Fatal refleeting.

[Pumpkin]

I read the thread diagonally. Not being on the American schedule, I have a lot to catch up. The ideas from both sides are very interesting and I think are we already boiling the topic down to its equilibrium.

I scraped my individual replies as they are a bit old now and many things have been said; I would only repeat someone's point (or worst: mine). So I just want to bring a little something new in the conversation: there is the ironman idea floating up in the air. If this is going to be (and it would be as a "really sure?"-warned optional-option), the game will need to somewhat let completing a game (below personal challenge level, of course) without savescum possible.

Minotaar made me think about it.

Spoiler for Hiden:

3) don't make tactical defeats game-ending! That just leaves "reload and don't mess it up this time" as the only way to proceed, and I don't think the game will be better for going in that direction.

Sometimes, in my experience, it goes this way. I try something, get wiped because of something that I could have foreseen (9x MkV self-destruct guardians on a core world, lately), insult myself, reload and try to figure a way (AoE-immune missile frigates, maybe?)
And it's fine, IMO. Well, if the failure is blunt enough with a complete fleetwipe and minor damage in the ranks of the AI, even with no big-reprisal-blam-you're-dead, I reload. 9x MkV self-destruct guardians on my whole fleet is not a "tactical defeat", that's a flat-out death-sentence.

Your ideas are interesting, nonetheless. I tend to keep what currently is in the game (fleetwipe-level-reprisal) and make the threat smarter at deciding when to call the game's end. My idea was a bit artificial, indeed.

Sure! I'm not saying reloading and general trying of stuff should be discouraged, it just doesn't need to be forced when a mistake like this happens. I feel like it should at least be possible to just accept it and move on, for people who are into that.

Sometimes, though, you do go in expecting serious or complete losses. I would dread to get an objective done at the cost of the whole fleet, and get destroyed by the Game Is Over Now Department as a reward.  This kind of mechanic also needs to be explained in advance to a new player, or, well, you know.

There is also another quibble I have with this whole idea of a "coup de grace" from the AI - namely, if this is a thing that can happen, why didn't it happen before the start of the game? The justification in AIW has been that the AI could do it, it just didn't care. Why does it suddenly care now? What is different in this new world that makes this possible?

So players will need to be able to move on with mistakes. So that "fatal refleeting" mechanism should quickly finish already-lost-games without just increasing the "move forward" difficulty.
Well, if the "ironman optional-option" becomes a thing, of course.

[Tridus]

Well, this thread got exciting when I went to sleep. I'll try to reply individually, simply because one post that's replying to five people and five things at once can get confusing very quickly.

I would say that's tricky.  If I take all my science and unlock defensive units, then shouldn't it hold out for much longer than MK I stuff?  Wouldn't you expect MK I fleets to regularly fail completely, especially tackling higher MK worlds?   Think of it this way, if you build a MK III world and send out MK I waves against a MK III AI world, what are your expected results?  Should be something similar to what you do to AI waves at that level.

At this point I'd say AIP needs to be included in this kind of calculation, because early on , those defenses should hold just fine.  I'd think they would hold well into mid-late game.

I'd say that's fair. What about during the HW assault? If your defenses are still impregnable at that point without any ship support, you've won. The only thing left to do is solve the HW puzzle (or zerg it, depending on how nasty it is). But there's no way to lose anymore. That's the condition we get in a lot of 4X games where the game is over but it takes 50 more turns for the game to realize that it's over.

To me, this comes down to a fundamental question of "what is the lose condition?" Is it strictly "you let AIP get too high?" If that's the case, impregnable defenses at the expected AIP is probably fine, since they should crumble once AIP gets high enough (although in Classic you can make them near impregnable, period).

Alternately, is "you threw away your entire fleet and the AI punished you for it" a lose condition, or at least a "you lost some planets and have to retake them" condition? I don't see how the AI is particularly threatening if it just sits there waiting for you even when it knows you have no means to attack it, instead of coming at you.

[Tridus]

If the AI is utterly incapable of defeating your defenses alone, haven't you effectively already won the game? At that point, it's just a matter of doing the actual labour of killing enough stuff to win. Losing is impossible.

Even going beyond the "buy only turret tech":

No.  Because the AI homeworld's defenses are stronger than mine.
I'm tossing a Mk3 fleet against a Mk5 world, whereas the AI is tossing a Mk3 fleet against a Mk3 world.

I don't see how that matters. If the AI can't beat your defenses, it doesn't matter how long it takes you to beat theirs. You will, eventually. It's only a question of time, and you have all kinds of time when you have unbeatable defense.

The AI HW's defenses also don't get rebuilt when they get destroyed. You can whittle it down piece by piece. The AI can't do that, because if it doesn't wipe out your whipping boy completely and press deeper in, it'll be rebuilt before they attack again.

[Tridus]

I would say that's tricky.  If I take all my science and unlock defensive units, then shouldn't it hold out for much longer than MK I stuff?  Wouldn't you expect MK I fleets to regularly fail completely, especially tackling higher MK worlds?   Think of it this way, if you build a MK III world and send out MK I waves against a MK III AI world, what are your expected results?  Should be something similar to what you do to AI waves at that level.

At this point I'd say AIP needs to be included in this kind of calculation, because early on , those defenses should hold just fine.  I'd think they would hold well into mid-late game.

If one decides to invest heavily in defensive techs, it is by default their offensive abilities will be more limited. So offensive actions are more likely to suffer heavy losses / be wiped and require a greater portion of commitment. However, these defensive techs allow one to more easily survive a wipe will still doing just fine. A mechanic that goes based on a portion of fleet lost punishes this defensive play style more.

Yeah, exactly. I don't know if people misunderstood me, or what. I never said "if your fleet gets wiped, the AI should instakill you."

What I said was "if you have no means of attacking the AI, the AI should free up some of its defensive ships to go on offense and counterattack you in force." If you invested mostly in defensive tech, you have a decent shot at still surviving that. All it means is that the AI isn't going to sit there passively with a thousand ships doing nothing while you build your fleet back up, because why would it? That's the perfect time to launch an offensive.

The AI has all these lovely space docks full of ships waiting to be attacked. If the AI knows it's not going to be attacked, it should use some of them to attack you. Hell, that could help your fleet out if you lure enough stuff in and take it out: you've just weakened the ship support in some of the AI's worlds.

The goal is *not* fleet wipe = game over. It's fleet wipe = AI aggression, and you need to deal with that until the AI has to back off to defend against you again. Certainly makes refleeting periods a lot more interesting because there's stuff organically going on.

[Tridus]

There is also another quibble I have with this whole idea of a "coup de grace" from the AI - namely, if this is a thing that can happen, why didn't it happen before the start of the game? The justification in AIW has been that the AI could do it, it just didn't care. Why does it suddenly care now? What is different in this new world that makes this possible?

The player losses have to be "big enough", for an appropriate value of "big enough"??

Also, at the beginning of the game "refleeting" might be building 20 fighters and 20 bombers, which shouldn't trigger an aggressive response.

So any definition of "fleet wipe" would need to consider number of ships lost, percentage (of player ship capacity) lost, current AIP and probably other things too.

The simplest definition is probably one of relative power. Is your fleet powerful enough to be any kind of threat to the AI worlds that are on alert/recently attacked/reachable/whatever (the changes in AIW2 change how this works somewhat)? If the answer is yes (and for mk I worlds near your start, it takes very little power for that to be true), the AI would want to keep its docked ships for defense. If the answer is no, then the AI can free up some of those ships for an attack.

I wouldn't expect that response to become massive until late game when you're doing HW assaults and such... but really, if the AI HW is under attack and it wipes your fleet, should it's other 20 planets really sit passively with thousands of ships on them doing *absolutely nothing*?

[kasnavada]

I wouldn't expect that response to become massive until late game when you're doing HW assaults and such... but really, if the AI HW is under attack and it wipes your fleet, should it's other 20 planets really sit passively with thousands of ships on them doing *absolutely nothing*?

I know you're probably not finished, but +10000 to this. With the limitation "can / should the player be able to withstand the counter-attack ?"

[Cinth]

I'd say that's fair. What about during the HW assault? If your defenses are still impregnable at that point without any ship support, you've won. The only thing left to do is solve the HW puzzle (or zerg it, depending on how nasty it is). But there's no way to lose anymore. That's the condition we get in a lot of 4X games where the game is over but it takes 50 more turns for the game to realize that it's over.

To me, this comes down to a fundamental question of "what is the lose condition?" Is it strictly "you let AIP get too high?" If that's the case, impregnable defenses at the expected AIP is probably fine, since they should crumble once AIP gets high enough (although in Classic you can make them near impregnable, period).

Alternately, is "you threw away your entire fleet and the AI punished you for it" a lose condition, or at least a "you lost some planets and have to retake them" condition? I don't see how the AI is particularly threatening if it just sits there waiting for you even when it knows you have no means to attack it, instead of coming at you.

In AIWC there is no such thing as an impenetrable defense.  Every line has a breaking point against brute force and the AI has some tricks.  I've lost due to Raids and EMP guardians.  I've lost due to AIP being to high for what I had placed.

Also in Classic, you get to a point where there AI just won't even encroach.  If you defensive firepower is high, it just camps wormholes.

The goal is *not* fleet wipe = game over. It's fleet wipe = AI aggression, and you need to deal with that until the AI has to back off to defend against you again. Certainly makes refleeting periods a lot more interesting because there's stuff organically going on.

I like this.  Though in Classic, those ships would end up camping wormholes (in my big games).

I wouldn't expect that response to become massive until late game when you're doing HW assaults and such... but really, if the AI HW is under attack and it wipes your fleet, should it's other 20 planets really sit passively with thousands of ships on them doing *absolutely nothing*?

I know you're probably not finished, but +10000 to this. With the limitation "can / should the player be able to withstand the counter-attack ?"

I think this ties in nicely with you most recent thread.

[Tridus]

This backs up to the more general question of refleeting speed (because none of the three child threads seem to be touching that directly), but in my experience I wind up hitting two things that make refleeting more annoying than it needs to be.

1) If I have a large quantity of metal in reserve and I want to dump it all into my fleet quickly I need a huge pile of engineers on station. Metal gates production and I get that if you don't have metal you don't get ships and that's fine, but sometimes I have the metal and its just a pain making it into ships. I tend to spread my engineers all over the place so that if any particular world gets its defenses bruised, there's ships available to fix the turrets but this leaves enough spare engineers for maybe two worlds to produce ships at less than a crawl. A mk 5 Zenith starship takes something like 220 minutes to build without any engineers assisting. If I have the million and a half metal or so available then its just busywork to have to wrangle all of my mk 3 engineers and drag them over and I usually don't bother. Does anyone else play this way or should I just be making better use of warpgates and stacking all my engineers at a core planet?

Anyway, what I'm trying to get at is I would like the option to, if I already HAVE the metal available, just spend it in bulk immediately. You still can't build more ships than you have ongoing metal income for in the long term, so no infinite fleet zerging. It just feels weird that because metal essentially IS time*territory (there isn't a way to get it other than paying time and AIP) I'm paying a time cost in metal and then paying a time cost AGAIN just waiting for the metal to spend. The counterargument is "well this is the AI's counterattack window" but it never actually counterattacks in such a manner that you lose.

I leave engineers in FRD in whatever system has my space docks precisely so they can assist build them faster. Trouble is that the metal cap is not hard to hit, at which point production is wasted... then you have to build a ship that costs the metal cap, which goes fast. Everything after that is now based on your per second metal, which is slow.

Refleeting would naturally be faster with a higher cap because you'd be able to stockpile more. Maybe the cap should increase per world or something.

Regarding the AI being more of a threat when the player has made a poor move, the threatfleet is probably the best mechanism for that. Just have preemption on at a moderately strong level as the default and the threatfleet can operate more frequently.

It could, if it has enough ships. Hence my idea of the AI freeing up some of its idle ships to do that when it knows it just flattened your fleet and you have little to no offense to threaten it. What good are idle defensive ships when there's nothing to defend against?

Actually... let me reframe the issue a bit. What SHOULD a failure mode look like? If the player has made a critical error and left themselves open to potentially losing the game, what exactly ought that sequence of events look like? Should it look different between an average difficulty game and a high difficulty game? A low AIP game and a high AIP game? What sequence of events could lead to a loss for a player playing a low AIP game at 7/7 or 8/8? SHOULD there be such a condition? (I think so else there is no real risk/challenge, but maybe someone feels that playing 'optimally' you should never need to reload a game. Ironman types perhaps.)

This is probably a question for a seperate thread, but I think it's a great question that gets at part of what the core of the game is. I'd suggest you start a new thread for it so we can talk about it without it being all refleeting based.

[Tridus]

I wouldn't expect that response to become massive until late game when you're doing HW assaults and such... but really, if the AI HW is under attack and it wipes your fleet, should it's other 20 planets really sit passively with thousands of ships on them doing *absolutely nothing*?

I know you're probably not finished, but +10000 to this. With the limitation "can / should the player be able to withstand the counter-attack ?"

Oh good, I'm glad we're finding some common ground.

The answer to that IMO is "maybe". It should not be an instant game over. It should be an attack of sufficient force that it's some kind of threat to you (just throwing some ships away isn't very interesting). If your defense is good enough, you can probably hold it off, especially if your tech is more heavily skewed defensive. If it's more heavily skewed offensive? Possibly not, but in that case you've invested more power in your ships and should be fleet wiping less anyway (plus if you ignored your defenses in terms of tech, you kind of expect that you'll need some ships in a defensive response fleet to help out).

[Tridus]

In AIWC there is no such thing as an impenetrable defense.  Every line has a breaking point against brute force and the AI has some tricks.  I've lost due to Raids and EMP guardians.  I've lost due to AIP being to high for what I had placed.

That's true.

Also in Classic, you get to a point where there AI just won't even encroach.  If you defensive firepower is high, it just camps wormholes.

Also true, and kinda silly. It won't have to do that now that it can idle ships in the docks that were added for AIW2, provided it can also release those ships when it wants to attack, or load them into carriers (aka: mobile docks).

The goal is *not* fleet wipe = game over. It's fleet wipe = AI aggression, and you need to deal with that until the AI has to back off to defend against you again. Certainly makes refleeting periods a lot more interesting because there's stuff organically going on.

I like this.  Though in Classic, those ships would end up camping wormholes (in my big games).

Cool. Conceptually, this seems to have some support. Hopefully Chris gets a chance to look at it.

[Cinth]

Cool. Conceptually, this seems to have some support. Hopefully Chris gets a chance to look at it.

I think it's a sound idea that kinda refines what Threat Fleet and SF should do in certain cases.  I'd think the veterans would support something like this also.

[kasnavada]

I wouldn't expect that response to become massive until late game when you're doing HW assaults and such... but really, if the AI HW is under attack and it wipes your fleet, should it's other 20 planets really sit passively with thousands of ships on them doing *absolutely nothing*?

I know you're probably not finished, but +10000 to this. With the limitation "can / should the player be able to withstand the counter-attack ?"

Oh good, I'm glad we're finding some common ground.

The answer to that IMO is "maybe". It should not be an instant game over. It should be an attack of sufficient force that it's some kind of threat to you (just throwing some ships away isn't very interesting). If your defense is good enough, you can probably hold it off, especially if your tech is more heavily skewed defensive. If it's more heavily skewed offensive? Possibly not, but in that case you've invested more power in your ships and should be fleet wiping less anyway (plus if you ignored your defenses in terms of tech, you kind of expect that you'll need some ships in a defensive response fleet to help out).

And I'll disagree with that part. The principle I'm ok with - if defensive techs are strong enough, should be able to defend, if offensive tech are enough, shouldn't have wiped.
I don't think that defensive techs are in any way "balanced" compared to offensive techs. By which I mean that the same number of tech points are going to boost your defensive capabilities way beyond than whatever your offensive capabilities can do. And, therefore, it's a bit too easy to make oneself next to invicible while defending, while the same when attacking... not that much. Not to mention that you can lose your offense while defending. Unless you let the AIP treshold go way beyond what's wise for your defense capabilities, you're mostly safe. Whereas a bad encounter on a deep strike can always happen no matter how strong you are. The point is to avoid the "if the AI can't do anything, then it might be long but you've won" answer you've done a bit above.

Hence https://www.arcengames.com/forums/index.php/topic,19120.0.html

EMP & raid

Here I must admit... I possibly have not have opened the full bag of tricks (I don't often play above 9, too tedious), but in my experience EMP can be hunted down via the search interface of the map screen and don't spawn with waves. Raid starships I countered with gravity turrets and "setting" the turrets to auto-target them first.

[Cinth]

Quote

    EMP & raid

Here I must admit... I possibly have not have opened the full bag of tricks (I don't often play above 9, too tedious), but in my experience EMP can be hunted down via the search interface of the map screen and don't spawn with waves. Raid starships I countered with gravity turrets and "setting" the turrets to auto-target them first.

The thing is to give the AI tools to subvert but not make defenses irrelevant.  It then puts it on the player to counter those threats.  I think that's a good thing to do.

[Pumpkin]

Remainder.

The AI will be able to take territory back. The consequence for a fleetwipe / weakened defense wouldn't be an immediate game over, but rather the opportunity for the AI to step forward and conquer one (or some) of your worlds. It would be a similar consequence for player carelessly taking more territory than can be economically defended (but with the per-planet energy, I don't know if that situation will be possible, unfortunately).

Anyway, a player would only lose a game after several iterations of the downward spiral or "fleetwipe -> frontier planet(s) lost -> less economy -> ai more aggressive -> frontier planet(s) lost -> etc".

And while I'm thinking about that, I think that per-planet energy killed something important without us noticing. I'll try to make another post about that.

[skrutsch]

Cool.  Conceptually, this seems to have some support. Hopefully Chris gets a chance to look at it.

I think it's a sound idea that kinda refines what Threat Fleet and SF should do in certain cases.  I'd think the veterans would support something like this also.

Some will; TechSY730 suggested essentially this solution in message #10 in this thread.  I blame that Chris guy for trying to introduce a new mechanic instead of fixing an existing one.