I was thinking, instead of making the fleets equal in production cost, make them equal in MP. Then run another comparison, both in viability of each fleet and whether you can get more glass cannon fleets/replacement ships for losses. If the GCD is 1/4 the MP of the Dreadnaught, but cheap enough to build 5-6 of them for the cost of a single dreadnaught, then it wouldn't matter if you lost 1-2 per battle to take out the dreadnaught.



Edit: Also, in the previous comparisons the dreadnaught fleets had smaller ships to fill up the CP cap, thus the GCD fleets had more targets. I think the better comparison would be to make the dreadnaught fleet be just dreadnaughts even if there is some CP left over. Then just make the GCD fleets with the same CP limit. So if your CP is at 19 that's only 16 CP worth of Dreadnaughts for most races, so the GDC fleet should only be 16 CP. That way it's 4 GCD per Dread, or in the case of the 3 CP Dreads, 3 GCD per.

GCD fleets aren't particularly interested in heroes.

for heroes, everytime a glass cannon fleet is destroyed, its hero is out of action. also glass cannon fleeet hero isnt really going to have defense traits and his constant injuries mean he might be a lower level. just some thoughts.

Durath wrote:
But with that in mind you can test a theoretical maximum for both sides and get an idea of what is going on at an equal tech snapshot somewhere in the middle (there are problems with this premise namely that the hp bonuses change).

Not really. The theoretical maximum damage for all races is much too high for any practical use. A better procedure would be to test at 3-5 stages (completely arbitrary numbers, pick one that makes you happy). In situations that mimic turns: 50, 100, 150, 200, and 300 perhaps?

I think you have to consider heroes. It is ridiculously easy to get 1 hero to level 25 even if you're using micromanagers (1).

How many turns roughly? What are you going to do before then? This seems like a single player experience idea, and as we both know, the AI is bad.

Your GCD deck is harmony so that means you've been fighting the entire game without heroes. Isn't that a little unfair to everyone playing every other race?

I also don't get the Harmony Racial bonuses. If you'd like to propose a valid comparison, go for it.

I think you have to consider the best case scenario pilot/adventurer heroes and very bleh heroes as well.

I don't think it will make much of a difference, but feel free to propose an even comparison.

What about the Emergency Shelter equalizer? You have to pick a counter for what I'm going to do. Without looking at the cards I know weapon overclock counters offensive retreat. I'm pretty sure Emergency Shelter is a defense and is countered as such. That means you can't counter both at the same time. So that means I can offensive retreat when I see missiles or I can hit Emergency Shelter. Either option should save me from your missiles and means I have a 50% chance to evade destruction (not to mention there is always plain old retreat)?

If you retreat, I take your planet with ground troops or chase you away from my own planet.

Even if you Emergency Shelter, I can just apply more GCD fleets. With every planet producing 1-5 ships per turn. Losing a few extra fleets here and there doesn't matter. I don't care about losing fleets, but losing even a single dreadnought is going to hurt. Once a fleet loses one dreadnought, it has been weakened enough that the others will quickly fall.

As the Sheredyn you can't retreat from me so I don't have to worry about that from you (not to mention you lack heroes so can't use this cheese trick).

Why would I retreat? My ships typically aren't going to survive regardless of what happens. A GCD with 1 health is almost as good as a GCD with 450 health.

I think there are other factors with the hero that could in theory mess up your numbers.

Sure. Illusion comes to mind. Then I need more Kinetic GCDs.

It is possible that tactician (2) could throw off your evasion calculations? What if I used the engine module for+15% evasion too? This latter option is probably a bad idea but I'm just throwing that out there. Those two changes would be an extra -28% accuracy change; is it possible that an extra penalty could throw off your evasion calculations?

Probably not. The difference would be roughly 3 weapon modules, depending on some details. Engines are a horrible idea because they weigh so much and do so very little.

I don't know of a way you could factor all this in to one equation but I'm just making some observations. I saved a game at the end recently and let the tech run a few turns to max out all the ship stuff. I had one admin hero and 4 battle heroes and did not bother using the other 2 slots. I think 4 high level battle heroes is a reasonable expectation but then again if my strategy might depend on a 50% coin toss I think I'd go for 6 instead.

On what turn? If it is after turn 150, you are going to be running low on planets. Mind you, a Harmony using this sort of strategy is producing 1-5 ships per turn per planet (on average), every turn. It is very dangerous and intimidating. And with the siege system how it works now, it just takes 1 round to lose a planet.

The only race that is going to come close to the harmony GCD challenge is the Sheredyn.

GCDs aren't designed to survive very well, they are designed to pump out more damage. If I have to use some defenses, then sure, but otherwise it is all in weapons. If you have to build more weapons on your ships, then you are losing ground on the idea that survivability is the important part. The more weapons, the more fragile your ships become, which means the fewer GCDs need to survive.

While their DN has a rather poor bonus you should definitely use the power module it when it is half tonnage. With +1 CP from big fleets you get 8 DN instead of the normal 5 (and don't have a CP remainder).

Sounds like a good plan to me, perhaps the best. The Harmony Dreadnought is also quite nice with -50% weapon weight.

This gives you an extra 3105 tons to work with and 9000 base health (assuming optimal space). That being said I have to ask if I can only bring Dreads or if I can mix and match my fleet?

Bring whatever you want. Mind you, if there are support ships, then Battle 1 is going to be GCD fleet set to spread fire and will most likely kill off anything that isn't a dreadnought.

Something else to consider is that unless I am mistaken you are using throw away ships and I'm building ships that are built on the premise of never being destroyed. Is it not logical to assume that I've had time to level up my Dreads as well? A level 25 hero with level 9 Dreads makes a huge difference in terms of survivability (now if only I could get those initial creation bonuses for hp on such a fleet!). I'm assuming your ships will be somewhere between rank 3 and 5 as you don't expect them to live long.

Where are you getting the experience for your ships? If you are fighting my ships, then apparently I've lost tons of them already and so you shouldn't need them to be level 9. If it is someone else, then why are they losing so many combats? I've always assumed straight from the factory experience levels, because that requires no further justification.

Gwydion5 wrote:
I feel the word rematch is not the correct word to use, as it implies that there was a conclusion or consensus behind some major point.

In my opinion our test was rather limited and not done thoroughly enough to give definitive answers and conclusions. It was more or less a quick exercise with the end conclusion....

This is consistent with what I thought I had posted. Apologies if that was not clear, I was trying to be brief.

... being that in all our battles you almost destroyed a dreadnought at the end and if you had a second fleet you most likely would have destroyed at least one, if not the whole fleet.

Yep. Given that there were 2-3 more fleets of the same type of GCD it didn't seem worthwhile to run those battles. The conclusions were foregone.

But it's not helpful to assume, because on your first bomber fleet (which are not glass cannons, but armored carriers) you thought you had a winning fleet, then we fought 6 more battles and that fleet suffered attrition to the point that my fleet was fully healed again without taking a loss of a dreadnought.

Your dreadnoughts were able to heal because bombers were shot down and not present for the next battle. This reduced the combat damage of the fleet. Eventually they weren't able to do more damage than you were able to heal. Again, this isn't important, because it isn't the first fleet that does the killing, it's the follow up fleets. You'd be facing 2-3 fleets with that design and each one would be attacking every turn. That damage couldn't be soaked using your dreadnought designs.

For our next test, if you wish to do some more testing, I suggest the following



Setup:

Mid-tier tech again...

Factions / Traits :...

No Battle Cards : ...

Sure. Whenever you have time we can do it. I'll use the Harmony design I've had posted for ages. I've been considering the advantages of more CP and more tonnage, but let's not worry about that.

Phase 1 : Testing Weapon Type and Industry Efficiency...

Phase 2 : Testing with Fleet Cap limits. ....

Sure.

Other Notes : I would also like to add a Phase 3 to test moderate ship levels with an admiral. Because the tough super dreadnought fleets I have faced in human multiplayer, have had high level admirals and ship levels of 6 and 7. And those things can be very difficult to take down. Though I'm fairly certain nosebreaker LRM GCD's with bombers will probably do the trick. But I feel Phase 1 and Phase 2 are plenty for us to start with first.

So against a Harmony fleet, does that mean I get to bring in Metal Memory? How much of a bonus? What level of Bombers & Fighters?

Gwydion5 already demands a rematch using Fighters and PD modules.

I feel the word rematch is not the correct word to use, as it implies that there was a conclusion or consensus behind some major point. In my opinion our test was rather limited and not done thoroughly enough to give definitive answers and conclusions. It was more or less a quick exercise with the end conclusion being that in all our battles you almost destroyed a dreadnought at the end and if you had a second fleet you most likely would have destroyed at least one, if not the whole fleet. But it's not helpful to assume, because on your first bomber fleet (which are not glass cannons, but armored carriers) you thought you had a winning fleet, then we fought 6 more battles and that fleet suffered attrition to the point that my fleet was fully healed again without taking a loss of a dreadnought.





For our next test, if you wish to do some more testing, I suggest the following.





Setup:

Mid-tier tech again, I'll post a screenshot of the tech tree filled out that I had in mind, it's pretty much the same as before, but this way we both have the exact same techs. We can also discuss if there are tech advantages for one faction over another.

Factions / Traits : Take whatever you want so long as it is a real build that you would play that is not horribly gaming the test environment. (IE +3 CP, +40% HP, +30% Tonnage, +36% Damage, +15% Defenses, +15% Accuracy and then -30% food, -30% Science etc...) I will also object to Cravers as that will require advancing 70/80 turns for the depleted status and their tech for +5% damage per depleted planet will fall outside our tech limitation of mid tier. I will concede that a craver faction would probably be pretty brutal with that tech and 20 depleted planets but in a normal game that's a late tech/strategy for the mid game usually.

No Battle Cards : I'm open to the idea that we just don't play them. They can make a big deal but they can be repetitive as well and from a testing point they are a significant variable we should try to do with out... at this point anyway.





I will be using the Amoeba faction build listed in the screenshot a few posts up from this post.

Phase 1 : Testing Weapon Type and Industry Efficiency

I will build a single dreadnought. You will build a GCD fleet of equivalent industry cost. If that fleet is destroyed and the dreadnought lives we will battle again with you adding an additional ship to the fleet CP from before. We will test each weapon type starting with kinetics, then beams, then missiles, then bombers and finally if you feel a hybrid of missiles and bombers is worth testing we will do that as well.



At the very least there should be 4 battles assuming you win all of them with industry cost equivalent starting fleets that destroy the dreadnought. Also after each battle I will introduce a new dreadnought to so the experience gains are neutral to the testing.

Phase 2 : Testing with Fleet Cap limits. Since the phase 1 test isn't really taking in to consideration fleet cap, we will do another round of testing similar to phase 1 but with both of us being max fleet capped for our factions. The difference here in testing will be that you will build a minimum of 5 fleets for a single weapon type and then we'll do 5 battles to see how many fleets of that weapon type it takes to destroy the dreadnought fleet.



I imagine we will have similar results to our first exercise in testing where missiles and bombers will actually do some damage that if continued will destroy the fleet. The other part will be if battle fighters make a difference to the bomber swarm since Point Defense modules seemed very ineffective from a "kill the bombers before they destroy my ships!" perspective.

Other Notes : I would also like to add a Phase 3 to test moderate ship levels with an admiral. Because the tough super dreadnought fleets I have faced in human multiplayer, have had high level admirals and ship levels of 6 and 7. And those things can be very difficult to take down. Though I'm fairly certain nosebreaker LRM GCD's with bombers will probably do the trick. But I feel Phase 1 and Phase 2 are plenty for us to start with first.

Ship maintenance costs have never been more than a minor inconvenience. While it does have an impact, I've found the long term impact is rather trivial.

Great job guys!

That is interesting results.



I just have 1 comment:

In calculating economic costs, shouldn't we consider maintenance? At 16CP each CP cost around 6, so each fleet cost around 100 per turn. Of course in full-blown warfare this doesn't matter as fleets are constantly wasted. But in the gearing up/cleaning up, it seem to make more sense to stack up dreadnoughts?

As to how much each is worth, it seem to me retrofit is the best exchange, and that is really cheap.

I still believe the GCD fleets are better.

• They are produced faster and can be more places than the dreadnought fleets.
• With Troops being the best way to do seige warfare, you need space superiority for a single turn before the planet is lost. It is far easier to attack and defend with more fleets than fewer fleets.
• They require fewer resources produce. Ships without armor or defenses don't need those types of modules.
• The Race we tested was designed to produce very good dreadnoughts and poor GCDs, in GCD focused races, they would do better.
  
  

However, the Dreadnoughts designs may be better.

• Gwydion5 already demands a rematch using Fighters and PD modules.
• We should also drop the Repair module and add another armor module. That will make them tougher.
• For Anti-GCD builds, more LRKs and less missiles.
• GCD fleets are micro-intensive. You'll be creating 1-3 ships per planet per turn and have to manage them. Even with waypoints you'll have to create and manage fleets everywhere.
  
  

Replies to Comments

• MP is virtually useless measure of combat power. However, I believe Monthar means CP.
• I believe the correct comparison is equal economy costs. If a dreadnought costs the price of 5 GCDs, then the dreadnought fleet has to fight 5 fleets of GCDs.
• The comparisons were made with equal CP and the dreadnought fleets consisted only of dreadnoughts (16 CP = 4 dreadnoughts). The GCD fleets were the same size.
• Very little math was done in these tests, it was mostly combat tests.
• AI ship designs are bad. Very bad. All AI designs are seriously undergunned and overpriced. They also use Nosebreaker more often than they should, resulting in very poor combat perform. The combination of bad ship designs with bad combat strategy means that the AIs only advantage is the massive economic boosts that they get. I strongly encourage people to try out Ail's Ship Design Mod (and play against the Harmony).
• GCD Math is easy. Build lots of ships with lots of long range weapon modules, and maybe 1 of each defense module. Apply liberally to the enemy. Don't leave home without a fleet or two to go with you. Spreadfire is your friend. Short Circuit is your friend.
  
  

Durath wrote:
And one question for Gwydion5.... do you feel that blockade breakers and merchants is worth 23 points? For 1 more point you can have businessmen 3 and just ignore trading as you should be set for finances. It has been my experience that the trade bonuses are a joke unless you spend a lot of points on them. Even with a cooperation agreement (can't have that with eternal war) it seems like the trading is an extreme afterthought that generates far too little revenue and depends on meeting enemy races.

Do I feel it's worth taking, yes. Do I know if it's worth taking? No. I should probably take out eternal war, because in multiplayer games, a faction like that will be at war because no human player would give them the benefit of trade. Beyond that I do like trade, in the mid game you can generate a decent amount of science and dust per trade route. When I was a big Hisso player, it was a night and day difference in terms of managing the dust requirements of your empire and having a decent boost in research. But that was long ago. I want to say that trade will make up about ~40% of your total dust production, and ~25% of your science production. Which to me is something I think you can justify 23 points for.



In addition, blockade breakers are immune to the system blockades of other factions. I'm not sure how well known this is, but if other factions (for an easy example) blockade the worm holes of the galaxy, you would lose trade with the other half of the galaxy (in a disk galaxy type anyway). Because trade routes need unobstructed paths to their destinations. Now I don't know exact details of the relationship mechanics behind this. I have seen it with other factions where one of my trading partners is at war with a third party and the blockades end up hurting my trade with my trading partner. Perhaps if I was at peace with both factions I wouldn't be affected by the blockades? In any case there is a mechanic there at play which adds value to blockade breakers. Plus the ability to have trade while in cold war helps a little in the early game. Very useful against the AI

Thanks for doing the math and the tests.



I'm back to using the Armor Support Module since I like to park in enemy territory and reuse my ships as much as possible. It might not be the most efficient, but it seems to help me survive when I can't go back to my own territory to refit and repair. My defenders seem to outlast most of what the enemy throws at me in the early game. By late game, they are no longer needed though.

I'm glad that someone took the time and make all the maths



The ones who take their time to make the maths - they usually have the knowledge to win - but mostly it requires simply too much time to make a real "profit" from it.



Well done, well done, I'd like to see more too.

soooo, with your current designs they are quite similar in Terms of Cost-effectiveness?



(don't get me wrong, I'm very interested in this Project and eager to see more results )

Gwydion5 wrote:
I would disagree. The kinetic and beam GCD's were ineffective. If you keep sending these at me you're just going to level my dreadnoughts to the point that those GCD's do no damage due to ship experience bonuses. Then again I don't think you used nose breaker with the first 2 fleets. Perhaps that would have killed one dreadnought.

Correct, I did not use NoseBreaker. Had I, I would have nearly killed one with the first wave, and the following waves would have continued that trend. I'd be curious if melee kinetics would work out here as well. They make progress, just not very much.

Those did the most damage and almost destroyed a dreadnought. I think this GCD build can be mitigated some, partly with a single fighter module as explained above, and by playing a better tactics card on phase 1. I have 3 cards to choose from to try and offset the damage missiles do and the first round is huge considering how many die. But I'm doubtful it will stop the loss of ships.

I've got my bets placed in regards to this one. We'll see how it turns out.

I'm sure the LRM + Bomber build would be even more brutal with Tectonic hulls.

Not to mention the price. I'm used to cheap ships with lots of guns. Not these high priced lemons.

Gwydion5 was unsatisfied with the PD modules and wants to try fighters. I think he'll be greatly disappointed but we'll see what happens.

I was thinking of starting out with just 1 fighter module and 3 PD's and see if that is better against bombers. As you said it took 6 battles to resolve. Having 10/10/13 T2 Defenses and 4 PD's against T1 bombers left me unimpressed with their effectiveness. I'm thinking a single "battle" based fighter module will help with that. At least this way it will be 24 bombers versus 20 fighters so only 4 bombers will be left alone, while having 12 PD's instead of 16. I guess we'll see.

All of the Glass Cannons would eventually get through the dreadnoughts, you just need to come in with the economic equivalent in GCDs.

I would disagree. The kinetic and beam GCD's were ineffective. If you keep sending these at me you're just going to level my dreadnoughts to the point that those GCD's do no damage due to ship experience bonuses. Then again I don't think you used nose breaker with the first 2 fleets. Perhaps that would have killed one dreadnought.

The LRM + Bomber ships are the most cost effective, but also the weakest to counter destroyer designs (i.e., no armor so they are very weak to few LRMK GCDs).

Those did the most damage and almost destroyed a dreadnought. I think this GCD build can be mitigated some, partly with a single fighter module as explained above, and by playing a better tactics card on phase 1. I have 3 cards to choose from to try and offset the damage missiles do and the first round is huge considering how many die. But I'm doubtful it will stop the loss of ships.

Amoeba Hulls are useless compared to the glorious Harmony Hulls.

I'm sure the LRM + Bomber build would be even more brutal with Tectonic hulls.

Some Results from our initial test

• A Long Range Kinetic Glass Cannon Destroy is ~128 Industry, giving ~11 per dreadnought or ~44 per dreadnought fleet.
• A Long Range Missile Glass Cannon Destroy is ~145 Industry, giving ~10 per dreadnought or ~40 per dreadnought fleet.
• A Bidefense Bomber Destroyer is around ~200 Industry, giving ~7 per dreadnaught or 28 per dreadnought fleet.
• A Long Range Missile Glass Cannon and Bomber Destroyer is around ~250 Industry, giving ~6 per dreadnaught or 24 per dreadnought fleet.
  
  

We built 16 CP fleets to throw against Gwydion5's dreadnought fleets. We only tested his first fleet.

• The LRK GCD did around 8k damage and suffered heavy losses.
• The LRM GCD did around 8k damage and suffered heavy losses.
• The Bidefense Bomber Destroyers did reasonable damage (8k+?) and suffered little damage. However, the bombers suffered from attrition. We continued this fight for several rounds. Eventually the Destroyers were destroyed but it took several rounds.
• The LRM + Bomber GCD suffered heavy losses but almost destroyed the dreadnaughts. The second wave of ships would have succeeded and are within the industry cost of the dreadnought fleet.
  
  

Conclusions

• Gwydion5 was unsatisfied with the PD modules and wants to try fighters. I think he'll be greatly disappointed but we'll see what happens.
• The dreadnoughts must not be allowed to use the healing cards. It is a very strong recommendation to use the Tactics that stop engineering.
• The tests were performed using level 1 bombers.
• We didn't do any systematic testing of battle cards.
• All of the Glass Cannons would eventually get through the dreadnoughts, you just need to come in with the economic equivalent in GCDs.
• The LRM + Bomber ships are the most cost effective, but also the weakest to counter destroyer designs (i.e., no armor so they are very weak to few LRMK GCDs).
• I believe this trend will continue with higher tech levels.
• You should probably use Nosebreaker and not spreadfire.
• Amoeba Hulls are useless compared to the glorious Harmony Hulls.
  
  

sheredyn get bonuses for power module tonnages right? would probably make a great late game dreadnaught fleet

Forgot that those Amoeba builds are -30% cost from Militarists.







I think if we have time we should try a Sheridyn build too. I've attached the faction trait pictures.

thuvian wrote:
GCD fleets aren't particularly interested in heroes.

wouldn't that place them at a big disadvantage against a high level hero with 80% bonus on defense and offense for instance? besides glass cannon fleets tend to only use one style of attack, so what happens when the entire dreadnaught fleet uses all defenses to block your only style of attack. and would sheredyn dreadnaughts be able to use the late game power module? their dreadnaughts do have a bonus for them

thuvian wrote:
This is consistent with what I thought I had posted. Apologies if that was not clear, I was trying to be brief.

No worries, I was just clarifying as I found the word dubious.

thuvian wrote:
Yep. Given that there were 2-3 more fleets of the same type of GCD it didn't seem worthwhile to run those battles. The conclusions were foregone.

I understand that and may even agree with you that it is a foregone conclusion. But rather than assume we should play it out just to be sure in such cases.

thuvian wrote:
Your dreadnoughts were able to heal because bombers were shot down and not present for the next battle. This reduced the combat damage of the fleet. Eventually they weren't able to do more damage than you were able to heal. Again, this isn't important, because it isn't the first fleet that does the killing, it's the follow up fleets. You'd be facing 2-3 fleets with that design and each one would be attacking every turn. That damage couldn't be soaked using your dreadnought designs.

Again let's refrain from assuming especially considering we haven't tested a PD vs Fighter mix. ~1480x4 = 5920 ~200x16x2=6400 so 2 fleets would be the industry equivalent. I think 2 armored carrier bomber fleets would favor you to win this battle though. Also I'm fairly certain fighters don't behave like they are supposed to (according to the wiki article). I have noticed in the past that fighters split to engage 50% against bombers, 50% against ship. And fighters are grouped, so a fighter squadron will target just one bomber. So I'm eager to test this as we should see 2 squadrons engage your ships and 2 squadrons engage your bombers.

thuvian wrote:
So against a Harmony fleet, does that mean I get to bring in Metal Memory? How much of a bonus? What level of Bombers & Fighters?

Metal Memory falls within the tech range, but metal memory is meant to address the lack of leaders for the Harmony. So I would say if we are testing without leaders, then we are testing without mineral memory. The tech level for T2 fighters and bombers is outside the mid-game tech level I have in mind. I'll try to post that screenshot tonight. Been putting in 12 hours a day at work this week and just unmotivated when I get home to go on the computer.