How to use attack points

[ncarson]

I searched the manual, the Guide, and even Google-steam, and no one seems to have any idea of what this is or how to use it. I'm trying to get my idiot admirals to group and attack all at one, rather than stream in and get killed one ship-fleet at a time.

[AKicebear]

I've also never figured it out - replying to follow any better answers.

[Colwolf77]

Attack points allow you to mark enemy targets before you begin a war. Once the war commences those fleets will try and destroy them then attack targets within its engagement range or other attack points.

[Nightskies]

There are three ways to use Attack Points.

First, the specified Attack Point for fleets centers their Engagement Range on the Attack Point when at war with the target. They will engage Threats and other potential Attack Points in that range, and any targets in the system they happen to be in at any given time. They may also be assigned specific missions by overall AI command (including in your empire, if enabled) that falls outside their normal, independent protocol. Fleets will actively reassess their mission and interrupt it with another. Its very common when giving manual orders to an automated fleet and flipping it right back to automation- something I'd call "fighting the automation", which I recommend avoiding.

Second, they will prioritize their designated Attack Point. That doesn't come close to a direct order to engage the target. It may not even engage the designated Attack Point at all, with reasons from lacking fuel and refueling locations to reach it from their current location, lacking a path to it at all (possible with poor exploration), assessing that the forces at the Attack Point are too powerful to take on, or that there are no worthy missions or targets at the Attack Point itself (such as a colony without stations with a 'no bombardment' policy).

But they will also coordinate with other fleets, sometimes incidentally, with the same designated Attack Point if they assess that they do not have the force themselves to handle it. I see fully automated invasion fleets time their assaults within seconds of arrival from completely different locations with far too much consistency to be coincidence- even in conjugation with manually controlled invasions (very satisfying, I must say). That's speaking of the invasion fleet's ability to drop troops in a fairly coordinated manner. They *can* fail, and it is absolutely frustrating to see something that *would* win under manual control to do so, but remember that the automation, under full automation, uses far more force than a manual invasion needs. Give it that.

I think that undesignated Attack Points do not get the same treatment, but am not sure about that. Side point - fewer assigned manual missions is better for automation. It can't anticipate your actions, but if it has enough time to act on them, it may do so. Naturally, that includes hostiles, if they have sufficient sensor coverage- and they do make lots of sensors. So should you, if relying on automation.

HOWEVER, the AI has questionable assessment of how powerful its target is relative to itself. This point has been touched on several times in discussion and in updates. I think it still does a mediocre job of bringing an appropriate amount of force, particularly regarding colonies with weapon batteries and defending fleets. This is probably the root of the issue, now capitalized with *much* stronger weapon batteries - even a single v1 Beam Battery can do a number against an early Shakturi fleet. Sensor coverage of the target helps though, *before* any moves are made. AI empires can still overwhelm well defended colonies in part because the AI still (and wisely) generally uses few specific designated Attack Points; few colonies can resist a third to half of the combined might of an empire's Attack Fleets.

Further, "time their assaults within seconds of arrival" does not mean the various fleets get to the target at the same time. For best effect, make a single big attack fleet with long range that others will sort of piggyback on, attacking the same Attack Point with it on the way.

THAT SAID, I must emphasize that fleets very actively reassess the situation. I have made many, many hostile fleets that I have eyes on (thanks to deep space spies) dead-stop and go home upon detecting the forces I've moved far ahead of them. Even so far as to promptly offer peace after the 'realization' that they don't stand a chance.

When it comes to automation, "fight to the death" generally works better for offensive fleets on a large scale. Retreating ships will make the fleet lose cohesion, and if the flagship of the fleet retreats, the others will follow. Defending and manual fleets work much better with more timid retreat settings.

Third method of using Attack Points, the list of targets can be used to prioritize them. Specific mission objectives can now be made (bombard, invade, etc). These are effectively Attack Points and Threats. The ones that qualify as Attack Points (stations and colonies) are more effectively engaged than Threats (fleets and ships, though sometimes stations as well). Threats can be targeted automatically when they are detected en-route to a station or colony, or are actively engaging them. AFAIK, threats will not be automatically directly targeted otherwise, even if selected as a priority (itchy memory is telling me from the back of the mind that there was an update note that says otherwise).

These are not at all handled like specified Attack Points, though the automation may assign fleets with these as their designated Attack Point. They are merely given higher priority, and will almost certainly be hit by individual fleets.

Despite efforts to address this in updates, the automation may divert an automated fleet to a target outside the designated Engagement Range, but this is separate from the automatic function of a fleet operating on its own instruction. Think of it like this: the AI 'commander of the fleet' has its own default instruction it will follow, but will obey direct instruction from the AI grand admiral that may not always respect the fleet's designated Engagement Range when selecting what fleet to attack a given target. It is supposed to *somewhat* respect Engagement Range, but not strictly (this is intentional, as specified in update notes).

Does that satisfactorily answer the question?