AI War Beta 5.080-5.082 "Bonus Round" Released!

[Hearteater]

You can Control-click Harvesters to build all harvesters regardless of enemy ships.  Still requires manual interaction though.

[Diazo]

Okay.

This is pretty much copy-paste from my AAR, but as it is feedback on the Zenith Siege Engines I figured it was worth a post here.

I am going to maintain that the siege engines are weird. My more recent sessions have been on 5.080 so I have the nerfed version at the moment and they are okay I guess, but weird.

On defense they rock as they can setup before the AI's forces arrive.

On offense they are more problematic. I can't figure out a use for them in my fleet ball as it moves too much so I'm using them on their own to take out guardposts.

This requires cloaking starship escorts to get them there though. Even then, they have to run away after a single salvo otherwise the AI's forces are able to close and eat them for lunch.

This makes them one of the most micro-intensive units in the game and I'm not sure they are worth the extra attention they require at the moment.

I'll see how they do after the buff in 5.081 but I do have a couple of mechanics tweaks that would make them more appealing, at least to me.

1) Due to their setup time and relative fragility, turn them into a one-shot artillery unit. Increase their setup and reload times, with boosted damage to compensate, and make them one-shot wonders. This crosses into the Zenith Bombard and the Artillery Golem however so maybe the overlap is too big.

1a) To compensate for how high-damage they are, make them self-destruct after a single (or two) shots. This is more a youngling thing though.

2) Keep the long setup time, but decrease the reload time significantly (adjusting DPS as appropriate). The logic behind this being you can better decide the balance point between doing damage and running away. More then once (in my current game) after the first salvo, the target has been sitting at 10% HP or less but I have to run away anyway because if I leave them stationary for 15 seconds the AI ships that are closing will eat them alive.

If the reload time was 4 or 5 seconds, I could make the decision of leaving them in and taking more losses but still getting that extra little bit of damage off to actually kill the target (while probably totally dying afterwards) or running away to keep my seiges alive and coming back to finish the target off later. Right now the long reload means I have to run away if that first salvo does not kill the target as the sieges will not survive for the 15 second reload.


Anyways, that's what I've found with sieges so far.

Anyone else figured out some way to use them on offense that does not require cloaker starships? They need escorts of some sort but putting them in the fleet ball means they never fire.

D.

[keith.lamothe]

2) Keep the long setup time, but decrease the reload time significantly (adjusting DPS as appropriate). The logic behind this being you can better decide the balance point between doing damage and running away. More then once (in my current game) after the first salvo, the target has been sitting at 10% HP or less but I have to run away anyway because if I leave them stationary for 15 seconds the AI ships that are closing will eat them alive.

FYI, 5.080 already decreased their reload time from 15 to 10 (maintaining dps).  I didn't want to decrease it too much, but 8 or even 7 might be ok.

Anyone else figured out some way to use them on offense that does not require cloaker starships? They need escorts of some sort but putting them in the fleet ball means they never fire.

I'm guessing shield bearers or riot starships (with shields) could cover them long enough.

[TechSY730]

2) Keep the long setup time, but decrease the reload time significantly (adjusting DPS as appropriate). The logic behind this being you can better decide the balance point between doing damage and running away. More then once (in my current game) after the first salvo, the target has been sitting at 10% HP or less but I have to run away anyway because if I leave them stationary for 15 seconds the AI ships that are closing will eat them alive.

I was thinking about the opposite. Reduce the setup time, but keep the long-ish reload time.
However, a slight tweak will be needed to prevent cheese, while moving or setting up, reloading stops (not resets, but does stop progressing)
(For this example, I will assume the setup time was reduced to 5 seconds, and the reload time stayed at 15 seconds). Without this tweak, this will effectively give them 10 seconds of non-consequential movement after every fire, as the 15 seconds will still recharge while moving, and during the last 5 seconds, if you can stay still, they will still be able to fire due to the 5 second setup.

EDIT: This basically makes them more Seige Tank like. Where the setup time is enough to make you think about placement since you need to be immobile to get the best out of them, but not long enough that you are afraid to move them if need be or the battlefield's location has shifted.

[Diazo]

2) Keep the long setup time, but decrease the reload time significantly (adjusting DPS as appropriate). The logic behind this being you can better decide the balance point between doing damage and running away. More then once (in my current game) after the first salvo, the target has been sitting at 10% HP or less but I have to run away anyway because if I leave them stationary for 15 seconds the AI ships that are closing will eat them alive.

FYI, 5.080 already decreased their reload time from 15 to 10 (maintaining dps).  I didn't want to decrease it too much, but 8 or even 7 might be ok.

I'm assuming setup time has to equal reload time then? My suggestion was assuming you could keep the 15 second setup time, but have a 5 second reload time once you started firing (and stayed stationary).

I see the change about the multipliers going back up in 5.081 which is nice, but I think most of the guardposts are hull types the siege engine does not get attack bonuses against and 99% of the targets I've killed so far have been guardposts. (Well, direct fire, I've killed several fleet ships with AOE so far.)

I'm hoping to get a nice long session in with them this evening as I should actually have free time so I'll actually watch them closer, the last little bit of game time I've been on the defensive thanks to multiple waves at once and the large amount of threat that raiding these Data Centers is producing.

I'll watch them closer when I'm able to take them on the offensive so hopefully I'm have something more concrete then 'I think...'.

D.

[Hearteater]

I wonder if Keith pulls his hair out when we complain about lack of need to micro and just blobbing everything, but when he adds a ship that requires micro we complain about needing to micro and not being able to just put it in our blob.

[TechSY730]

I wonder if Keith pulls his hair out when we complain about lack of need to micro and just blobbing everything, but when he adds a ship that requires micro we complain about needing to micro and not being able to just put it in our blob.

I think the complaint about "blobbing" was that micro overall didn't feel rewarding enough because of how the AI manages units and how the combat system works (among other reasons).
One unit built around more "micro" doesn't feel as well integrated into the system so it feels like it stands out (in a bad way), and doesn't solve the overall cause of the "blobbing over-incentive".

Or at least that is what I think the answer would be; I don't feel very strongly about the whole "are fleet tactics not rewarding enough" debate.


But yea, this sort of feedback must be frustrating. 

[keith.lamothe]

I'm assuming setup time has to equal reload time then?

Doesn't absolutely have to, but the mechanic was easier to implement that way.  I could probably make the "movement => reload = full-reload-time" thing into "movement => reload = max(5,current-reload-time)" or something like that.  Not sure I want to go that low on it.

I wonder if Keith pulls his hair out when we complain about lack of need to micro and just blobbing everything, but when he adds a ship that requires micro we complain about needing to micro and not being able to just put it in our blob.

Not frustrated, actually, but the temporal juxtaposition of the recent discussions on blobbing/micro and the reaction to this unit (which was designed that way before those discussions, incidentally) has been a cause of amusement

[chemical_art]

Just a question: Can forts be moved from ultra-heavy to structural?

I ask because I am glad to see the multipliers going back up for the siege gun. It should help make it so that it is more useful on sieges at the cost of being a deployable turret. If it's still too strong dps wise I'd get the base damage more but leave the multipliers alone.

But then I was thinking, it still is an awesome turret because it smashes ultra-heavy bad. So I was thinking what would you want to siege on offense? I came up with forts and shields. If both had the same type of hull (structural) then you could leave the 5 multipliers on everything it has a bonus against, but keep the 10x against structures which would seem the best for sieges anyway.

[Diazo]

Well, there's micro and then there's micro.

With my fleetball I hit '1' and right click to move, it kills stuff, I right-click it onto the next guardpost. (Note that I very rarely right-click a unit, I move my fleet close and let auto-target sort it out.)

With Siege Engines I have group 2 and group 3 both assigned to them. Group 2 is the Siege Engines and the Cloaker Starships, 3 is just the siege engines. I do this so I can power-down the sieges, losing them all to an autoattack against a random fighter sucks (and happens so freaking often.)

So, to use the sieges I have to:
1) Clear the Tachyon guardian on the warp point with other units.
2) Hit 3, K, to powerdown the sieges
3) Hit 2 and warp to the AI system, making sure the cloaker starship goes first.
4) Shiftclick-move around the tachyon guardians on the other warp points and then using ZX find the correct position so the sieges are in range.
5) Wait for the sieges to move and reach position.
6) Pause the game, hit 3K to power up the sieges and right-click the guard post.
7) Wait 15 seconds, watch the salvo get launched and while the shots are still enroute pause the game.
8) Hit 3K to power down the sieges, hit 2 to select the Cloakers aswell and issue the run away orders.
9) Unpause the game and pray the guardpost dies and hope the AI's response only kills 1 or 2 siege engines before the cloak timer runs out and the siege engines are protected again.
10) Repeat until enough siege have died that the single salvo they get off is not doing worthwhile damage and retreat to refleet.

I suppose I do need to say that for the record I'm not one of those who are complaining about the lack of micro. There's plenty of micro available, it is just a matter of where your personal micro vs. losses balance point is.

Note that I consider micro and anti-blob mechanics to be two very different things.

(Expect actual feedback after my game session tonight.)

D.

[orzelek]

resource collectors do autorebuild nowadays.. The problem is when the enemy has cloaked units afking on your planets.

You see all is well when said unit is not a chameleon. That one is cloaked but it's not - tachyon burst (from champion) is not decloaking it. So you have 1-2 of these left over on your planet and currently there is NO way to find them.

[keith.lamothe]

Update: 5.081 hotfix is out to fix some important bug and balance issues in 5.080, and make a number of other changes/fixes that should hopefully improve things overall.

[TechSY730]

resource collectors do autorebuild nowadays.. The problem is when the enemy has cloaked units afking on your planets.

You see all is well when said unit is not a chameleon. That one is cloaked but it's not - tachyon burst (from champion) is not decloaking it. So you have 1-2 of these left over on your planet and currently there is NO way to find them.

Sure there is. Look for some space debris that doesn't seem like it belongs there. Send a unit over to it, and that should bait it into attacking.

Unless the AI is putting its chamo units into stand down mode.


Actually, that is an interesting point. Right now, there is no counter to the chamo mechanic itself; if the unit is in stand down mode, you are pretty much stuck. Sure, you could try doing something like detonating a lightning warhead near it, but if that unit is AOE immune, that won't work, and even if it did that is some AIP cost. (Another reason why it would be nice for warheads to get an immunity to AOE immunity). In that case, your only hope is a nuke...and that will fail if the chamo unit is nuke immune too...

[TechSY730]

Wait, if I am reading the notes correctly, does this mean the counts for starships in waves will be much like how fleet ships are now (except with a pretty decent penalty)? Cool. It struck me as odd that if the wave computation can show that the AI can "afford" more than one starship of a type even with their low caps and large penalty for being a starship, they would get that count chopped to 1 anyways.

[keith.lamothe]

Wait, if I am reading the notes correctly, does this mean the counts for starships in waves will be much like how fleet ships are now (except with a pretty decent penalty)?

Not really, in that it basically just adds 1 starship, normally, and after that the wave calc would play around with it and then get brought back to 1 anyway.  But there are a few cases where multiple starships can happen, and now that won't be artificially prevented.