「Stellaris」Combat Disengagement Explained

[@leobastian_]

To summarize: Combat disengament will make you weaker(from lost bonuses and cause ships will retreat earlier), but If youre already losing anyway, or still winning, its worth it

[@rowbot5555]

So basically, it's good to make your looses less loosey and your wins less costly

[@AzureRath97]

People under-estimate the value of stalling a losing war. Trickster helped me a lot in a pinch.

[@loafofbread9400]

Early game multiplayer its usually lose now to lose later, seen alot of people try to stall out but they just dont have the eco unless person trying to box you in gets attacked by another

[@Snoogen11]

@@loafofbread9400 I agree, stellaris is an economy game with a military side tacked on. Stalling wars only work if you can replace ships faster than your opponent, otherwise you'll be on the wrong side of the resource attrition. That being said, if you can get more *actual* kills than deaths, stalling wars can set you up for victory in the next war.

[@Arbaaltheundefeated]

@@Snoogen11 Well, in multiplayer specifically, stalling also weakens your opponent more in the long run. The longer he's spending time and effort and resources the less of an advantage taking you out is going to give him... and I don't know about you, but if I'm being taken out of a game I'm damn well going to do my best to not see the one who did so go on to win the game, I'll replace buildings with strongholds and planetary shields with every last mineral just to force a slow grind, heck I'll do scorched earth tactics if the situation is hopeless, make myself as unrewarding a fight as possible :D

[@Stefan_Anon]

Seems that y'all want to see how trickster/hit and run does against no retreat. So I did some testing! NO RETREAT/UNYIELDING (100 destroyers, 25.4k fleet power) vs HIT AND RUN/TRICKSTER (100 destroyers, 20.7k fleet power): 36.2 losses vs 21.9 losses. No Retreat wins 10/10 times. Follow up engagement: NRU (64 destroyers, 16.1k fleet power) vs HRT (78 destroyers, 16.2k fleet power): 34.5 losses vs 17.6 losses. No Retreat wins 10/10 times. Another follow up engagement: NRU (30 destroyers, 7.6k fleet power) vs HRT (61 destroyers, 12.4k fleet power): 30 losses vs 4.9. No retreat wins 0/10 times. This set of tests nicely demonstrates the usefulness of No Retreat to boost your fleet if you are equivalent to (or slightly behind) an opponent in terms of ships. However, the losses you sustain mean that you will be destroyed in the long run against someone that makes use of disengagement and engages in battle multiple times. I'd say it's best used when on the defense for a few reasons: 1) Stations. No retreat boosts your stations, meaning that your defensive bastions become better. On the other hand, you will take losses against enemy stations when attacking due to lack of disengagement. 2) Reinforcements. It's very important to keep your fleets at capacity with no retreat, and it's easier to do when in home space. 3) Importance of victory. When on the defensive, losing battles will mean losing stations and planets, which hinders your capacity to fight. You can afford to lose some progress while on the offensive, but if you lose when on the defensive you can lose your entire empire. A few tidbits on No Retreat otherwise: 1) It's useful in the lategame when disengagement does not matter, but at the same time you have to sacrifice Rapid Deployment's range. If your opponent is going for range, you will likely lose. 2) I would not recommend using it if you are superior to an opponent. You will win anyways, but sustain significantly more losses than with any other doctrine.

[@JabberwockOclock]

Can you also spawn in Reth Undol too? His trait is just Unyielding-plus, and he can come with Unyielding too, giving him -66% disengagement chance.

[@kerbalkerman6827]

Thanks my dude

[@jonrussell739]

Thanks stefan! I would love to try some no retreat builds.

[@JabberwockOclock]

@Clark Kent You misunderstand. Reth Undol's trait is Hell's Heart, which makes him less likely to disengage. It also comes with -5% evasion. Paired with Unyielding he might as well be on No Retreat doctrine.

[@thetaomegatheta]

Hormone Replacement Trickster

[@AshenVictor]

I like to stack sublight speed as much as possible. You lose 100% of battles you didn't show up in time for, and sublight speed is the biggest determinant of how long it's going to take a fleet to get to where you need it to be. It also lets you dictate when and where a fight is going to start so you can do it on favourable terms. (Especially useful if you can shadow an enemy group on follow orders and snag the trailing fleet as the leading one hypers out). Rapid Deployment plus as many Gale Speed admirals as practical.

[@noobsunleashed9703]

just have a tug corvette

[@JupiterCobalt]

@@noobsunleashed9703 Does that still work? I read somewhere that fleet speeds were no longer determined by the fastest ship, but rather the slowest. Haven't bothered checking in game

[@noobsunleashed9703]

@@JupiterCobalt still works :)

[@KatyaAbc575]

@@noobsunleashed9703 Some of us dont like to rely on exploits.

[@antred11]

@@KatyaAbc575 Why some people may add corvettes as an exploit, I like to have "balanced" fleets (for role-playing purposes), and balanced means that my capital ships must have proper escorts.

[@mustapha9918]

Very nice video :D. Combat Disengagement Chance is great when you re playing defensively. When attacking an empire tens of stars away, I largely prefer "no retreat" war doctrine. That's especially true when I m playing military and built a fleet that is way over my naval capacity. (because blitzkrieg is actually a decent strategy) So I would love to see you commenting on the other war doctrines !

[@ChaoticNeutralMatt]

Building over and throwing waves of ships at the enemy?

[@jiminfested]

Congrats to the detailed lifelike 3D rendered corvettes, destroyers, and cruisers

[@blahmaster6k]

Haven't even watched the whole video yet but I can already tell I'm going to love it because of the incredibly cute MS Paint drawings.

[@LCInfantry]

You his mom?

[@ravenwiley9918]

I love stuff like this, I hope you keep this up!

[@crynsinia3182]

Yes!!! Please do more of these videos! One about leaders in general would be really nice. Like if it's worth it to get very high level leaders.

[@feartheghus]

I’ve always cared about disengagement because I like not losing people, I’m sentimental so I want to keep each of my ships intact and alive. It also helps with my battleships because it means o can have entire fleets of elite ships that simply don’t die, even if I lose I only lose like 2 ships at most from large fleets while the winner can lost half of his own and more battleships than me.

[@Mexterinfinite]

Does the bonus for "friendly" in the equation apply just to home territory or also to allied territories? Would there be a difference between federation allies and allies you called into a war if so?

[@ryorai5804]

Its also important to take into account the force disparity bonus.

[@morunaga]

MS paint ships are the reason I came, thanks.

[@valtteriahokas6238]

When is the mod that makes your drawings into ship appearances ingame coming?

[@pooface1041]

i always just ignored combat disengagement, but i might have to re-think it for peaceful/inward perfection civs

[@sumikae7635]

I agree with many points, but the tests fail to consider a player's active decision to emergency FTL before their ships become severely outnumbered in a losing engagement. With Trickster's bonus to emergency FTL, it's a strong trait for conserving your forces for both losing and winning battles in a war of attrition. From experience, I've won wars against my friends who had slightly superior economies even though I usually lose half the battles, and it's because trickster minimized my losses after every engagement. I always try to get trickster early and mid game because I find the reduced impact of war on my economy by minimizing losses worth it.

[@therealbahamut]

Disengaging fleets can save a lot of resources on rebuilding ships because, although you "lose" the engagement, you keep more ships out of a smaller fleet. It's also an interesting, albeit expensive way to bring ships home from distant stars. The fact that the enemy wins engagements can lead him to overconfidence. If he saunters up to your territory and finds a bunch of repaired fleets that escaped earlier, he can lose the war he was certain he had won already, or at the very least he can be relentlessly stonewalled away from taking your systems until a peace can be declared. Thus, I think the "trickster" admiral trait is quite apt.

[@Drachenfurie]

Could we get a video on No Retreat? How good/bad is it really? And how good is it early on, since Gestalts can use No Retreat if I remember correctly?

[@donkeylover4449]

I don't know how good stats wise it is, but I have done amazingly risky stuff with it. Such as take on 6 empires at once, 3 being players. And won the war. I felt like such a luck spiritual empire.

[@ShadowcatAlfa]

10/10 drawings

[@Sephiroth36977]

Just started playing Stellaris two weeks ago, this is officially the newest Stefan video I have ever seen.

[@joshuabauer591]

260 upvotes and 0 downvotes. We love you stefan

[@donaldobrien2614]

I swear it's not been less than a month since another Stefan video told me to max combat disengagement chance

[@Zan-tx6jv]

How do strike craft and disengagement chance work? If 1 ship in the fleet disengages does their strike craft also leave? I've had encounters where the entire strike craft flotilla stuck around until the last ship died/left.

[@LeafBowParadox]

As a Fanatic Pacifist. Yes. It is. Carriers and Destroyers for win hahahah

[@ChaoticNeutralMatt]

Just had an interesting consideration, still at 0:55 in, but if your ships survive and can regenerate, then I'd say it's probably worth it.

[@JOHNCENA-ho5ts]

hey stefan do you have any gameplay of starnet grand admiral on the latest versions? I want to see how well you do on it. btw do you have any advice for organic/hive mind ringworld?

[@Gwynbleidd93]

The interesting question is how do hybrid setups fare against those? For example, would No Retreat + Trickster be better than NRU?

[@leafybean]

Woohoo more content! Thank you!

[@batska4237]

I would like to make a little addition to what Stefan said in the beginning about combat disengagement chance. So I wrote python script the other day to se how well ships will disengage if hit by different weapons. The result that I got was quite surprising to me. If ship is shot by a larger weapon, it has a higher chance to disengage. Say, ship hull can be destroyed by 4 shots of a weapon. It means that chance is 1.5*(1/4) = 0.375 when below 50% ship can take 2 hits. The chance that ship will not disengage in 2 hits is (1 - 0.375)^2 =~ 0.3906 So the chance that it does leave the battle is 1 - 0.3906 = 0.6094 or about 61% If same is done to the ship that is destroyed by 20 shots we get 1.5*(1/20) = 0.075 1 - (1 - 0.075)^10 = 0.5414 or roughly 54% Of course, this is just maths but my python script does not take any of that into account. It just calculates ship's health and disengagement chance after each hit. the results I got were: BBs at 3000 health that got fired upon by proton launchers (not in friendly territory) (468 - 1040 dmg) out of 100,000 about 68.2k BBs disengaged If same ships are shot by large gauss cannon (84 - 276 dmg) about 62.9k BBs disengage And for strike craft (6 - 17 dmg) 60.8k disengaged So yeah, just wanted to point that out. Bigger weapons are not as effective at destroying ships as smaller weapons are Upd: Boy was I wrong. I forgot that proton launchers have a boost (+75%) against hull. Yeah, in this case most BBs get just one chance do disengage so their overall chance of surviving is reduced dramatically and only ~52k out of 100k survive In cases when damage is so big it just dismantles ships, it is impossible to disengage. But if ships can survive multiple hits, going for smaller weapons may result in bigger losses for the enemy. However, this is true only if damage per cycle is the same for both fleets. You will not be able to inflict high losses, if your fleet gets dismantled first

[@Stefan_Anon]

I erroneously got similar results in my first attempts at calculating theoretical disengagement chances. You are missing two factors: that the kill shot cannot disengage, and that the HP the ships has when it CAN start disengaging is going to likely be significantly below 50% HP with large weapons. In-game tests conclusively show that larger weapons cause significantly less disengagement. When I tested smaller weapons on the destroyers, losses went from an average of 42 with large weapons to 25 with small and medium ones.

[@batska4237]

My script just takes ship's hp and applies damage to it. If after hit 0.0 < hp/max_hp < 0.5 it calculates chance to disengage. So it does take into account that ship has to be alive and it has to have less than 50% hp. However, I never got to test that in game. Setting up everything seems like a hassle. So with a real result a can do nothing but agree :) Results that I got probably make sense in some cases, but in real battle there are other factors that impact disengagement and losses that I did not take into account (e.g. that ships fire back at you and you don't have time to wait until your small weapons do damage)

[@zlorfik2428]

I'd be interested to see this comparison with the No retreat option as well, since a lot of the community seem to hate no retreat

[@mcawesome9705]

no retreat/unyielding and hit and run/trickster comparison is in the comments

[@Amyante]

I tend to build a defensive station in a black hole and then also build the comms jammer on it. No disengagement for you, AI :P

[@thefinn12345]

I constantly use trickster/hit-and-run unless I have some reason not to. I play defensively a lot though, so... It's rarely an even fight.

[@mytiliss682]

No retreat all the way. Unless your fleet have no chances to win, so you have to either prolong war till you mass it or cut losses and capitulate. Rapid deployment to catch small fleets and no-mil race.

[@Zentenski]

Do you think you could do a video on the value of the crystal plating and things like that? I'm having a hard time figuring out if it's worth adding those over more armor.

[@volodymyrboitchouk]

It depends on what weapons the enemy uses. In the face of neutron launchers, tachyon lances and perdition beams you should spam armor, but against disruptors, lasers, raulguns and missiles it's a fine choice.

[@Hiro1oo1]

I wish there was a way to ask the game to spit out some of the more raw weapon modifiers we've gained over time. I mean as like, if I'm playing vanilla, I want the weapons to show me their base stats, then show their stats when my empire's modifiers are on it. If I pick up a mod and it gives me some ways to go wild with range, I want to be able to see how big of a boost various +range modifiers have an impact. I want to see how significant my previous % boost damage tech just changed up my weapons. I just want more information at my fingertips.

[@Stefan_Anon]

You might be able to see more info if you use "debugtooltip" in the console

[@MijnAfspeellijst1234]

a battleship with full crystal hull, is this good or not? or do you get ducked by neutron launchers anyways?

[@tzucaza9801]

2:09 i dont 100% agree with that statement. Even if all ships of both fleets are within range the moment the battle starts not all weapons fire simultaneously (not even on the same ship). Because of that there is still a good chance to disengage even in the end game. e. g. Neutron lunchers - befor they fire there is always a delay of 1-19, after that they have a Cooldown of 150 (fom the files) This means: - the more they fire the more they get out of sync - If all weapons are out of sync they dont instant kill anymore - due the big hits on avrage a Battleship has (if within disengangement range) has on avrage a 85% Disengangement chance on neutron hit This also explains why Neutron Lunchers are so good (if we ignore the range advantage of 10) - They deal a lot of damage per neutron shot so they are less likly to give the enemy a chance to disengage - They have a long Cooldown so they wont get out of sync as fast as their kinetik counterpart (means again less disengangement) In short: Disengangement in the endgame plays a big part if the fight goes on for a while (if it is not ended in the first few volleys) Also I dont understand why you didnt touch "Crystal infused Plating" and "Crystal forged Plating" in a Disengagement Video where you explain that Hull affects Disengagement chance. Because they have an interesting relationship with Disengagement chance: - The more Hull you have the less likly you are to disengange - The more Hull you have the more chances you get to be within the disengangement chance and the more hits you can take within that range This means you shouldnt ever go full plating even if you want to maximize disengangement chance. For that you want to fill on average half of all available Utility slots with them (doenst matter what type of plating). At max only 1 more then that after this point the use of any more platings doesnt really effect your disengagement chance anymore.

[@AceDreamer]

saw another video how neutron vs kinetic and neutron vs lightning wins all the time, this video compounds it. The meta seems to be "spam neutron launchers and nothing else" (besides x weapons) and i think this really needs a tweak as to become les boring.

[@jamesmattila-hine1133]

i was wondering what are good weapons to use on your ships i something i haven't been able to work out

[@matthewmcneany]

That 25% home system disengagement chance explains a lot. I assume that there's a similar effect for fleet recovery time too. It always seems to take way longer for my fleets to return if I manually retreat or am defeated than it does for the AI I don't play enough PvP to have noticed any effect either way. I guess there's either a home system modifier or a distance to home starbase modifier affecting the time it takes a fleet to recover, there might also be a fleet strength modifier - i.e. larger fleets take longer to return. Would be interested to know what the actual mechanics behind this are too. The wiki just says distance is a factor with no details.

[@alexander-mauricemillamlae4567]

any ship you dont have to build twice is another ship you can build. Trickster helps with that.

[@RoboMuskVsLizardZuckerberg]

Yes this is my Mega Corp main battle style. Minimize cost and maximize profit

[@chesh05]

I was hoping you'd also talk about whether or not the Subspace Snare on the Titan is still worth using in the late game as well.

[@StevenKell]

Awesome, thank you.

[@SteveBlewett]

I adore the ship drawings.

[@ribhuhooja3137]

Does the AI get any bonus to disengagement? In my games, even when I have both trickster and hit and run policy, I take FAR more losses than my enemy (even when I have a stronger fleet and win the combat). Like, they lose something like 1-2 battleships while I lose 75-80% of my battleships. This is when I have trickster along with hit and run, and they don't have trickster. Makes no sense

[@Karina-winsmore]

Unyielding admiral and No Retreat war doctrine *maniacal laughter*

[@Pizza7478]

WE WILL WIN THIS FIGHT OR DIE TRYING!

[@Darlf_Sevil]

full run chance bonus builds :3

[@grahamturner2640]

How does ship design even work? Is there any good guide on selecting weapons for your ships?

[@Razzlion]

check out montu plays, hes done both ship type spam vs AND different outfits vs eachother statistics etc. well worth the watch!

[@maddorori817]

ship design basically means the difference between wearing a leather armor and a full plate metal armor in a swordfight

[@KaizerKlash111]

Early game : strike craft, if not possible autocannons, if not possible lasers. Late game : Neutron launchers+ Mega cannon/ FAM with a ration of 1 strike craft battleship for 3-4 artillery battleships. If possible, rush crystal hull plating for your early game ships, it is great and increases indirectly your disengagement chance (more hull = more opportunities for you to take a hit and run)

[@soulstealer5625]

The only things I design is a carrier dreadnought and various corvette designs for specialised swarms (whole corvette fleets).

[@ChaoticNeutralMatt]

Steffan Anon did a video covering this specific topic as well.

[@aldonisdraconis2933]

What I'm surprised about is cruisers being useful for something specific. May possibly do well against BS in economic costs per battle in favorable footing.

[@gabrielclark1425]

I like it because of the higher ship xp.

[@henrylancaster1234]

imy self look for fastist so warlike or the one that boost it by 20 i really like that way i can hunt down ships

[@Beef1188]

Devastator torpedoes to counter this!

[@sevsxes]

Soo, 25-30 year cruisers with trickster admiral + "hit and run" policy literary unkillable until protons, right?

[@Stefan_Anon]

Pretty much

[@soul1d]

Protons aaaaaaaaaah! Fighter of the Hull man aaaah!

[@elordis287]

What I discovered in my tests is that combat disengagement chance is apparently rolled per-day rather than per-hit. See: forum.paradoxplaza.com/forum/threads/on-disengagement-math.1442054/post-27100534 forum.paradoxplaza.com/forum/threads/on-disengagement-math.1442054/post-27101875

[@Stefan_Anon]

Im not so sure about your statement. Giga cannons deal too much damage to cause disengagement in the first place.

[@elordis287]

@@Stefan_Anon The key result was with Stormfire tests actually: 1) Two runs of 20 tests of 1vs10 Stormfire BB. 14/20 and 16/20 disengages. 75% disengages. 2) Two runs of 1vs20 Stormfire BB. 8/20 and 11/20 disengages. 47.5% disengages. If disengage chance is rolled per-hit it shouldn't matter at all how much attacks per day ship suffers. But we see here a large discrepancy in real disengagement chances depending on how many attackers there are. And since we use autocannons there is zero chance of oneshotting or even twoshotting BB.

[@Tobascodagama]

I wonder if combat disengagement chance could be kept relevant in the late game by computing per tick rather than per instance of damage.

[@stc2828]

No love for the focused arc emitter?

[@BlankTemplate123]

Huh. Wasn't there a thing where even if ship disengages, the Strike Craft still continue fighting? And that is used with Disengagement to become strong.

[@maddorori817]

that applies to defense platforms with Hangar bays, not with ships I believe

[@BlankTemplate123]

@@maddorori817 Heart works with both. That's why I asked.

[@Bruh-ff2tw]

Any chance of showing off an OP megacorp build

[@1Maklak]

Test "no retreat" versus "hit and run". Come to think of it, disengaging from combat is loosing a fight and therefore warscore.

[@Stefan_Anon]

You get war exhaustion from losing ships, not losing the combat. Its possible to cause more war exhaustion with a lost battle as long as you kill more ships.

[@1Maklak]

@@Stefan_Anon War exhaustion generally doesn't end a war anyway. War score does and to get that, I need to win combats and occupy systems, usually almost all of them.

[@mindmanna8135]

Hit and run, and the like, are great for defending against overwhelming forces, guerilla style wars of attrition, wherein you continually recuperate and rebuild few ships, and the enemy is deep behind enemy lines with stretched resources and supply lines, fighting the same disengaging/disappearing ships which return to finish the job.

[@marshalll4w359]

protons go brrr

[@AbdulHannanAbdulMatheen]

👏🙂

[@catisreckless4647]

I find it somewhat stupid and not very reliable. An ai empire finished becoming the crisis, so I got dragged into a war where they already had two star eaters. I got so tired of chasing them around, dealing a shit ton of damage, only to have them run after destroying a few ships from my fleet. The very definition of tedious. The Ai either ignores the chance entirely, or it's just not done well. Even with the starbase module the reduces escape chance, the ai can run as soon as their ships get weakened, even if I was able to do 90% or more damage within a few seconds. I never really had an issue with this until I had to deal with those stupid ass star eaters. The entire mechanic needs a redo, at least where the Ai is concerned, because they damn sure aren't playing by the same rules we have to. I finally sent 700k worth of fleet power to bumrush them, which is just frankly asinine. I was super excited to play this new version (I'm on console), but if I have to deal with this bullshit every time a genocidal empire decides to become the crisis, I'm going to have to prevent them from spawning in my games, period. The legit only time I was frustrated playing Stellaris.

[@alexanderormwiklund]

Banan jump away because jump is junpy jump with jumps for jumps and anti jump jump in big grabitational jump yes

[@bobdole6691]

so basically its more useless than it is useful and is very niche for specific situations in the early game and so it might not be worth teching into regardless until ur able to get the tech in no time at all. perhaps using an admiral with a trait, otherwise f it just focus on more important shit

[@opeox9356]

Добавляй русские субтитры что ли, наверно что-то интересное рассказываешь

[@Logan-dk8of]

fleet retreating is a stupid game mechanic anyway. according to the games own rules you cannot ftl from within a gravity well yet fleets do it in battle anyway and also go multiple hyperlanes away despite not having jump drives so it should be impossible. so it does not even fit the own game's lore, and it is annoying af because you go fight some ai and have double or more than their fleet power and then when you win you only kill like 10% of their ships anyway

[@Zorak9595]

I'm 3 minutes and 40 seconds in, and maybe you address it a little bit later, but.. your test makes no sense. Your test assumes a control fleet vs a fleet with the trickster skill to show that the trickster skill actually hurts your chance in combat. But this completely ignores that the enemy admiral has a 20% weapons range advantage. So the non-trickster fleet fires first.. so OF COURSE they win every battle. How can you conclude the trickster fleet loses BECAUSE of the trickster skill when the other fleet has a combat advantage? Had you wanted to do a real test, you should have chosen an enemy admiral with a neutral skill, like Ship Upkeep -10%.

[@verty6006]

not first :