Author Topic: Fatal refleeting.  (Read 5480 times)

Offline Vyndicu

  • Sr. Member
  • ****
  • Posts: 319
Re: Fatal refleeting.
« Reply #60 on: September 17, 2016, 06:20:47 PM »
Seem like a lot has been said in the last few week. Sorry if it has been said somewhere else or earlier.

I would like to bring a few points to this.

Obviously we do not want an "easy" low AIP route strategy but also at the same time we do not want to make it possible to create a defense line that AI can not "breach" with a 2000 AIP at 7/7.

So while it make sense to have AI do little more striking fairly short after a fleet losses, I feel this may be gamed* too much to AI's detriment. How?

Lets say I went with a low AIP (hovering just above AIP floor) until my "third spire city" fallen spire. Then afterward I crank the AIP just high enough to see golem/spire husks falling apart to my defense line. Guess what? I can pretty much crank my metal income higher than it should be theoretically possible at that AIP via salvage or what not. This may not happen in the version 1.0 but later expansions but is something worthwhile consideration. I am only bring this up because I don't want to create a strategy where it is possible for both side to become chokehold/staled on offensive side. Plus make not possible to crawl up to AI HW's wormhole with minimum defense (what one would call low AIP route).

So my question to all present. Is where do we draw the line on how much it can be gamed* and when to allow? For offensive/defensive?

In an ideal world we would try to strike a balance between AIP low route and defensive mutual stalemate. But AI war is much more likely fluid so all mechanic if any at all will also likely need to be fluid themselves.


Just to be clear. When I say gamed*, I make the AI do something silly or abuse a mechanic to favor the gameplay it is supposed to overcome in the first place. Building a heavily fortified line at low AIP route is definitely one of those gamed situation. Example: 7/7 fallen spire you can take at minimum 10 solar system for a 5 spire city layout and hold 15k threatfleet at the gate.

Offline Draco18s

  • Resident Velociraptor
  • Core Member Mark V
  • *****
  • Posts: 4,072
Re: Fatal refleeting.
« Reply #61 on: September 27, 2016, 11:22:09 PM »
Stumbled into this channel today on Youtube and while he talks briefly about the concept in board games, I think this is the effect that is missing from AIWC.

https://www.youtube.com/watch?v=_3M49nj0nUI

Offline kasnavada

  • Hero Member Mark II
  • *****
  • Posts: 911
Re: Fatal refleeting.
« Reply #62 on: September 28, 2016, 01:27:15 AM »
Stumbled into this channel today on Youtube and while he talks briefly about the concept in board games, I think this is the effect that is missing from AIWC.

https://www.youtube.com/watch?v=_3M49nj0nUI

Well... it's going to be complex to keep the "the player sets the pace" from AI War I and have this concept too.

2 other articles from wikipedia about board games which I find useful:
https://en.wikipedia.org/wiki/Eurogame
https://en.wikipedia.org/wiki/American-style_board_game

I don't like the style of what the article is writing because whoever implies that american-style game are more complex clearly never stumbled on Vladimír Chvátil's (among others) creations... and 'cause monopoly's rules are dumb as dirt. That said, a lot of people don't count it because Monopoly is nearly a parody of the American-school of board games. And quite of a few Eurogames don't have depth (like catane without expansions).

AFAIK, Chris leans heavily on the US-style of board game when designing games.

Offline Pumpkin

  • Hero Member Mark III
  • *****
  • Posts: 1,195
  • Neinzul Gardener Enclave
Re: Fatal refleeting.
« Reply #63 on: September 28, 2016, 01:59:10 AM »
Stumbled into this channel today on Youtube and while he talks briefly about the concept in board games, I think this is the effect that is missing from AIWC.

https://www.youtube.com/watch?v=_3M49nj0nUI

Well... it's going to be complex to keep the "the player sets the pace" from AI War I and have this concept too.

2 other articles from wikipedia about board games which I find useful:
https://en.wikipedia.org/wiki/Eurogame
https://en.wikipedia.org/wiki/American-style_board_game

I don't like the style of what the article is writing because whoever implies that american-style game are more complex clearly never stumbled on Vladimír Chvátil's (among others) creations... and 'cause monopoly's rules are dumb as dirt. That said, a lot of people don't count it because Monopoly is nearly a parody of the American-school of board games. And quite of a few Eurogames don't have depth (like catane without expansions).

AFAIK, Chris leans heavily on the US-style of board game when designing games.
All these links were very enlightening. (I wasn't aware of that American/European distinction at all.) Thank you both.

I don't feel that bringing the player to the edge of winning/loosing and letting it there really fits into the AI War's meta: there is, OMHO, a grander sens of progression and clear phases in AI War (early game, scouting, opening / mid-game, conquest, increase in power / end-game, big final battle, result of past decisions).

As Chris added that in the document ("Killing metal-poor players"), I believe the topic can be moved into the "resolved ideas" subforum.
« Last Edit: September 28, 2016, 02:05:00 AM by Pumpkin »
Please excuse my english: I'm not a native speaker. Don't hesitate to correct me.
Pumpkin>> Do I need another cure about paranoia on top of overexcitement?
Mal>> We play AI War, enthusiasm and paranoia are both required!

Offline kasnavada

  • Hero Member Mark II
  • *****
  • Posts: 911
Re: Fatal refleeting.
« Reply #64 on: September 28, 2016, 03:44:57 AM »
I don't think this topic's over.

As written, metal death condition part is never going to happen unless the ressources somehow become finite, or the AI starts to nuke everything in sight in "all" game modes. Which I doubt to be a very sane or popular move. It could be a possible game mode that forces an ending though.

My second point is that the whole of the wave mechanisms is meant to be the "edge of losing" part. Whether it's doing a good job at it, at highest diffs or (average diff + higher AIP) yes... but I think I spoke enough of that subject in the low AIP threads.
« Last Edit: September 28, 2016, 04:43:19 AM by kasnavada »

Offline Draco18s

  • Resident Velociraptor
  • Core Member Mark V
  • *****
  • Posts: 4,072
Re: Fatal refleeting.
« Reply #65 on: September 28, 2016, 01:12:06 PM »
I don't feel that bringing the player to the edge of winning/loosing and letting it there really fits into the AI War's meta: there is, OMHO, a grander sens of progression and clear phases in AI War (early game, scouting, opening / mid-game, conquest, increase in power / end-game, big final battle, result of past decisions).

The reason I pointed it out was because there's a certain threshold where the game is exciting because you're kept in suspense, just shy of being able to win or having lost, but aren't pressured by anything other than the fact that the win is juuust two moves away, if only you'd be allowed those two moves.

AIWC tries to pressure a loss by throwing a CPA or Exo at you and while, yes, there's the sudden excitement about "ooh, this might be the one that kills me!" when you survive you're not left in a state where you can make a move towards winning.  Instead you're rebuilding back towards a state where it'll happen all over again.  A constant cycle of "near-loss, rebuild" erodes the tension.

If we take the analogy to Jenga, a sudden CPA is like getting your turn only for your friends to switch on a fan, creating a sudden cross-breeze that makes it impossible to take out the next piece (or just outright knocks over the tower: you lose!).  Only if you manage it, then it's turned off for their turns.  At some point you're going to call someone an asshole and walk out.

The difference here is that the player's action dictates the tension: removing a tile from the tower might be the move that ends it, but if not, the game continues.  The next move draws the Sheep from an opponent's hand, giving you the resource you desperately needed.  The dice come up 9, getting you the stone you need.  Doubles, roll again.  Sure, there are some random elements involved, but they appear to be under the player's control: the player is the one rolling the dice: "Big bucks big bucks, no whammy no whammy!"

As opposed to "I'd like to kindly remind you that you must pay the bank 80,000 credits before Saturday."  You need those credits to keep playing and need those credits to create still more credits.  Constantly being knocked back to square 1 isn't progression, it's stagnation.

I don't know how to make this happen in AI War.  The only similar mechanics I can think of are more appropriate to a zombie apocalypse: enemies just keep coming.  Imagine that an Exowave is just a constant stream of units into your worlds, an unending tide (obviously at a strength level more close to that of a standard wave).  Now it's a pressure that can either be kept at bay or will, eventually, overcome your defenses and take you out.  You can't dedicate any of your fleet to solving this problem, because any units you do send are going to be there indefinitely.  They aren't dead, per say, they're engaged with the enemy, permanently, reducing the amount of ships you have to make attacks on other planets.  You can't repair (effectively) because your engineers just get shot, but the repairs you do make exactly equal the amount of damage you're taking.  You can't forget about it for very long, but you have breathing room to focus your attention to other tasks.

Hmm...maybe I did think of something.
(This is the TLDR)

Take that idea (prior paragraph) and put it into play.  The attacks will stop when you accomplish a task...say, completely removing the AI from an entire solar system.  The waves start when you gain a strong foothold (say...a third of the planets?) and doesn't cease until you control all of them.  (Note: the waves would not come from the same direction as your advance, or you'd be unable to send your fleet in to perform further conquests.  The continuous waves would be a 'diversion' of sorts that would allow the AI to make ingress back into the system, unless your fleet fights it off and captures more territory).

If you've expanded into multiple solar systems at once, you'll get multiple waves you need to fight off simultaneously.  Perhaps you've overextended.  If you can't handle the pressure and capture a whole system quickly, you're going to lose (the combined assault overwhelming your ability to repair faster than your ability to conquer).
« Last Edit: September 28, 2016, 01:13:56 PM by Draco18s »

Offline kasnavada

  • Hero Member Mark II
  • *****
  • Posts: 911
Re: Fatal refleeting.
« Reply #66 on: September 28, 2016, 01:40:08 PM »
I like the idea you're proposing. But I see issues with it - it will probably need a lot of tuning to work. A bit too weak and it's like the mechanic is not there, and too strong and you'd have to fall back. That said it encourages a faster conquest, which I like.

Offline Draco18s

  • Resident Velociraptor
  • Core Member Mark V
  • *****
  • Posts: 4,072
Re: Fatal refleeting.
« Reply #67 on: September 28, 2016, 02:25:11 PM »
I think the strength would primarily depend on factors like AIP or it's analogous counter part.  Early game taking territory will generate these waves, but they'd be largely ignorable.  It's when it gets to the midgame that they become a problem and you've "expanded as much as you can expand" effectively.  At that point you can either risk more territory or you can rely on deep strikes (which have their own risks).  It's no longer a stagnant rinse and repeat cycle of build, defend, rebuild because you're not being actively challenged every 2 hours.  Instead you're taking risks in order to accomplish goals.

Offline motai

  • Newbie Mark III
  • *
  • Posts: 49
Re: Fatal refleeting.
« Reply #68 on: September 28, 2016, 08:05:16 PM »
it seems to me that the only thing missing in aiwc for these concepts(fleet loss impact) to be real was that the ai could not build to push the frontier back. one step forward 2 steps back couldnt happen because the ai would not rebuid. hopefully aiw2 will have ai building options for advancing and rather than the outright death mechanics being brought up we will find that we have to retake positions(likely easier than original since they wont be mk4) but still giving a better back and forth feel that was missing and seems to be the core drift of this discussion. in fact many of the building options could be leading to the ai having those mk 4 planets in the specific path of advance because of the advance and lead you into the need for more scouting or strategic choices.

 this would in alot of ways make the refleeting question more dynamic and less in need with it having to be fatal. the counterbalance seems to be not that you would reload from loss, rather that you would reload because you realize you just cost yourelf another hour of progress. so how do we design a mechanic that makes the ai smarter/harder but doesnt overpunish risk vs reward?

1. quest/strategic/diplomatic goal sets? could provide timely reinforcement for completing objectives at the cost of the fleet to lessen the refleeting impact for correct actions.
2. refactoring the salvage options. currently these provide too much impetus into a defensive war that you must appear weak but remain strong to take advantage of. maybe refactor salvage to being a system resource that is capturable. this would help reduce the refleeting impact by letting you capture and recover the fleet that you lost. rather than harshly impacting the time lost on refleeting.
3. control points/golems maybe make these "hulks" indestrucible. they can be defeated but then either side would start repairing them to use them again. (ai would be recovering them to throw somewhere not necessarily at the player) could be another counterplot escalation like ai progress would be the superweapon factor(how many teams of recovery agents are added as the game progresses time/events). 20 minute timeframes would give you more tension on do you let the ai get something or move to strike it. (warnings from intel leaks to lure the human fleet?) also links to point 1. also this would lead to the ai not wanting to send them forward because they can be defeated and used by the player if calculations are wrong. this also would enforce the refleet and die issue with a few ai plots because they would see that as the time to send the golem spaceforce to kill you.
4. finding a way to prevent map gridlock. some maps became this refleeting slug because of chokepoints in the map layout. x and snake being a prime example. you have no way around the front so all the freed threat is always in your way. i like how nomads felt about providing temporary gateways around and providing strike points. if the ai could rebuild this would not have such a one sided effect in the players favor. though i do know some people have been wiped out by the asymmetric ai fleet cause by the same issue. this is plannable though, and at the heart of strategy and planned allocation in my opinion. i would like to see potential nomad links even without the hacking and using them as weapons options. i thought it was kinda a shame that in your map design you would plan the orbits but not use them to good effect like them having wormhole orbital configuration changes during gameplay. hopefuly this becomes a future plan.
5. set the ai response to fleet wipes to become launching cross-planet attacks to set a time limit but otherwise enfore the fatal response issue. potenially linked to idea 1 of providing a reward of metal of ships from surviving and allowing better back and forth pushes in escalating manner.

just a few thought i hope us help move forward to a better game in the end. or at least add more options to our game setups.


Offline TheDeadlyShoe

  • Full Member Mark II
  • ***
  • Posts: 194
Re: Fatal refleeting.
« Reply #69 on: September 30, 2016, 01:42:47 PM »
fwiw, i feel raiding is already great activity for the player to conduct during fleet downtime. the biggest problem is that raid starships are very specialized (not many ships can do badly what they do well)  and also very expensive themselves, lending them to suffer hard from refleeting problems. Especially if you lost them at the same time as your fleet.  Champions are also similar in that they are a downtime activity, but they arn't a core mechanic, so.

one problem with refleeting is that if you suffered massive defensive losses you are also rebuilding all of your defenses and that creates massive competition for metal.    Perhaps turrets shouldn't cost metal?

i guess this is where i go bluh bluh does metal even add anything to game that's 100% about your power (fuel) cap and your ship caps in terms of actually accomplishing things.  I guess, other than encouraging you to take your (soft, fleshy bits) out of the (chainsaw meatgrinder) rather than (spamming gratuitously).



The reason I pointed it out was because there's a certain threshold where the game is exciting because you're kept in suspense, just shy of being able to win or having lost, but aren't pressured by anything other than the fact that the win is juuust two moves away, if only you'd be allowed those two moves.

AIWC tries to pressure a loss by throwing a CPA or Exo at you and while, yes, there's the sudden excitement about "ooh, this might be the one that kills me!" when you survive you're not left in a state where you can make a move towards winning.  Instead you're rebuilding back towards a state where it'll happen all over again.  A constant cycle of "near-loss, rebuild" erodes the tension.

If we take the analogy to Jenga, a sudden CPA is like getting your turn only for your friends to switch on a fan, creating a sudden cross-breeze that makes it impossible to take out the next piece (or just outright knocks over the tower: you lose!).  Only if you manage it, then it's turned off for their turns.  At some point you're going to call someone an asshole and walk out.

i am unsure about your premise.  I've gotten whammed continuously by exos a few times, but even the most Brütal CPAs just don't chain hard enough.  Exos are mostly a bonus mechanic designed to balance humongous player bonuses like golems and spire, not a core mechanic; and neither exos nor CPAs are sprung upon the player suddenly. 

Offline x4000

  • Chris Park, Arcen Games Founder and Lead Designer
  • Administrator
  • Zenith Council Member Mark III
  • *****
  • Posts: 31,164
Re: Fatal refleeting.
« Reply #70 on: October 03, 2016, 02:13:42 PM »
Now on the doc:

Trickling Exogalactic Waves (Default Off)
Based on a suggestion from Draco18s, this is something that would allow for a different kind of play where you aren’t being hit with huge exogalactic waves that might overwhelm you all in one burst, but rather instead are facing that same set of threat in a more gradual, constant fashion.

It’s the difference between being hit in the face with a bucket of water or having it poured slowly through a tiny funnel onto your head.  Strategically and tactically this would have a lot of ramifications, and as kasnavada points out this could be very tricky to balance well: “A bit too weak and it's like the mechanic is not there, and too strong and you'd have to fall back”

This is not too complex to implement, so we’re making it a new option that is off by default.  Worst case, most people (or everyone) will continue using exo-waves like happened in Classic.  Best case, this works so well that we change this to being default on.
Have ideas or bug reports for one of our games?  Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline Draco18s

  • Resident Velociraptor
  • Core Member Mark V
  • *****
  • Posts: 4,072
Re: Fatal refleeting.
« Reply #71 on: October 03, 2016, 05:34:01 PM »
*Salute*
Sounds good.  We'll see how it works out (I'm sure I'll end up using it by default)