wasted points) approach as this run, but it would be miserable and require LOTS of RNG to make up for the lack of damage.  I'm not 100% certain even, but I think it's 51%+ probable that it can be done.

This run result has also convinced me that Levels 5 and below are borderline impossible.  The only 5 or below suggestion I am currently aware of would involve Level 4 being reserved for Quick Tinkering.  This would be mean that Interloper cannot be selected and that even with 11 Constitution +2 from Food a character could have at MOST 180 HP. 
Perhaps +1 from footwear is possible while still having enough stealth as well, but the point is that even at 190 HP Tchort will very quickly be dealing 1/3 and 1/2 health pool damage (damage that DEMANDS immediate healing).  "Slow" damage sources such as mines I don't believe can possibly be set fast enough to kill Tentacles and then Tchort and if Tentacles are alive there is NO CHANCE that a Lvl5 character will
be able to survive a gauntlet of even a single tentacle plus Tchort floor shocks (unless it is exactly a tentacle that is 100% absorbed by shield...but that still means 2 tentacles had to be killed).

I simply don't believe that slow damage can work against Tchort given the other threats present and the time it would take to deal with them.  Want to lay a bunch of traps to lock away the adds?  Even placing the traps to lock off the chokes takes too much time, you will have to do it in shifts because of the damage adding up but since the damage increases over time you will never be able to lay more than a small number which they can get past in a single digit number of turns...

And if you don't lock away the adds?  You will never have enough stealth to walk past them, ESPECIALLY without interloper, so you would either have to depend on RNG that they NEVER leave the last Tank TNT area or you would have to kill them all off (again how?  you don't have time).

In fact, based on my experience, I doubt a Level 6 mine-based approach makes sense (I could be wrong of course, but Quick Tinker would have to be SUPER Quick, much faster than I am imagining, to outpace the tremor damage).

However, there might be a non-mine or mine-hybrid approach that could work that hasn't even been discussed...I haven't put much thought into what is possible at Level 6 so I don't feel as comfortable fully ruling it out...but that's my current thinking, that Level 6-8 are the bare minimum levels and the true bare minimum might just be Level 8 or Level 9 with ARs and perfect RNG, perfect everything (maybe even perfect Tentacles though again I'm not sure if there is a strategy where type of the remaining 3 would matter).

At present, with the data I have obtained from my playthrough I am only honestly comfortable suggesting Level 9 as likely possible.  I really truly don't even know if Level 8 can work as the lower HP and the lower accuracy as well as the lower quality gear will start to all add up.  As mentioned, I definitely would need to see some contravening evidence or examples to address the parallel threats of Tchortling/Spitters/Devourers Adds, Tchort Tentacles, and Tchort tremors at this point for any claim below Level 9 AR (which I think the math supports, but just requires insane ~99% crit RNG + perfect craft gear).

I suppose I should look into that Level 8 Thermodynamic Destabilization suggestion...because if that's actually viable at Level 8 I can see it blowing ARs out of the water.

Anyway, thanks again for reading and suggestions!  I may need to come back and edit this response a bit as it has gotten quite long and I'm not sure if the double-post is mod-acceptable...

15
PART 1 of 2

Hey guys, I'm back with an update.  This time I have quite a lot to say so hopefully I won't go over a post character limit or something like that (20,000 characters, I did go over...).

After my previous success with a Level 14 Assault Rifle character, I decided to try psionics and/or traps based on all of the feedback, and that's the character I started to create.  However, after spending a significant amount of time re-reading your feedback posts, looking at your suggested builds, and sitting at the character creation screeen, I realized I still wanted to try one more AR run.  My reasoning was as follows: I don't have any deep knowledge regarding psionics or trap use but I know a lot about AR and if Level 14 AR isn't actually too tough (which it isn't) then I might end up wasting time heavily optimizing for something that AR can still beat.  So I wanted to find the limit of AR thinking that that would give me much faster feedback as to where I should start targeting optimization for other archetypes.  An additional reason: I really like ARs, Dakka Dakka / Lead Storm for life  :P!

Having decided to try another AR run, and given that I knew the "minimum" AR Level which had to be well under 14 (since 14 wasn't too bad), I came to what I think is an obvious conclusion/initial guess: target Level 10.
Level 10 seems like a good balance because in comparison to (a fully optimized) Level 14 removes a healthy fraction of crucial gun skill, excludes any sort of extreme top-level crafting possibilities, drops a non-negligible number of maximum Health points, and blocks skill-check-based workarounds for virtually all end-game content/dangers (except stealthing).
At the same time, Level 10 still gives access to Concentrated Fire and that Feat is absolutely core to all Assault Rifle builds and provides crucial damage potential.  Level 9 would be a massive drop in power due to the strict loss of this Feat coupled with even lower Gun skill at a point where every single Gun skill point matters (and weaker everything else as well).

I have now completed Dominating at Level 10 using:

Str: 7
Dex: 3
Agi: 7
Con: 3
Per: 9
Will: 3
Int: 10

Feats:
Echoing Soliloquy (No reason not to get this right?)
Gun Nut
Opportunist
Recklessness
Suppressive Fire
Interloper
Full-Auto
Concentrated Fire

Skills:
Guns 60 (85)
Stealth 60 (149 with just gear and skills, with almost all gear/buffs effective 215...I had been using Adrenaline already but now really begin to appreciate it and ALSO Jumping Bean is really a big deal here, more discussion below..always using Rathound Regalia because 10% movement speed +25 stealth with Interloper seems to me to be worth more like +100 true stealth)
Traps 60 (I used 6 EMP traps during Arke Fight and a couple crawler poison traps for Tchort Fight, but never got too fancy here, used plain old bear traps virtually everywhere else)
Mechanics 60 (90)
Electronics 60 (90)
Chemistry 60 (90)
Biology 60 (90)
Tailoring 60 (90)

I crafted (with House):
- 9mm Smart Muzzled Hornet 17-56, +46% special attacks (and had a second one with +41% for swapping out if durability got low during a single fight...but turns out never needed the backup gun)
- Efficient High Frequency Shield Emitter 1132 Capacity, Conversion Rate 10.7, 7/23/96/141/180
- +61 stealth cloaking device, 97 energy, turn on costs 27, consume 9.32 every 5 seconds
- +31% Special Attacks Smart Night Vision goggles (low 42 energy capacity, turning on costs 9, consumes 3 energy every 5 seconds)

Noncrafted:
- Rathound Regalia (I never ended up crafting Super Steel String but maybe I could have crafted infused SuperSteel XYZ that is better?  probably not possible to craft better at Character levels below 10 though, I'm guessing)
- Chemical Assault Unit Armor (I just kill the CAU members outside of the Free Drones base when it's "Go Time" and I don't care about burning Protectorate bridge)
- Standard +17 stealth balaclava (didn't even bother crafting a better one because I didn't want to waste rare high quality cloth on +1 or +2 stealth), ChemicalAssaultUnit Armor for Tchort and MushroomForest, RathoundRegalia for early Tchort fight and almost everywhere else except early (before Core City main quests)
- +23% movement speed/32 stealth Ninja Tabis (I *know* I could have crafted better here because I had some almost perfect-skilled cloth but I killed Carnifex and he was wearing these and I wanted to see if I could cheap out on the cloth in case I needed it for later and never bothered to upgrade later since stealthing was a minimal concern).
  I wore these Ninja Tabis for the entire Tchort fight again because I'm not actually sure what a better choice would be...maybe infused constitution boots but didn't end up needing to go that deep into crafting
- Bullet Belt (of course)
- W2C bullets for all the "major" fights in Deep Caverns obviously (Faceless +300 is mostly enough, I also brought 300 W2C rounds with me from buying nonstop through the run, I finished Deep Caverns with 250 rounds left)

Final experience bar (before boarding Train to North Underrail): 203427 / 10000

Notes of interest:

Overall

- I needed to save and load a lot more, but not as much as one might think.  I would say that the maximum saveloading was for Carnifex (totally optional) and Tchort Fight.  Versus Carnifex with Agility Steak active I was able to get initiative/first turn once every ~20 loads.  The second time I flashbanged him and that's that (see below).  I personally spent dozens of saveloads on the Tchort Fight but excluding experimentation/learning the "winning strategy" only took ~4 save snapshots and ~30 loads (totaled from ladder to end combat with tchort dead). The fight description/summary this time was very convoluted (see below).
- I took significantly more time in Deep Caverns and on Tchort compared to my Level 14 and Level 16 runs.  The gatekeepers (Mushroom Forest, Arke Fight, Tchort Fight) all took ~3 times as long at least this time around.  To be fair about this comparison though, my Level 16 run was practically a speedrun.
- I don't believe at any point in this run I needed to saveload more than approximately *20 times* to force "super RNG" (Carnifex initiative was definitely the single most-abused saveload, moreso than any Tchort Fight point).  I did end up needing to saveload substantially more, but I would say it is more a question of requiring a few extra saves and load all over the place rather than any single place requiring an absurd number of saves/loads (which is a good thing for reproducibility/consistency).
- The Tchort Fight is absolutely the gatekeeper of the run and it is getting very dangerous/hard even at Level 10.  I don't think I do my next run "blind", which is what I have been doing up until now and including this current run. By not playing "blind" I mean I will probably want to use Cheat Engine and/or character editing and/or at least Export/Import to pre-fabricate a character to test the Tchort Fight before spending ~50 hours to get to that point and possibly fail).
- Based on this run, and in particular the Tchort Fight experience, I no longer believe/accept that a Level 4 run is possible, though of course I will always remain open to video or detailed writeup evidence.  More discussion below.
- Stealth is insane, but stealth with stealth +movement tabis, rathound regalia, plentiful Adrenaline, and also plentiful Jumping Bean is so good that even at Lvl 10 I'm stunned at effectiveness (easier stealthing than my Level 16 run).
- I really put Molatovs, Flashbangs (extremely good for getting burst hits to land...as well as stunning enemies so you don't die), and Toxic Gas Grenades to work
- In addition to toxic gas grenades, this run also really utilized a ton of medical crafting and some fishing so I felt like I was almost playing an expansion pack already (so much new stuff that I never used before).

- I didn't even craft 100% optimal gear (didn't have gear that was crafted using the exact skill requirement that I reached at House with Workbench buffs) and I think that in retrospect at least one point of Intelligence would have been better placed into Perception. 
  The relative gain of one more intelligence point is actually not that much in terms of crafting item/weapon quality, and for weapons I think ~5 skill requirement can be on the order of +1 damage on a 17-56 damage weapon which is basically nothing compared to the bonus of +5 Guns at such low gun skill levels.

- If I did this same run again and anticipated that I would 100% need to make SuperSteel I would probably reallocate SOME points from Chemistry+Biology+Traps to Mercantile because scrounging for money is a pain.
- If I did this same run again and knew that I would 100% not be making SuperSteel, I would spend way more freely and not waste a ton of time walking sometimes instead of taking train (though I didn't do that TOO much, it was more that I was constantly loading after findout out a shop was a bust).

Tchort Fight Consumables

- Focus Stims - This run I learned how to craft them and came prepared with a stack of 6.  Didn't need to do the old killing Coretech employees routine.
- Regenerative Mixture - I still don't know how to get regenerative mixture except by killing Caerus Residential Vents guy "Gunter Vasilica" so I still did that. IMPORTANT NOTE: I almost forgot to bring +1 Perception Goggles for the 10 Perception check in the vent, I got lucky and found some on one of the shelves somewhere in Deep Caverns and avoided a very embarassing load to hours earlier.
- Aegis - This run I learned how to craft them and came prepared with a stack of 6.  Fun fact, I had never fished anything ever before in my ~250 hours of play until this run for medicine crafting.
- Iron Gut - This run I learned how to craft them and came prepared with a stack of 6.
- Jumping Bean - I crafted 9 of these but didn't need all of them.  These things are amazing!
- Super Health Hypos - Since I'm friendly with Faceless I have 100% reliable Super Health Hypos to pair with Morphine (although I guess you only could ever really use 1?)
- I saveloaded for +2 Constitution -2 Dex (or something useless) Junkyard Surprise (this is definitely worth doing instead of just accepting a +1 constitution burger).

- Adrenaline Shot - I didn't end up using these vs. Tchort because the downside is too dangerous/slow (every second counts because of the monotonically increasing Tchort shock damage) and by the time I knew I could win the fight, I was more afraid of the adds killing me without being able to burst than of not having the damage for the Eye.
- Bull Head - I didn't use these or need them, but I crafted a couple anyway just in case.


Tchort Fight Strategy Summary

- Descend ladder in full stealth except for Gas Mask swapout when needed.
- Place two TNT charges on *the left/West side* Mutagen Tanks and blow them.
- Place two TNT charges on the *right/East side* Mutagen Tanks and blow them.
- At this point, *wait for parade* then dodge back to the left side.  I say "wait for parade" because there were 12 (TWELVE !!! I screenshotted and counted) Tchortling/Spitters/Devourers that all grouped up on the right side platform in a nice square ball to investigate the TNT and stayed densely packed on the order of seconds.  I saved/snapshotted upon seeing this, I hope it's consistent that they don't split apart and spread out most of the time as this save is core to my "successful fight" line.
- While walking back the left side, near the center-right choke, I placed two crawler poison traps (not a perfect lockout) for use later (and so tentacles not respawning while trap-laying).

- Mutagen Puzzle was Completed, so now Three tentacles to fight: I got cryo, mind control, and acid. (My Exitus 1 "seed" was: MU L8 DX R1 ET MH AL PZ UP WW KG XT E6 B4).  I've been stealthed 100% of the time until now so Tentacles don't actually have any involvement and the way I kill them I don't think type matters at all.

- The left-side (West side) shack is crucial (maybe central shack can work but not sure) because you can go back there, exit combat and get a save snapshot, switch to full combat and then PICK OFF TENTACLES WHILE DODGING BEHIND SHACK DOOR after each burst.  The Tentacles appear in a perfect formation such that none of them can hit you inside shack but all can easily be hit from right outside shack door.  The Tchort-adds on the other side of the map DO NOT aggro or come over (I think they are too far away for sound).
- I killed two tentacles this way (exited combat) then walked back toward center and third tentacle popped up and since I was able to point-blank it I killed it in one go (plenty of focus stims for all these separate shots).  Again exit combat and get a save/snapshot in.

NOTE: I did lose *some* time up until now but I am clicking/hotkeying fairly efficiently and yet Tchort is hitting me (without Aegis or Morphine) at this point (all three tentacles just killed) for 40-75 electricity damage (he has quite a damage range) NON-critically!!  That's A LOT of damage considering I have 150 HP total *with* my CON buff food.
I believe I dipped into the middle shack on my way from right to left after the final TNT and SUPERStimmed to stay alive and opted to bandage-wait-out the stim cool down at that shack before starting combat at the left shack so I probably advanced the counter 2-3 extra shocks since I don't think the waiting was necessary...but to be clear here the damage gets very high much faster than I think one would expect.

- Key moment: a spitter (and some other Tchortling enemies) are random-walking around toward middle right (where they were all investigating that final TNT) as I am coming back from killing the tentacles on the left.  I initiate combat and run to the northern choke and am able to get the spitter to lock off that choke with Tentacle spawns (especially likely if spitter got caught in trap).
- With tentacle spawns on the ground blocking access to all other Tchort-adds and even keeping them at an out-of-range distance, I have free reign to chug irongut/aegis/focus stim/remaining regen mixture, go to melee range on Eye and unload bursts nonstop.  Although melee range+nightvision is now required, Tchort is still dead in 3 turns (thanks concentrated fire!) and nothing is hitting me the whole time except a couple spitter tentacles.
- There are a LOT of mini-Tchort adds but I chuck a molatov and a toxic gas grenade on the bridge choke and hide at max distance next to the Eye corpse and basically they still mostly take care of themselves with the madness debuff.


Tchort Fight Notes

- I did Mutagen Puzzle because tentacles are very hard to kill now and my HP is just too low (and my current fight approach requires tentacles completely out of the fight for a while).  I can no longer reliably one-burst tentacles though it is still possible to do.  However, not sure if it's possible for ALL tentacles.  Some tentacles I believe it is not possible or virtually not possible unless point blank AND very lucky.  The Cryo Tentacle can only take 1/3 life damage from one of my bursts depending on my luck.
- The Cryo Tentacle basically one-shots me 80% of the time, same with the Mind Control tentacle, so I *had* to use a strategy that involves never taking on more than 1 at a time (in fact, in this case, none of the tentacles ever get a shot on me).  Before (Level 14+ with evasion) I was able to potentially take a shot from one tentacle while killing the other but that's off the table now.
- I can't even 100% kill the adds in a single burst because of misses, but if they are approaching me on a choke/bridge they'll still never get a hit off as I certainly two-burst them superconsistently at any medium/short range.
- I am chugging my consumables before every mini-fight segment and refreshed everything after the Eye died just in case the Tchort-adds got ugly, I'm glad I made a lot.



Arke IRIS Fight

- I placed 6 EMP mines outside surrounding the bottom left (south west) corner console room and thats where I made my stand.  When the south entrance bots initially spawn they will be in a group and you get a good hit/free bursts there.
- I also places ~5 bear traps around the console room for good measure.
- W2C bullets mandatory for the Industrial Bots.
- This was the closest my gun came to breaking (actually just barely got into yellow as I killed last loose bots) and I can imagine that if I were slightly less lucky or lower level that a backup gun would be necessary
- The four center turrets are easy to clean up at the end but I was also able to exit combat and get a save in before killing them
- I did NOT end up needing/using any consumables such as Focus Stim or Aegis.


Arke Basement

- I still very reliably one-burst Greater Coil Spiders.
- Now that I was mentally prepared for this to be a pain, it wasn't so bad.  I refined my silly strategy where instead of killing everything with traps, I planted one TNT and lured almost ALL spiders out of big electricity room and into hallway.  I let them have a civil war fighting scorpions FIRST (just waiting, without TNT) then used TNT further down the hall.  I was able to actually sneak by without fighting (again except for earlier, the first small room with two spiders for door key to open the big electricity room).
- Similar to Caerus Residential, there are two key strategies for Electro Spider killing:
  Need to kill one spider and you are near door?  Burst spider then close door.
  Need to kill two spiders (Wasi Abdul keycard, Arke basement key)?  Burst one spider, adrenaline, flashbang, burst second spider.
  Need to kill three spiders?  You're doing something wrong, there is never a time this is needed.  However, perfect adrenal/flashbang probably works for up to 3. 
- You can always easily just go right to the ladder in the last room of Arke Basement, saveload if the scorpions bump you.  You can obviously trap if you really need to but it's not even necessary.


Mushroom Forest

- Is this actually doable without CAU Armor?  I'm even more convinced it's not.  These guys do *serious* damage to me even when wearing CAU armor unless I ALSO have Iron Gut active.  Even then they can hurt me.
- At lower levels I may need to actually grab the Biohazard Boots and come back here (I have not been wearing those) because the damage is insane (when factoring in that they reduce your HP pool by 10-20%).
- Jumping bean + adrenaline actually makes running past the second half of room before the heart room quite easy (doable in one try).
- Taking the "northern" path to the heart and hiding behind the rock or whatever allows me to alternate between killing the sentry and shooting the heart from relative safety HOWEVER
- I now needed to first aggro and "pull" the northern chompers/spitters and kite back left while killing them from some amount of cover as they are too dangerous to fight in place.  I can't reliably one-burst them.
- I am glad I crafted a bunch of Iron Gut as I am not sure Mushroom Forest is doable at Level 10 (or below) without it.  I also needed to exit combat/heal up before going in for the sentry/heart alternating kill.

Pages: [1] 2