Poll: Worst Unit Of The (time interval) Award (IV)

[keith.lamothe]

In theory I could have it only count stationary non-snipers that are actually within range (say, range*1.2 to account for the fact that the AI needs some room to maneuver) of the relevant wormhole for the purposes of "do I go through?".  That would be some additional cpu cost but it would be on the AI thread and it's ok if it slows down a little (the "mistakes" that the AI makes are not typically caused by slow cycles, etc).

In that case, it could consider those turrets as having the full firepower, but it would probably still lead to some frustration at the threat-ball becoming a threat-wall.  On the other hand, maybe it just encourages not bottlenecking right on the wormhole unless your goal is literally preventing entry by all but an insane attacking force.

[TechSY730]

but it would probably still lead to some frustration at the threat-ball becoming a threat-wall.

That's why the overall adjustment to how much firepower ratio the AI waits for before attacking would be needed if how much the AI penalizes firepower of turrets is reduced.

In theory I could have it only count stationary non-snipers that are actually within range (say, range*1.2 to account for the fact that the AI needs some room to maneuver) of the relevant wormhole for the purposes of "do I go through?".  That would be some additional cpu cost but it would be on the AI thread and it's ok if it slows down a little (the "mistakes" that the AI makes are not typically caused by slow cycles, etc).

In that case, it could consider those turrets as having the full firepower

When I first posted my mantis issue, this sort of dynamic turret firepower adjustment came to mind. However, I now realize how finicky, and in the right hands, exploitable that would be. I think a properly chosen denominator for a better "rough approximation" of "turrets being stationary consideration" may work just as well in most cases.

Then again, this sort of dynamic lookup would allow configurations like Nodor described to still not hold off threat from coming through very much (though, maybe with a slight adjustment such they come through in slightly larger "batches"), allowing people to easily "bait" the AI with unusual turret configurations at the penalty of not having as good "alpha strike" damage or "critical point defensibility", but still keep the AI from sending in 5 ship "waves" into a very aggressive turret ball right at the wormhole they are entering.
This way, if people are willing to risk placing turrets such that the AI could possibly slip past them, the AI would be more willing to come in and try to do just that, allowing people to still easily "bait" small threat balls. But for those who place turrets in such a way that the AI cannot slip past them, the AI will consider that too and wait for a bigger buildup to stand a chance.

[Hearteater]

Too bad the AI doesn't tracking stuff over time, or you could have the 10% value scale upwards each time it tries to assault the planet and fails.  Reset it back to 10% once it breaks through the system.  That would mean it would start by being real brave but over time would wait for bigger and bigger forces.

[Volatar]

Too bad the AI doesn't tracking stuff over time, or you could have the 10% value scale upwards each time it tries to assault the planet and fails.  Reset it back to 10% once it breaks through the system.  That would mean it would start by being real brave but over time would wait for bigger and bigger forces.

That actually sounds like the best way to handle this. Makes the most sense for a force without scouting as well.

[TechSY730]

Too bad the AI doesn't tracking stuff over time, or you could have the 10% value scale upwards each time it tries to assault the planet and fails.  Reset it back to 10% once it breaks through the system.  That would mean it would start by being real brave but over time would wait for bigger and bigger forces.

That actually sounds like the best way to handle this. Makes the most sense for a force without scouting as well.

Yea, this seems like the best way to handle it without fiddly math.

How much coding work would it be for the AI to track state based on past information on a per-planet basis?
Also, would it violate one of the design objectives of the AI, not model the AI as a centralized intelligence but rather a distributed one, with a "distributed" knowledge base to?

[Hearteater]

A pretty straight forward setup could be each planet gets a single number initialized to 10.  Each time a wave or threat invades a human system that number is decreased by 1, to a minimum of 1.  Whenever a human Command Station is destroyed in a system is destroyed, the number is reset back to 10.

[keith.lamothe]

Giving the AI some persistent info isn't too hard to do, but I can't think of a concrete case where it would obviously be an improvement.  Rules like "increase the 'do I attack' threshold by 10% per attack-attempt-on-this-planet since the last time I blew up a human command station there" are very prone to manipulation by the player: just have a buffer system that you let them kill every time, stop them on the second planet in, and rebuild the buffer.  Or the other way around, if you specifically prefer large threat balls.

I think for a bit we tried a rule like that if there were enough failures it would just spawn an H/K to take care of business, but iirc that didn't go over so well

[TechSY730]

Giving the AI some persistent info isn't too hard to do, but I can't think of a concrete case where it would obviously be an improvement.  Rules like "increase the 'do I attack' threshold by 10% per attack-attempt-on-this-planet since the last time I blew up a human command station there" are very prone to manipulation by the player: just have a buffer system that you let them kill every time, stop them on the second planet in, and rebuild the buffer.  Or the other way around, if you specifically prefer large threat balls.

I think for a bit we tried a rule like that if there were enough failures it would just spawn an H/K to take care of business, but iirc that didn't go over so well

Hmm, so even though the "consider current state only" method may be more fiddly to get right, it is less subject to "mind games" from the players manipulating this information.

[TechSY730]

So far I have seen 4 possible things to do with turret firepower calculations by the AI
1. Keep as is (considering turrets only 10% as effective as normal) (The reasoning being that making the AI be stupid when considering turrets is less bad than trying to get them considering it correctly but not hampering player progress too much or introducing worse problems with a new system or cutoff)
2. Adjust the ratio (bump it up to like 25% or something)
3. Make the ratio dynamic based on current state of turret placements (like Keith's suggestion)
4. Make the ratio dynamic based on past success/failures of raids on the planet to be attacked

Of course, some combinations of the above may work (like it considers both the current and past states, and it has the "minimum ratio" bumped up some)
Also, if threat balls start turning into threat walls, some adjustment to the "how much firepower ratio do I need to enter" (particularly, lowering it in general) may be needed.

[Eternaly_Lost]

I think for a bit we tried a rule like that if there were enough failures it would just spawn an H/K to take care of business, but iirc that didn't go over so well

That makes me wonder, is there a way outside of the EXO-waves to get an H/K to spawn? Like the AI just decides to send one of them rather then a wave of ships? Or is that the only way to get them?

[keith.lamothe]

That makes me wonder, is there a way outside of the EXO-waves to get an H/K to spawn? Like the AI just decides to send one of them rather then a wave of ships? Or is that the only way to get them?

It's the only way   For most of the past year they were defined in the game but could never spawned.  Then someone asked if they could be included in exos.  We get strange requests, sometimes

[TechSY730]

Well, if my exo-wave like logic to determine standard wave composition idea for an AI modifier gets implemented, that could potentially be another way to see H/Ks, though it would be late game before standard waves would get strong enough to afford one.

[Nic]

I think for a bit we tried a rule like that if there were enough failures it would just spawn an H/K to take care of business, but iirc that didn't go over so well

That makes me wonder, is there a way outside of the EXO-waves to get an H/K to spawn? Like the AI just decides to send one of them rather then a wave of ships? Or is that the only way to get them?

With the new exo-like carrier logic, you can get H/Ks from sufficient amounts of unit deploying from multiple carriers. I think this only peaks out at mk1 H/Ks though.

[keith.lamothe]

With the new exo-like carrier logic, you can get H/Ks from sufficient amounts of unit deploying from multiple carriers. I think this only peaks out at mk1 H/Ks though.

Oh, right, forgot about the carrier logic.  Have you actually seen an H/K from one of those?  I wasn't aware that 1000 of any ship that can fit in a carrier would add up to that much.

[TechSY730]

With the new exo-like carrier logic, you can get H/Ks from sufficient amounts of unit deploying from multiple carriers. I think this only peaks out at mk1 H/Ks though.

Oh, right, forgot about the carrier logic.  Have you actually seen an H/K from one of those?  I wasn't aware that 1000 of any ship that can fit in a carrier would add up to that much.

If you managed to get a few hundred mother ships in a carrier, you could probably make it spawn a few H/K if you pop it early.

Then again, if you managed to get a few hundred motherships to spawn, you probably got bigger issues then worrying about when to pop carriers.