Current balance of starships in exo waves?

[keith.lamothe]

Mind you, with manufactories gone, any plasma siege that gets in range of the home command FF is going to damage the home command, you don't have 25+ structures under your home command FF any more.

Yea, I guess building infinite science ships to tank would be a bit expensive   Though if you had mkI fighters on loop-build they'd probably do just as well as the manufactories if not better.

edit: Does the relative strength include health somehow?

Basically all relative strength is meant to be is "how does a single unit of this type compare with a single unit of a mkI fleet ship  standard-cap type?"  (standard cap being 24/48/96/192 depending on cap scale).  This doesn't count bonus types as stronger than triangle types (in the end the triangle types have gotten buffed over time to the point that I'm not even sure it's true that the bonus types are generally stronger, but I'm not worried about that), so a mkI fighter and a mkI IRE are considered equivalent, but a mkI armor ship is considered a bit stronger since its cap is a bit lower.  And it only counts mark levels linearly despite the fact that the linear increase in both dps and health creates a more-than-linear improvement; that's fine with me because the escalation works well.

So it doesn't try to factor in dps or health or whatever.  It's just an overall "how strong is this thing?".  For fleet ships, starships, and so on it's generally just computed as ((cap/192)*mark).  For other things that don't really have a cap per se it's just an override number I've put in for it.  Or just the default of "mark level" if I haven't tried to give it a strength number.

[Diazo]

Ah. Hmmm.

Unit caps, right.

If we are standardizing on the relative strength thing, would you be open to changing it to a basing it (roughly) on a formula?

If not, I'd suggest leaving the FS exos calculated as they are and just adjust the numbers manually for the ships that are out of whack (really just the Zenith and Spire I think). If you are just entering override numbers you are doing it manually anyway, just on a different scale depending on which type of exo it is.

Where I'm going with this is that by just manually putting the RelativeStrength numbers in, you get situations like there where the RelativeStrength costs are (in my opinion) out of whack. But as the are entered manually, it is purely based on a person's opinion what the RelativeStrength value for a unit should be.

For instance, I prefer to let ships into a system and slow them while letting my long range units pound on them. That means radar dampening gives me fits and so in my personal opinion, Radar Dampening should result in a higher RelativeStrength value then to someone who defends the warp point with tractor beams and short range turrets who is not (or less) affected by the Radar Dampening.

If you are looking at redoing the FS exos to relative strength, I'd like to see a formula of some sort used as a starting point. It would require at least a rough agreement on what should go into it however and would have a ManualBonus variable that would be set manually for non-numerical abilities like Attack Boosts or the Shield Bearer.

IE:

(UnitDPS * 0.8 ) + (UnitDPS * 0.2 * AttackMultiplier) + ((Health + (Armor * 10)) * 0.001

Then all divided by unit cap and then multiplied by the ManualBonus so:

((UnitDPS * 0.8 ) + (UnitDPS * 0.2 * AttackMultiplier) + ((Health + (Armor * 10)) * 0.001)  / UnitCap * ManualBonus

As a note, that comes out to a RelativeStrength of 1 for a Fighter Mk I. (With a ManualBonus of 1, or 0%)

It also comes out to a RelativeStrength of 7131 for a Zenith Starship Mk I, so it isn't workable as is, that high health really throws things off.

It gives an idea of what I'm talking about though.

D.

[keith.lamothe]

If we are standardizing on the relative strength thing, would you be open to changing it to a basing it (roughly) on a formula?

I'm not dead-set-opposed, but I don't think it's necessary:

For example, if two bonus fleet ship types with a standard cap size are so different that player consensus would rank one of them as a 0.5 (at mkI) and the other as a 1.5 (at mkI), then wouldn't it simply be a case that the two types are way out of balance?

I've used formulas when establishing dps/health numbers to try to bring things into balance (largely, it worked for setting a baseline), but I wouldn't want the game to be literally telling itself "well, this unit type is seriously UP, so I won't charge the AI as much for it, but this one over here is really OP so I'll charge double...", etc.

See what I mean?

For instance, I prefer to let ships into a system and slow them while letting my long range units pound on them. That means radar dampening gives me fits and so in my personal opinion, Radar Dampening should result in a higher RelativeStrength value then to someone who defends the warp point with tractor beams and short range turrets who is not (or less) affected by the Radar Dampening.

But look at it this way: why should a particular potential defensive situation affect the value of the attacking ship?  That is, why should the AI pay more for ships that will work against your defenses as-they-stand and less for those that won't work as well in that case?

I'd think that, instead, it gets them according to their basic utility (which should be similar across caps of the same mark for human-buildable units, though starship did just move to more of a "half cap" balance point; still seeing if that's true or not), and if you think you can respond with tactics that are disproportionately effective against that particular composition... well, good for you   But sometimes the AI will hit you with a composition disproportionately effective against your tactics.

Are we on the same page, or talking past each other?

[Diazo]

I think we are kind of half-way on the same page?

I was going for something with much more fine control over the RelativeStrength then it looks like you are intending to use.

I understand where you are coming from, my intent with the formula was to prevent the 'opinion effect' where everyone is going to have their own opinion on what the RelativeStrength should be for a unit as it sounds like a lot of units have their RelativeStrength manually set.

Going with that, my only thing would be I'd like to see the Zenith and Spire starships get a slightly higher relative strength as compared to the other starships.

Their extremely high individual health warrants it in my opinion (other opinions will vary of course).


Erm, looking at your formula, shouldn't that be (192/cap)*Mark ?

That would give fighters a RelativeStrength of 1 and a Plasma Siege Mk I a RelativeStrength of 96.

D.

[keith.lamothe]

I understand where you are coming from, my intent with the formula was to prevent the 'opinion effect' where everyone is going to have their own opinion on what the RelativeStrength should be for a unit as it sounds like a lot of units have their RelativeStrength manually set.

Actually only a few of them are manually set.  Basically everything on that list that's human-buildable is just (192/cap)*Mark (thank for the correction; I should not be trusted with math or heavy machinery).  The only time I set it manually is if there's no meaningful cap in the game, like with guardians and golems.  And with H/Ks, if I wanted them to have a meaningful strength value.

Going with that, my only thing would be I'd like to see the Zenith and Spire starships get a slightly higher relative strength as compared to the other starships.

Their extremely high individual health warrants it in my opinion (other opinions will vary of course).

To that my question is: shouldn't their health just be lower, then?  Otherwise, doesn't their need for a higher value aren't they overpowered compared to the other starships?

Or are you saying that "in the role of an AI attacker..." they're better than the others, whereas in the role of a human defender/attacker (or AI defender, for that matter), they are basically on par with the rest?

[Diazo]

Going with that, my only thing would be I'd like to see the Zenith and Spire starships get a slightly higher relative strength as compared to the other starships.

Their extremely high individual health warrants it in my opinion (other opinions will vary of course).

To that my question is: shouldn't their health just be lower, then?  Otherwise, doesn't their need for a higher value aren't they overpowered compared to the other starships?

Or are you saying that "in the role of an AI attacker..." they're better than the others, whereas in the role of a human defender/attacker (or AI defender, for that matter), they are basically on par with the rest?

I'm saying that their high health makes them excellent deep strikers to allow them to bypass defenses and kill their target. Therefore in the exo-wave scenario when the AI ships have a definite target they will not deviate from, that plays to the strength of the Zenith and Spire SS, that of having very high health.

In general, because AI ships do not get "Kill this target" orders, or if they do it's on a random unit, not a "High-value" unit that the exo waves pick from, Spire and Zenith SS are fine most of the time.

However, as exo waves composition is picked differently from other AI attacks, can we not account for the fact that exo waves play to the strengths of the ship in this specific case?

My current impression is that 10 Spire SS is a lot more dangerous then 10 Flagships and 10 Plasma Siege in terms of killing their deep strike target.

The 10 Flag and 10 Plasma will rip up your defensive line a lot more then 10 Spire, but the 10 Spire are a much larger threat to that Adv Fac in the system.

Again, this is opinion from the limited playtime I've had with the new starships. I'll see if I can actually back this up this weekend when I get playtime in.

D.

[keith.lamothe]

Hmm, yea, I do see that some things will be specifically more dangerous as an exo attack unit, regardless of how well the overall balance is.  I could simply add a 1.5x factor or something for use when pricing them as exos.  Much as I have one for pricing certain units in buying them for standard waves (specifically, the bomber is more expensive, and the fighter less so).

[Diazo]

That's what I was thinking of.

Having had 5 minutes to think on it I need to add a caveat that these impressions are from early game FS exo-waves when I do not have sufficient firepower to kill Zenith and Spire in a couple salvos.

Later game once I actually get enough units out that I can kill them in the time it takes them to cross my defensive line my opinion might change.

I'm specifically thinking of one case where the last survivor of the exo-wave was a high mark Zenith SS, it took all the defenses I had left in my chokepoint system 3 or 4 salvos to kill it.

It was up against my FF in the middle of my turrets so the radar dampening should not have played a big part in this.

D.

[keith.lamothe]

In that case for those units I'd still want the internal "strength" number to be whatever it would be for general purpose (which really should just be the standard based on cap and mark unless it has no meaningful cap or it's balanced on a different scale like a fortress), and a separate number for "exo suitableness".  Not suitableness specifically relating to a kind of _opposition_ (near or far defenses, etc), but specifically relating to a kind of use by the owner of the unit.  Which indirectly includes the forms of opposition such a use can encounter.

[Diazo]

Well, at this point I'm willing to say I don't see changing over to relative strengths being an issue. Most of the outliers should have their manual RelativeStrengths set from the Golem and Spirecraft exo waves already right?

It would just require some math to convert the exo-points that is currently used into the RelativeStrength values and that is just number crunching.

Then playtesting feedback to see if any units require manual RelativeStrength settings.

D.

[TechSY730]

If you do retire specific exo-costs and shift them to relative strengths instead to be the costs, make sure you set ones that don't have a relative strength set yet. Especially H/Ks (imagine the AI spamming H/Ks as freely as it does laser gattlings.  Actually, don't imagine it; it's too terrifying.)

Also, did the Z Bombard get it's relative strength increased a bit like the recent "In AI hands" nerf in the patch notes said it did?

[keith.lamothe]

If you do retire specific exo-costs and shift them to relative strengths instead to be the costs, make sure you set ones that don't have a relative strength set yet.

Um, yea.  Otherwise... yea, not thinking about otherwise.

That's actually one of the things I'm asking about here, though: are the relative strengths on the AI-only items reasonable?  Obviously the H/K doesn't even have them and needs values; what should they be?  Just exactly how many mkI fighters is a mkI H/K worth (ignore current exo costs) ?

But also the guardians and the golems and spirecraft (which aren't AI-only but don't have normal caps to compute from, etc).

Basically at least a sanity check on the whole list's strength values would be good.

Also, did the Z Bombard get it's relative strength increased a bit like the recent "In AI hands" nerf in the patch notes said it did?

I actually gave it a 1.5x multiplier on its relative strength so that the AI pays extra for it everywhere, yes.  That's more to do with its extreme range combined with its alpha strike than the raw health/dps numbers being off, though.

[TechSY730]

That's actually one of the things I'm asking about here, though: are the relative strengths on the AI-only items reasonable?  Obviously the H/K doesn't even have them and needs values; what should they be?  Just exactly how many mkI fighters is a mkI H/K worth (ignore current exo costs) ?

But also the guardians and the golems and spirecraft (which aren't AI-only but don't have normal caps to compute from, etc).

Basically at least a sanity check on the whole list's strength values would be good.

Hard to say. Have not looked at the list extensively for most of these items.

For the H/Ks, maybe have the Mk. V H/K be 4/5ths the cost of the mothership? And scale from there?

Also, did the Z Bombard get it's relative strength increased a bit like the recent "In AI hands" nerf in the patch notes said it did?

I actually gave it a 1.5x multiplier on its relative strength so that the AI pays extra for it everywhere, yes.  That's more to do with its extreme range combined with its alpha strike than the raw health/dps numbers being off, though.

Wait, so relative strength also plays a role in number of units of that type they get in normal waves? I don't recall seeing that step in the logs (though I could of simply missed it).

[keith.lamothe]

Wait, so relative strength also plays a role in number of units of that type they get in normal waves? I don't recall seeing that step in the logs (though I could of simply missed it).

Yea, the stat actually came from that usage when we kept hammering out situations where waves weren't adjusting properly for the ship caps of the unit in question.  That's why strength is inversely related to cap.

[Toranth]

If you do retire specific exo-costs and shift them to relative strengths instead to be the costs, make sure you set ones that don't have a relative strength set yet.

Um, yea.  Otherwise... yea, not thinking about otherwise.

That's actually one of the things I'm asking about here, though: are the relative strengths on the AI-only items reasonable?  Obviously the H/K doesn't even have them and needs values; what should they be?  Just exactly how many mkI fighters is a mkI H/K worth (ignore current exo costs) ?

Well, in what sense?

An H/K Mk I fires 25 shots of 520,000 damage each every 3 seconds, and has 100,000,000 HP.
A Fighter Mk I fires 1 shot of 4,080 damage every 4 seconds, and has 165,000 HP.

a)  To do equivalent damage, it takes 4,248 fighters to match the DPS of an H/K.
b)  To do as much damage in a single salvo, it takes 3,187 fighters.
c)  If you had Fighters vs H/K 1-by-1 until the fighters killed the H/K, it would take 24,510 fighters.
d)  If you had a group of fighters vs 1 H/K, it would take a group of 1,093 fighters to get the kill.

Which of these numbers is more indicitive of the usefulness of one unit vs another?
I'd think either a) or d) would be the most useful, depending on your viewpoint.
Is your idea to cram as much damage into an Exo as possible?  Or to 'match' the most 'equivalent' number of ships?