(Bonus Ships) Comparing Swarmers to Pseudo-Starships

[Bognor]

It's based on the boosted-unit's circle radius (called in the code and ref sheet "LoneUnitRadius").

Therefore based on my experiences, this is a significant advantage for lower cap armies.

I agree.  It's much easier to saturate a low cap fleetball.

There's no "LoneUnitRadius" in the export spreadsheet (the one you make via the in-game reference tables), but there is "ShipRadius".  I made the attached graph on the assumption that each ship consumes an amount of boost proportional to the square of its ShipRadius, which seems logical.  The data points represent all triangle ships, bonus ships, and player-available starships at Mark I.  The roughly flat trendline indicates that on average, a cap of high cap ships consumes just as much boost as a cap of low cap ships.

Or is boost consumption actually proportional to ShipRadius, rather than the square of ShipRadius?  If it is, you get a strongly positive trendline, consistent with Wingflier's experiences.  And you could easily flatten the trendline by making my initial assumption correct.

Edit: There's actually 78 data points on the graph, but many of the data points lie in little piles that look identical to single data points, which is also why the trendline appears to poorly fit the data.  The outlier at (96,614400) is the Neinzul Scapegoat.

Edit2: The second attachment, "Total cap ship radius by ship cap.jpg", is pertinent if boost consumption actually is proportional to ShipRadius.

Edit3: So finally I've found the explanation from Keith about how it's supposed to work.  Based on my reading of Keith's explanation, boost consumption is supposed to be proportional to the square of ShipRadius, ie the first graph, inconsistent with Wingflier's and Keith's experience.  Is it possible there's a bug causing boost consumption to actually be proportional to ShipRadius, ie behaving like the second graph?

[KDR_11k]

• Wrath Lance: need ships with individually high hit points --> starships/superweapons/pseudo-starships (definitely not high-cap ships)

Does that actually stand a chance? I know when I tried to deal with a WL it just hit everything with so much damage that high HP units got vaporized anyway and its penetration through low-HP units seems limited, it doesn't clean up Infilballs that quickly.

[chemical_art]

Does that actually stand a chance? I know when I tried to deal with a WL it just hit everything with so much damage that high HP units got vaporized anyway and its penetration through low-HP units seems limited, it doesn't clean up Infilballs that quickly.

For me, low cap units can survive glancing blows at longer range of the lance. In addition, if they can get into knife range, they can survive for much longer since they are easier to manage.

[keith.lamothe]

Not that this matters greatly to the overall picture, but:

• Wrath Lance: need ships with individually high hit points --> starships/superweapons/pseudo-starships (definitely not high-cap ships)

Does anything survive actual contact with those beams?  I'm thinking nothing in the fleet ship range has much chance of that.

Bear in mind those use the photon-lance mechanic, so they don't just hit everything on the line: they spend their strength against the first thing hit, then whatever's left goes against the next, etc.  So it doesn't necessarily do better against high-ship-count target-blobs.  A zenith-beam would do better against dense blobs, because everything it hits (up to the cap of 9 per shot) gets the full strength of the beam.

Anyway, the smaller the ship's unit radius, and the further away from the lance, the more chance the ships have of falling "between" the beams as they go by.  But if your fleet ball actually gets "caught" like that it's mostly done anyway, right?

[Diazo]

Not that this matters greatly to the overall picture, but:

• Wrath Lance: need ships with individually high hit points --> starships/superweapons/pseudo-starships (definitely not high-cap ships)

Does anything survive actual contact with those beams?  I'm thinking nothing in the fleet ship range has much chance of that.

Bear in mind those use the photon-lance mechanic, so they don't just hit everything on the line: they spend their strength against the first thing hit, then whatever's left goes against the next, etc.  So it doesn't necessarily do better against high-ship-count target-blobs.  A zenith-beam would do better against dense blobs, because everything it hits (up to the cap of 9 per shot) gets the full strength of the beam.

Anyway, the smaller the ship's unit radius, and the further away from the lance, the more chance the ships have of falling "between" the beams as they go by.  But if your fleet ball actually gets "caught" like that it's mostly done anyway, right?

If it's hit by the beam, nothing survives. I cheated and spawned all the spirecraft while testing this. A single pass of one beam killed all the shield bearers and most of the rest of the ships.

It is probably low, but each beam does at least 500million DPS.

In practice however it has targetting issues. When I was taking one out a few games ago I had teleport raiders. I'd send them in and 90% would die in the first pass, but the other 10% would take 5 or 6 passes to kill as the beam just 'missed' somehow. I think there are small wedges inside each beam that don't get hit, but these wedges move slightly on each pass.

D.

[keith.lamothe]

In practice however it has targetting issues. When I was taking one out a few games ago I had teleport raiders. I'd send them in and 90% would die in the first pass, but the other 10% would take 5 or 6 passes to kill as the beam just 'missed' somehow. I think there are small wedges inside each beam that don't get hit, but these wedges move slightly on each pass.

Yea, during development I had it doing a smooth rotation, but the truth is that our FInt (fixed-point-math) library uses fairly rough approximations for the few core trig functions it needs.  I need those functions to determine the endpoint of a beam of length X at angle Y.  The problem is that if I try to move Y smoothly through a range the results are really... well, wrong   If I use normal floating point math with the standard library trig functions it works great, but that would lead to desyncs in multiplayer because floating-point-math is not deterministic across different machines.

So it doesn't rotate smoothly but instead shifts a certain angle per "tick" and just pulses a few times at that angle before moving to the next.

Incidentally this makes it possible to not suffer 100% losses from one of these.  Without this quirk I really don't think anything would survive short of a massive FS fleet, and even that would be gutted afterward.

[KDR_11k]

Also I think its maximum number of hits per tick is limited, against my Infilball it only killed the first few layers on each turn.

[TechSY730]

I actually think someone might be able to use teleport ships to good effectiveness when taking on a wraith, if you had good micro, for similar reasons that fast ships can be useful in this fight.

[keith.lamothe]

Yea, the ability to get from point A to point B without crossing all the space between does have special relevance to the wrath lance

[Diazo]

Lol.

It's not that easy. When ships teleport on a right-click attack, they teleport to the center of the unit and then 'unstack' around it.

You can't see them behind the unit sprite, but the beams all emit from the center point of the wraith lance.

I gave my teleports that right-click attack order and they vanished. They did not even last a single frame to be drawn at the wraith lance, they just teleported away.

Expensive lesson when a full cap up to Mk IV is lost to learn that....

And as the wraith lance has AIP on death, I could not get them to attack without that right-click order.

I had to right-click move them next to the wraith lance so they were in range, and then right-click attack.

Still, teleporters are probably the unit that the easiest time with a wraith lance, but easiest isn't easy in this case.

D.

[TechSY730]

But back on topic, any idea what a good "increase" in cap stats the balance target for high cap stuff should have compared to the "normal" balance target that would be fair compensation for their downsides? Should the cap health be buffed more than cap DPS, DPS more, or roughtly equal amounts of both.

NOTE: the following is based off of mid cap scale game statistics. The "standard" cap is 96 in this case

Just an idea, for very high cap stuff (laser gattling cap, which is around 200), how about something like a 1.3x in cap health and a 1.5x in cap DPS adjustment to balance target compared to that of the normal balance target? Scale these multipliers towards 1 as needed for high cap ships that aren't as high cap as laser gattlings. (At around 110 to 120 or so, these multipliers become so negligible that you can pretty much treat them as 1 at that point)

To help with the mid cap vs low cap stuff, how about something like .85x in cap health and a .8x in cap DPS adjustment to balance target compared to that of the normal balance target for ships with a cap of, whatever the lowest fleet ship cap is right now (3 is it?). Scale these multipliers towards to 1 up for ships that aren't quite as low cap. (At around 60 to 70 cap or so, these multipliers become so negligible that you can pretty much treat them as 1 at that point)


Yes, these multipliers aren't very extreme, but that is intentional. You WILL start hitting balance issues they were like 2x or something like that (or .5x for the low cap stuff). 2x and .5x are actually rather severe modifiers for cap stats. Multipliers that high are almost certainly going to be overcompensation.

[Diazo]

My only concern with something like that is what happens when the AI swarms you with a couple thousand of these buffed swarmers?

I'm not sure tweaking the target cap HP and DPS numbers as the primary balance method is the way to go, especially as they already have slightly higher numbers don't they?

I'd rather see the economic costs get slashed, swarmers are cheap and weak and you will lose a lot of them.

Right now a Cap of Laser Gatling costs almost double what a Cap of Fighters do. (They cost half a missile frigate and a third of a bomber cap.)

I like to see a cap of swarmers cost at most what a cap of fighters do. Probably no lower as the super-cheap and super-expendable neinzul already have those roles.

The real issue with swarmers (high-cap units) is that so many game mechanics favor low-cap stuff. Not only the obvious like Eyes, but the fact that DPS drops faster as for high-cap units as opposed to low-cap units in combat.

That's not the numbers being out of whack though, that's the game mechanics screwing them over. Boosting their stats is not fixing the problem, it's just band-aiding it.

D.

[Hearteater]

The economic costs do need to go down, possibly energy too.  Per cap, they should be cheap to use, but not youngling cheap.  Their cap build time is fine, and I might even tweak it up a touch as it is already pretty close to younglings.  For damage, I think I'd favor extra damage via bonuses, rather than raw DPS.  The player is better placed to take advantage of that on both offense and defense, and that avoids the problem of what to do about 2000 Laser Gatlings.  I'd be ok with their health staying the same for now, or even coming down a little.

[TechSY730]

Scaling the target capwise economic costs down for high cap stuff would also work.

Good point with the bonuses. Many high cap ships have below average to really sad bonuses, which when combined with only average-ish base cap DPS, that is an issue. (There are exceptions, like the auto-canon minipod which has pretty strong bonuses, but lackluster bonus magnitudes for high cap stuff is the general trend)

[Wingflier]

My only concern with something like that is what happens when the AI swarms you with a couple thousand of these buffed swarmers?

As opposed to what? It swarming you with a 100 low-cap pseudo-starships?

I'll take the buffed Swarmers ANY day of the week.

With Riot Control II's I can paralyze them constantly. With Siege Starships I can stop huge swaths of them right in their tracks. With Champions there are all KINDS of modules to deal with things like this, such as HBCs modules or Siege modules. With many different kinds of bonus ships, such as Saboteurs, Railguns, or Fireflys, I can start a cascading chain reaction that kills most of them within less than a minute.

Not to mention that with many of the Golems I can kill them en masse (Black Widow, Wasp Golem), and if I have Botnet Golem, I can convert the entire swarm to my side in about 10 seconds.

All of those counters I listed would not work for low-cap pseudo-starships. I shudder to think how much we would have to buff high-cap bonus ships to be on par with, in the AI's hands, low-cap ships.

In the player's hands it's a bit different because the AI doesn't have nearly as many ways to shut down swarms as you do.