Current balance of starships in exo waves?

[keith.lamothe]

All we need now is for the game to decide rather then spawning a single Hunter Killer Mk5. It spawns 262144 Laser Gattling MK1.

Yea, I can see the logs now:

Code: [Select]

Picking 262144 Laser Gatling I
Deploying...
Over ship limit
Creating 263 carriers
Still over ship limit
Putting 263 carriers into another carrier


Somehow I get the feeling those would win against the Hunter Killer and be a lot harder to kill then the hunter killer would.

Lightning warheads aside, I believe you would be right.

That's one iffy thing about the pricing system it currently uses: as the budgets go up and the ship-count-per-battlegroup rule pushes it "up the scale" so it has to buy the more expensive stuff just to fit it all, I don't think it's really getting a very good deal on some of those "upgrades" compared to the little stuff.

Realistically speaking it's never getting to buy many (if any) laser gatling Is since even the starting budgets have enough of a "push up" effect to get it into the higher tiers even for the pickets.  But I think for example that 32768 CoreMicroFighter's would be a lot more dangerous than said HK.  Or even that 128 LightStarship's (the mkI flagship) would be more dangerous than a 16184-point golem.

So to some extent I think if you plotted "average danger" over "exo budget" what you'd get would be kind of a crazy-curve where it actually gets easier in some segments and then stays mostly horizontal for a while before gradually picking up into an upward curve as it hits the top-end units and starts resorting to "the same, just moar".


The funny thing is that, play-experience wise, this seems to work out just fine

[Cinth]

What if those small "swarmer" fleet ships weren't buy one at a time.  Say maybe they come in groups of 10 (some fitting number) and have the batch priced accordingly?

This assumes we want to actually see them.

[keith.lamothe]

What if those small "swarmer" fleet ships weren't buy one at a time.  Say maybe they come in groups of 10 (some fitting number) and have the batch priced accordingly?

This assumes we want to actually see them.

If they still spawn as individual units that would defeat the purpose of the count rule: not incinerating your CPU.

I can understand why that consideration would not occur to you, however

[Marmu23]

 I find that exos are currently harder at exo level 4 and AI level 8 than last year when i did a challenge game at AI level 10. The combo of fallen spire + spirecraft + golems seems to scale harder and faster than it did back then. Hopefully once gravity properly applies to exos things will be better (see the 6.13 thread for a save showing half the exos ignoring gravity), but perhaps slightly decreasing the size of exos would be a good idea.

 For solo play I guess it's ok since I can put the game at +10 and build super massive chokepoints over a couple hours while i do other things but for group play we never play with them since it means so much time is spent farming resources for building defenses rather than actually playing.

[Hearteater]

What if those small "swarmer" fleet ships weren't buy one at a time.  Say maybe they come in groups of 10 (some fitting number) and have the batch priced accordingly?

This assumes we want to actually see them.

If they still spawn as individual units that would defeat the purpose of the count rule: not incinerating your CPU.

I can understand why that consideration would not occur to you, however

If only there were some kind of stacking unit...sorry 

[keith.lamothe]

If only there were some kind of stacking unit...sorry 

Oh yes, you came to mind immediately as I started writing that post

Basically it does unit-stack, preferring mkV laser gatlings instead of mkIs.  I don't think a mkV is any different than a 5x stack in your example.  Though in this case a mkI costs 1 and a mkV costs 16.  Which is one area where the relative-strength value is better: it's right on 5x for the mkV version.

[Hearteater]

Sounds like its time for Mark X ships!

[keith.lamothe]

Sounds like its time for Mark X ships!

We call them "guardians"   And so does the exo logic, really.

[Eternaly_Lost]

Sounds like its time for Mark X ships!

We call them "guardians"   And so does the exo logic, really.

And then those things come in MKV, and several dozen Carrers filled to the brim with them (also known as the Special Forces) makes the Fallen Spire fleet just roll over and beg for mercy. With my CPU having already done it a long time ago.

[Diazo]

Okay.

Plasma Siege Mk V: Single Target: 450,000DPS, max AOE: 1,153,125DPS (25 under FF). 50million HP, 2048 ExoPoints
Spire Starship Mk V: 1,800,000DPS when full beam duration, such as an FF, 120million HP, 1536 ExoPoints

That seems backwards, the Spire has well over twice the DPS (depending on how much AOE you give the plasma siege) and over twice the health, but only costs 75% as much for the exo?

If you cut the Spire Starships stats in half, so a cap of 1 -> 2, it is an awfully lot closer to what would seem appropriate.

However:

Spirecraft Siege Tower Mk II: 106,667 DPS, 8million HP, 2048 ExoPoints.

How is that Siege Tower remotely close to the Plasma Siege Mk V in threat that it costs the same amount of ExoPoints, much less the Spire Starship Mk V that it costs 25% more then?

Is there come reason different 'tiers' of the parts of the exo-wave, (flagship, escorts, wingmen) use different point values that would explain that?

I don't have time to compare more (bedtime here), but it looks like there is wonky stuff going on to me.

D.

edit: erm, did I miss something? Your "IndividualMaxDPS" column has the Plasma Siege doing significantly more DPS (6,750,000) then the Spire Starship (5,400,000) when my own calculations have that DPS going the other way. (Spire does more then the Plasma Siege.)

[chemical_art]

Okay.

Plasma Siege Mk V: Single Target: 450,000DPS, max AOE: 1,153,125DPS (25 under FF). 50million HP, 2048 ExoPoints
Spire Starship Mk V: 1,800,000DPS when full beam duration, such as an FF, 120million HP, 1536 ExoPoints

That seems backwards, the Spire has well over twice the DPS (depending on how much AOE you give the plasma siege) and over twice the health, but only costs 75% as much for the exo?

If you cut the Spire Starships stats in half, so a cap of 1 -> 2, it is an awfully lot closer to what would seem appropriate.

However:

Spirecraft Siege Tower Mk II: 106,667 DPS, 8million HP, 2048 ExoPoints.

How is that Siege Tower remotely close to the Plasma Siege Mk V in threat that it costs the same amount of ExoPoints, much less the Spire Starship Mk V that it costs 25% more then?

Is there come reason different 'tiers' of the parts of the exo-wave, (flagship, escorts, wingmen) use different point values that would explain that?

I don't have time to compare more (bedtime here), but it looks like there is wonky stuff going on to me.

D.

To be fair, it has been a long time since exo-costs have been examined, and even then, they have never been thoroughly examined.

Not to say changes are good for it; it can somewhat "average out" costs, making it so that the cases of "hoping for a good exo-roll" less needed.

[Cinth]

What if those small "swarmer" fleet ships weren't buy one at a time.  Say maybe they come in groups of 10 (some fitting number) and have the batch priced accordingly?

This assumes we want to actually see them.

If they still spawn as individual units that would defeat the purpose of the count rule: not incinerating your CPU.

I can understand why that consideration would not occur to you, however

I don't think it would qualify for the smallest exo my games produce

It might work for ai difficulty below 740 of early exos though.  I'm not one to crunch numbers

Anyway, just an idea I thought I would throw out there.

Actually.... what if the ai bought sets of ships instead of individuals... buys a b c ship types for x points as a packaged deal.  Not something quick and easy but maybe worthwhile?

[Diazo]

For immediate changes, I'm thinking limit it to tweaking the exo-points cost of the starships that just got overhauled.

I don't want to go through and redo everything, overall the balance is correct (or close enough).

We are not limited to using existing values correct? I could add a 3072 points 'tier'?

If yes, I'm going to mull this over, then look at posting a list of suggested changes tonight after work to solicit feedback.

(From your earlier posts, it sounds like you are open to tweaking the exo-costs Keith.)

D.

[keith.lamothe]

Okay.

Plasma Siege Mk V: Single Target: 450,000DPS, max AOE: 1,153,125DPS (25 under FF). 50million HP, 2048 ExoPoints
Spire Starship Mk V: 1,800,000DPS when full beam duration, such as an FF, 120million HP, 1536 ExoPoints

That seems backwards, the Spire has well over twice the DPS (depending on how much AOE you give the plasma siege) and over twice the health, but only costs 75% as much for the exo?

They might do better in the same tier instead of the spire being one lower, but remember how much people fussed that the spire starship's real-world DPS was so much worse than other units due to the photon lance's mechanic (i.e. if either you or your target are moving relative to one another, the beam does much less real damage)?  So, what's it gonna be: either the photon lances are a finicky mechanic and need higher dps to give equivalent value, or they don't?

Also, I'd say a plasma siege has a much greater damage of doing something unpleasant to a human player, due to its ability to damage stuff that is legitimately protected by an FF.  Spires can just punch through it eventually, but generally heavy human defenses don't understand this "eventually" concept.

All that said, I don't mind changing the costs.

Spirecraft Siege Tower Mk II: 106,667 DPS, 8million HP, 2048 ExoPoints.

How is that Siege Tower remotely close to the Plasma Siege Mk V in threat that it costs the same amount of ExoPoints, much less the Spire Starship Mk V that it costs 25% more then?

Oh, I think you're absolutely right on that one.  These numbers just haven't seen attention basically ever   Kinda funny that it's worked out well despite obvious issues like this (not that the Spire V even existed until recently, but there are other mismatches).

Is there come reason different 'tiers' of the parts of the exo-wave, (flagship, escorts, wingmen) use different point values that would explain that?

When I say "Tier" I mean one of the following:
Drone
Cutter
Corvette
Frigate
Destroyer
Cruiser
Battleship
Dreadnought
Mauler
Planetoid

And a sub-tier of "Light", "Medium", or "Heavy".

For reference, here are the costs by tier:

Light	Drone	1
Medium	Drone	2
Heavy	Drone	3
Light	Cutter	4
Medium	Cutter	6
Heavy	Cutter	8
Light	Corvette	16
Medium	Corvette	24
Heavy	Corvette	32
Light	Frigate	64
Medium	Frigate	96
Heavy	Frigate	128
Light	Destroyer	256
Medium	Destroyer	384
Heavy	Destroyer	512
Light	Cruiser	1024
Medium	Cruiser	1536
Heavy	Cruiser	2048
Light	Battleship	4096
Medium	Battleship	6144
Heavy	Battleship	8192
Light	Dreadnought	16384
Medium	Dreadnought	24576
Heavy	Dreadnought	32768
Light	Mauler	65536
Medium	Mauler	98304
Heavy	Mauler	131702
Light	Planetoid	262144
Medium	Planetoid	393216
Heavy	Planetoid	524288

I don't have time to compare more (bedtime here), but it looks like there is wonky stuff going on to me.

Certainly   It's worked well enough in the past, but I figured it was time to haul it out and make it sane.

edit: erm, did I miss something? Your "IndividualMaxDPS" column has the Plasma Siege doing significantly more DPS (6,750,000) then the Spire Starship (5,400,000) when my own calculations have that DPS going the other way. (Spire does more then the Plasma Siege.)

May have something to do with how we're computing the AOE damage.  It's possible the I wrote the reference computation wrong in the case of the siege plasma and it's over-counting it, but remember that it gets 100% damage vs the main target, 12.5% (iirc) against X (20? I forget) nearby targets, and ALSO extra damage against stuff under a forcefield that it hits either of the first two components.

For immediate changes, I'm thinking limit it to tweaking the exo-points cost of the starships that just got overhauled.

I don't want to go through and redo everything, overall the balance is correct (or close enough).

We are not limited to using existing values correct? I could add a 3072 points 'tier'?

I'd have to add an override for a specific value rather than just "use the tier's value" but it could be done.

What I'd rather do, however, is get the "relative strength" metric to handle all exo-eligible types in a reasonable fashion and just convert exos over to using that.  I'd really like to move to only having one internal "how much is an individual ship of this type worth?" metric.

(From your earlier posts, it sounds like you are open to tweaking the exo-costs Keith.)

Absolutely   I'm up for a complete overhaul of these numbers, basically, but I think the results would be best if we hashed it out here rather than me trying to thinker with it by myself.

[Diazo]

Alright.

Thanks for those answers.

Also, I'd say a plasma siege has a much greater damage of doing something unpleasant to a human player, due to its ability to damage stuff that is legitimately protected by an FF.  Spires can just punch through it eventually, but generally heavy human defenses don't understand this "eventually" concept.

The thing is, with their health the spires now reach their target and go stationary, so the beam gets full DPS on their target. On 6.012 I was having several spires reach my home command FFs and I was watching 300million HP of FFs vanish as the full salvos were landing.

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.

I'll mull on the numbers but anything serious will have to wait until I get home and can break the math out.

D.

edit: Does the relative strength include health somehow? One of the things that has made the new starships more dangerous is that their new health allows them to survive long enough to reach their target. It is not as important as DPS, but it does need to be a factor.

edit the 2nd: That might have been 6.011 that I was watching the FF just melt, not at home to check. Either way, the new starships are scary! 

edit the 3rd:

What I'd rather do, however, is get the "relative strength" metric to handle all exo-eligible types in a reasonable fashion and just convert exos over to using that.  I'd really like to move to only having one internal "how much is an individual ship of this type worth?" metric.

This makes sense to me. Are you willing to share the calculation though? And how granular is it? I'm looking at the RelativeStrength column on the chart on the last page and I'm a bit confused on what it is doing.