I was using the Drengin tree with a custom race.

Yeah those were my thoughts on "Good" you have to give up all the planatary bonuses to get there then once you do you end up with the worst of the lot.

I would expect good to have different but just as effective skills as evil, with neutral a bastardisation of the two (sorta like those starwars games like Kotor etc where good makes you awesome at some force skills but worse at lightning and chocking people and evil visa versa)

I am generally good but almost always get the "leaning evil" because of colonization events, by the time I research xeno ethics 10bc / turn for 1000 turns is like ... nothing, and even sometimes I can outright pay it

BUT I don't always take all evil choices in colonization events because some don't justify the management overhead (aka me, the player, need more time to handle planets with too much specifics) like +9% research or +20% growth... (improved growth on a single planet don't really give that much of an advantage when you got more than 40 planets ...)

some tech trees also get more benefits for being good in TA like the drath which get a +3PQ building ...

which imho is really nice along with terraforming to make all your planets PQ 10+ (smallest being PQ4 + 3 from terraforming + 3 from building)

Also someone said to me in another post, that there's more power in overunning an opponent than sheer weapon technology dominance...

@Roxlimn

PD Combo is an extra 1700 research over Telepathic Defense, and has the same stats-but is cheaper (30 to 40).  Barriers III is roughly equivalent to Subspace Rebounder-as this necessitates an extra 2700 research, that may be worthwhile.  Dynamic Shielding is unlocked right after that, and isn't equaled for another 6200 research, so it's actually fairly good.  Kanvium III is almost equivalent to Arnorian Battle Armor-Duralthene isn't justifiable for the space alone-which is an extra 3500 there.

In the Terran tree, psionic beam is not near equaled until doom ray, which is an extra 20600 research.  Psionic shredder requires another 22200 until black hole gun.  Photon Torpedo III is another 10400 beyond psionic missile-by far the lowest of the three.

(These numbers are all assuming you skip the named tech and go on to the one that equals it, by the way.  Add another 1k or so if you don't for weapons, and another 200-400 if you don't for defense.)

Even in TA, you can max out roughly 1.5 defense trees per 1 weapon tree.  Back in DL and DA, you could max out all three defense trees and still have room left over before you even maxed out beam weaponry.  This is a combination of the weapons trees costing significantly less in addition to the fact that the defense trees cost slightly more.

Also, the combat mechanics in DA/TA tend to favor attack ships more so than defense ships.  While the defense modules can help with this somewhat, it's still not enough.

I've already lost my point once or twice due to ranting (not at you, by the way), but I think I've shown that going evil saves you more research.  And again, in DA where defenses are cheap and weapons are not, there's no point in going good because you can max out a defense tree in about the time it takes you to unlock a single evil weapon.

-

It's worth noting that your military rating is generally higher with a ship with one weapon and filled out with the highest defense you have than filled with the highest weapon you have (which can be useful for intimidating the AI via the spin control center) due to the fact that the tech tree provides a minimum 10% defense bonus while there are generally not weapons bonuses in it.  However, some races do get weapons bonuses in their tree, so in general all-weapon ships will produce a higher military rating for them.  Obviously, if you go Good, all-defense, one-weapon ships will give you a far superior military rating.

Free terraforming is nice, but since you have to have the tech already (fixed a bug from early DL I believe), I'd rather just build the orbital terraformer and go evil.  There's also the bonus that it doesn't let the AI get it.  Admittedly, 800 production is hefty, but it's worth it to me.

And not to beat a dead horse, but just for the record, the rush buy ability does not work in DL and DA, so anyone still playing either of those versions has less incentive to go Neutral than they otherwise might.

-

whismerhill, valid point, I neglected to fully consider TA's ethical trees, which do tend to shake things up a little bit.

I haven't seen a PQ4 that didn't turn into a 15/16 in a long time, but I do see them every now and then in TA.  Still, I think your lower limit there would tend to be a PQ5, as those always turn into PQ8 (and nothing better) via terraforming.  Adding in an extra 3 tiles from biosphere modulator or weather control zenith takes it to 11, which, incidentally, is the point you were trying to make-as PQ10 planets don't get the approval bonus, but PQ11 planets do.

You don't have to go good to get weather control zenith off of the Arceans, though-they just have to research it.

-

Just for the record, I started out playing Neutral going for NLCs (and the free terraforming was a nice boost), but a few games later I switched to evil and have never looked back.  I don't personally have trouble with the AIs when I'm evil, even when they're good, but that may just be because they're afraid of my military.

All things being equal, Evil alignment is superior long term to Good in Twilight (and vastly superior in DA/DL due to the MCC).  However, I think the situation is a bit more complex than Evil>Good in Twilight.  Timing windows matter in this game.  An small advantage early can be more important than a large advantage that comes later, since it accelerates your development relative to the AI opponents (which can give you the chance to execute some strategies before the AI gets strong enough to stop them).  Evil is basically useless starting out.  Its advantages show up later, when Good and Evil and Concepts of Malice have been researched, the Artificial Slave Center and No Mercy Invasion Center have been built and you've reached the point you can start setting up your starbase array (the free starbase upgrades speed things along then).  Why does this time lag matter?

In Twilight, stabilizing your economy post-colony rush is critically important or you can hit a dead spot in your development where you've got to cut total spending.  On Suicidal, that really hurts.  In my experience (and I'm far from the best player out there), getting the economy stabilized depends pretty heavily on getting your population up, which means you've got to keep morale at 100%.  That means cutting your tax rate, and that in turn means not a lot of revenue coming in unless you can get your morale up.  Harmony Crystals are very useful for that purpose (+20% morale empire wide), but you've got to select an alignment when you research Xeno Ethics, which unlocks the Crystals.  Switching from Good to Evil zeroes out your treasury in most cases and adds a 10/turn lease payment.  This can kill your economy early game. 

To avoid that, you've basically got three options:  Choose a method of stabilizing your economy that does not involve research Xeno Ethics (some combination of any or all of taking heavy morale bonuses in race set-up, building morale buildings early on each planet, making morale resources a very high priority, research techs that give economy bonuses early, building economic buildings early, getting the econ capital build on your homeworld early, or getting a bunch of survey ships out).  Bottoming your economy out to go Evil, then selling tech to get the capital you need to keep production going until you can get the Harmony Crystals built and your tax rate up (not a great method since it's often more favorable to trade tech for planets or unique racial techs than cash) is option 2.  Third, just go Good.  Then, Xeno Ethics becomes part of the solution rather than part of the problem.  You save 50/turn upkeep immediately if you have at least 6 planets total, you don't zero out your treasury, and you don't pick up a 10/turn lease payment.  This means you can get the Crystals built earlier, which can make a difference of 10-20 turns on how quickly you can get your economy ready to support the next phase of your game (in my case, usually conquest, though there are several ways to go).  10-20 turns may not sound like a lot, but if it comes in year 0, it can make the difference between rapid, painless absorbtion of a nearby empire or a protracted war to achieve the same end.  And the point at which you can absorb another empire usually is the point at which the game starts rolling downhill in your favor (at least in my experience).  The earlier it happen, the quicker things accelerate. 

This applies primarily with the Iconians.  The Altarians obviously should go Good, since they have a gob of techs unlocked by going good and can't access any for going Neutral or Evil (though the above reasons going Good early is helpful all still apply).  The Drath can easily fix the early game cash crunch through war profiteering, and may be better served going Neutral.  The Torians usually get their population levels up quick enough that there aren't severe post colony rush economic problems, which means less drastic measures to get your income up can be sufficient.  However, the Torians have a native tech that provides I believe a passive +15 defense, and if you combine that with the +15 defense you pick up by researching each of the good defenses and +20 for the empathic defense wonder, you've got +50 defense, which is a big enough defense bonus to matter.

Not really disagreeing with what has been said above about the advantages of Evil aligment, but just wanted to point out that there can be reasons to go Good.  Too often people thing going Good is only worth considering if you want to handicap yourself to make things more interesting, and that's not really true anymore in Twilight.

Sole Soul:

I believe that you don't really get the point of the defensive techs.  tetleytea certainly appears not to.

The point of the defensive techs is to rapidly research up the defense tree that's appropriate to your enemy's offensive line to the point where you can cost-effectively build a ship that's invulnerable to his fleets.  You don't research up this line until you have established that your enemy is pursuing a particular weapons line.  As soon as your decisive war is over you abandon researching the defense tech branch.

So that 1700 research that you saved on Telepathic Defense is 1700 sooner that you would have otherwise gotten your war-winning tech AND 1700 research you could be putting into Miniturization or some such.

It's irrelevant that Psionic Beam deals so much damage because you can never build a single Psionic Beam Ship that'll be invulnerable to all comers - or even any specific comers.  You'll take more and more damage until you're just a 1 hp ship and then you have to build enough of them Psionic Beams into your ship to get the Hyperion Dreadnaught 1 hp tie rule, which tends to be prohibitively expensive.

Sure, you get to Psionic Beam and then stop all weapons research - you'll need to pour those into Miniturization and Econ techs instead to allow you to fund your hideously expensive ships.  The maintenance alone on those things is killer.

I do not agree that the combat mechanics in DA and TA favor attack over defense.  In fact, I believe it's quite the opposite.  You need good economy, good logistics, and good weapons to win with weapons.  You only need good defenses to win with defenses.  I've done this over and over again in my games (I only play Tough though, if you must know).  I've reloaded games and researched offense over defense.  In every case, researching on-type defense was more cost-effective and more decisive.

Defense techs win wars.

Too, as Good, you get something else which hasn't been mentioned as yet - +25% Diplomacy bonus.  This translates to more tech-sharing and faster overall tech advancement by Good Civs.

JustinSane4:

Noticed that effect, too, with the Iconians.  Their early buildings are so uber compared to everyone else's that as long as they can get enough with them on it and enough cash to feed the furnaces, you can win easily.  Too, the Iconians' signature Organic Hull Plating is insanely good on the smaller ship hulls, making defensively oriented Smalls and Mediums that much more powerful.  Had a 100+ HP Medium at one point, from XP and tech bonuses.  Wow.

As Neutral, my preferred method for stabilizing the economy post-rush is with Ion-enabled Frieghters.  Would that be too slow on Maso?  I don't have plans to play anything above Tough - don't like artificial bonuses - but I'm curious.

@Roxlimn

In 1v1 battles, yes.  Assuming same tech level (let's use psionic beam vs. dynamic shielding + particle beams II as an example, from the TA Terran tree), once you have two or more ships on each side of the fight, all attack is king.  As a note, I have assumed cost equality as well.  Feel free to adjust that for your own calculations.

The following pseudo-random example is based upon my theory that your argument only begins to hold water in a midgame environment, as all attack clearly rules late game while in the beginning you simply can't fit much more than a weapon or two on a given ship.

On a medium hull with 40% miniaturization, one particle beam II and 14 dynamic shields use up our space, and will cost us slightly less overall than 5 psionic beams.  (Note: 7 will fit.)  This leaves us with a 2 attack/84 defense ship before applying our defense bonuses from researched techs (in this case it can be assumed to be 20%), which brings us to 100 defense.

On average, facing off against a single all attack ship with 5 psionic beams, our ship will do 1 damage per round and take -20 damage per round (which essentially means we take no damage).

However, if we compare two all-attack ships vs. two all-defense, one weapon ships, we have our defense fleet averaging 2 damage per round while our attack fleet averages 10 damage per round.  It gets successively worse from there, as adding a third all attack ship brings us up to an average of 40 damage per round, while our defense fleet is only at 3 damage per round.

You can not agree all you want, but in DA/TA, not only do all weapons roll separately (which prevents attack ships from wasting excess firepower on an already destroyed ship), but defenses deplete during each round (although they are refreshed for each new round), which leads to the above scenario where defenses can be "overwhelmed" on a per-round basis to destroy ships.  This can be overcome if your single ship has sufficient defenses to counter my fleet's firepower, but defenses aren't -that- much cheaper...particularly Good defenses.

Additionally, if I put even first or second-gen defenses on my all-attack ships, your all-defense, 1 weapon ships can't hardly scratch me, as even off-type is sufficient to grant me a defense value of 1.

The short answer is defenses are not that good, and should really be more worthwhile than they are (see comment about how them being significantly cheaper than weapons trees in DA was a good thing).

https://www.galciv.wikia.com/wiki/Ship_combat

JustinSane4:

I only have experience up to the Tough difficulty level, but up to that level, defense ships are extremely powerful and useful.  They don't, in fact, go from "all-powerful" to "junk" very quickly, and retain a great deal of usability even where their total defense might not, in fact, make them invulnerable to fleet action.

The key point of usability is based on the fact that up to a certain attack value, a Defense ship is completely invulnerable - capable of multiple kills and constant action.  As long as it has something to attack that it's invulnerable to, it's a useful ship.  For instance, when taking on single-ship defenses on invasion, it's often better to use a Defense Ship to batter down orbital ships 1v1 rather than take HP losses from your main attack fleet while also wasting valuable move/attack on depleting orbital defenses.

Not all AI fleets, or even my own fleets, are going to be up to date at max attack value.  At Tough, Defense Ships using Arnorian Battle Armor and Subspace Rebounders retain great value for a long time, without upgrading.  I cannot say for higher difficulty levels, but I'm supposing that I would be able to utilize them in similar fashion there, as well.

Sole Soul:

You example and flawed analysis only convince me further that you simply do not know how to design defense-strategy ships, not that they're useless.  I KNOW that they're not useless because I've had great success with them - moreso than with Psionic Beam ships, I might add - and those don't really require much of any gray matter to use or design, begging your pardon.

First of all, I don't analyze my own ship designs against what I can fleet or throw together because I will not be attacking my own ships.  Clearly, when we compare the efficacy of various options, we ought to be considering what our goals are, not whether or not our designs will win in theoretical gladiator matches within a theoretical ship arena in our worlds.

IF I had to face down Psionic Beam ships, I would not overkill the defenses in the name of theory.  In that case, 100 defense is clearly much more than we need, and as you point out, skimping on offense would render us too easy to counter.  Two Particle Beams on Dynamic Shielding would probably work best on a Medium hull along those lines.  Assuming we both had researched or acquired Logistics tech that allowed us to fleet two ships (but not three), I would naturally NOT pit 2 Medium defense ships in a fleet against 2 attack ships.

I would, however, probably be able to use my Defense Ships to attack and win any number of battles to which they are suited, and I can make it so that this occurs frequently.  I don't, after all, need to be stupid about tactical considerations.  For instance, I can concurrently design and field all-attack Particle Beam II ships and fleets that would be sufficient (and would also be relatively cheap to create and maintain) to take out 1 of the Medium ships in a Psionic Beam fleet before kicking the bucket.  I could then send in my Defense Ship to mop up.  I would lose a Particle Beam II fleet, but he would lose a Psionic Beam fleet, and I get to pocket the XP in change.

I wouldn't need a lot of Dynamic Shielding ships here.  Just one or two to accompany a bunch of Particle Beam 2 ships and together they would be able to take on a variety of fleets and challenges that purely Psionic Beam ships alone would not be able to handle.  Furthermore, my Dynamic Shielding ship at 81 defense would have 9 off-type defenses as well, so it would also be able to take on any fleet with that attack or less and come out without a scratch.

I found them much more powerful in DA because the defenses were cheaper, but their being slightly more expensive in TA hasn't prevented me from successfully using the same strategem - albeit with much less in the way of mindless hammering - I now have to think about tactics in order to sweep fleet upon fleet under the rug.

A Psionic Beam fleet is incapable of constant action - even inferior fleets with inferior attack ratings will eventually deplete its HP store given enough of them.  You only need 1 Defense ship to kill fleet upon fleet of the right targets.

You imply that logistics (as you mentioned that 2 fleeted Mediums would tip the balance towards offense) would tip the balance towards offense, but I counter that Larger ship hulls and Reinforced and Organic hull components tip it right back towards Defense.  Compare, for instance, your 3 Psionic Beam Ship Medium Hull fleet using Advanced Logistics against a Large Hull Dynamic Shielding ship.

I personally don't understand why more people don't use planets to regenerate hit points faster.  Even on my most damaged huge hulls, it generally only takes 3 or 4 turns before they're battle-ready again (albeit not at full hp).  But Jonnan's mention of that upgrading loophole is probably it-why wait 4 turns when you only have to wait one?

-

There's also the fact that the AI-and I've noticed this all the way up to Suicidal, though it's less prevalent-has difficulty researching the ethical techs.  The primary relevance here is that in general you won't be fighting an AI with either dynamic shielding nor psionic beams.  Given that little bit of information, we can conclude that your defense ships against an AI fielding particle beam IIs or even harpoon IIIs will have more than a fighting chance and will in fact lead you to believe that defense is superior to offense.  But of course they will-you're facing ships with 1/6th to 1/3rd of the firepower that you otherwise might be.

In the examples above, I've considered tech investment, ship cost. and miniaturization to be effectively equal.  On Maso+, this can no longer be assumed to be true-the AI is going to outresearch you a good 7 or 8 times out of 10, and at Maso and above they start getting miniaturization bonuses.  In this kind of environment, the research that going for psionic beams saves you is far more precious than the research that going for dynamic shielding saves you.

Roxlimn, I do believe you're missing my point.  I'm not saying that defense ships are useless, nor have I been.  I'm just saying that they pale in comparison to attack ships for what has been in my experience a very large portion of the game, and secondly that defenses should be better than they are.

If you'd like to compare a large hull defense ship to a medium psionic beam ship, and accounting for the above math where it was proven that a single defense ship could beat back a single offense ship, so extending it to two ships, we wind up with a large hulled defense ship with 2 attack/120 (144 after bonuses are accounted for) defense, at a cost of 1195BC.  While two of these ships could beat back two medium hulled, 5 psionic beam ships (5, not 3 as you stated at the end of your reply), as I stated in my earlier reply, seven will fit.  This gives us 84 attack per ship, at a cost of 1130BC each (!), which again has our two ship medium attack fleet defeating our two ship large defense fleet-at lower production cost, and lower tech investment (as I didn't bother to equalize tech costs based on the research of superior hulls and large scale building, which are 500 and 3000 respectively).

I freely admit that my examples are perhaps not the most appropriate ones, and therefore could be considered flawed.  However, the math is not.  I encourage you to do the math for your own examples, and to consider unequal tech investments (with the advantage to the AI) as well as unequal miniaturization.  But I would like to point out that I did try to choose a point where your argument does have some weight-and as you can see by the fact that a single defense ship in our calculations can defeat a single attack ship, I was not wrong in that regard.

You are correct in mentioning that I have neglected to consider the additional hp modules that a ship may have, but for my calculations they are effectively irrelevant, more so as any hp modules on said ship will result in that ship having less defense, and therefore taking more damage.

-

While whismerhill's view of balanced is not entirely wrong, it nevertheless fails on a few accounts.  On the one hand, if we are ignoring tech cost equivalency, then we can choose virtually any point within the tree and make our argument against a lesser version.  I assume that this is not what was intended by that statement, though.  So the logical conclusion becomes to compare end-tree techs.  Now, aside from the fact that a ship with both will have an increased research cost, which whismerhill has kindly allowed us to ignore, we must still fit enough defenses into said ship that we are able to meet or beat our opponent's offense.

On a 115% miniaturization huge hull with 322 space:

Our options are 575 attack or 25 attack/440 (484) defense or somewhere in between.  With each extra weapon, we gain 25 but lose 22 (20 * 1.1), so maybe we should just put more weapons on it...oh, wait, that's the ship we have on the left.

It's worth considering that ZPA has the best absorption/space of any armor and that black hole generators are the worst to put on a huge hull, so you might have something if you try it that way.  We wind up with a mere 416 attack this way, compared to 16 attack/510 (561) defense, so there's something to be learned here-don't put mass drivers on huge hulls.  Incidentally, our defense ship winds us costing us nearly 600BC more-if we equalize cost we're looking at 16 attack/420 (461) defense, which is quite a bit closer.

With beams we have 440 to 22/459 (504).

But stop and consider something.  This is one ship.  This is a 1v1 fight.  Didn't we prove a few posts back that defense wins 1v1 fights handily?  Hell, in the first example I gave you, your defense ships almost (but not quite) won a 2v2 fight-and that was against psionic beams!  And here it is with a larger tech cost, and in one fight it loses, while the other two are much, much closer than they are at other tech levels.

nice strategy but I don't think it's THAT easy to pull off on immense maps

JustinSane4:

If I were trying to be rude, I would call Sole Soul a noob.  I haven't.  He may be knowledgable about game mechanics, but it seems clear to me that his knowledge of game mechanics is leading him to incorrect conclusions - so much so that he doesn't even seem to want to try out something that I have personally found to be devastatingly effective.

I know that the HP modules change the equation somewhat, and it kind of changes the tip off point for which you want offense/mixed/defense ships based on the opposition.  Having said that, I find it more efficient to repair ships in orbit and rotate them out because my defense ships do not normally require repairs.  This means that I have an always-active screen of ships for taking out defenders, transports, mining installations, freighters, and other such small fry.

Think of this as a "frigate" class ship.  Even when they're not powerful enough on their own to take out the biggest war fleets, you design them so they're fast enough to avoid anything that can kill them, and powerful enough to kill anything they can chase down.  They're remarkably useful for that, at least.

Sole Soul:

But that is exactly my point.  We're fighting the AI, not ourselves.  We do not design ships to take out what would theoretically be our own ships.  We design them to take out the enemy.  Given a choice between a ship with overwhelming firepower and a ship with overwhelming defenses, I choose defense every time, and I'm often rewarded handsomely for that choice.

Unless you can build one ship with enough firepower to overtake an AI's entire fleet, you cannot take advantage of the 1 hp rule.  Until you can do so, bringing more defenses to the table pays off better.

I did not miss those points.  I believed defenses to be too-powerful in DA and that they're about right now, though they're still enormously powerful - easily the Good military overpowers the Evil one.

As I said, equality can be tossed aside at Maso because we're playing the AI, but by that same token, we also expect it NOT to have Psionic Beams very often.  Even if it did, you would still be best advised to take defenses over offense because invulnerable ships recoup IP disadvantages quite handily.

Of course not.  Many of the best HP modules in the game are extremely small relative to ship carrying capacities so the odd space you can't fill with weapons or defense can easily be filled out with a HP module or two.  Too, if the offense/defense math eludes you, you could also use HP modules to tip the balance in your favor, and this often favors a mixed approach or a defensive approach.

In our example, you assume again that I'll design a ship with 1 weapons module and then all defenses.  Not so.  That's overkill, and is not particularly useful.  I'll only need so much defenses, and I wouldn't be using these exclusively anyway, as I mentioned.  At miniturizatoin levels where a Psionic Ship can have 84 attack per ship, I could build a Meduim or Large hull with 20 attack - all I need to suicidally kill one Psionic Beam ship in a fleet of two, after which my Medium Defense ship will handily mop up the remainder.  I lose one ship to his two, and his two ships lost would have had expensive Psionic Beams on them to boot!

As long as a defense ship can beat an equally leveled or even slightly more advanced offense ship 1v1, it remains the backbone of my fleets.  My all-offense megafleets will still have uses, of course, but they will largely be only for the isolated set-piece megabattles - happens perhaps once in a war.

NOW, assuming that my megafleet will lose - and if the AI is better in every respect, it should - then I would rather have a fleet of Defense ships to mop up whatever remains of the enemy fleet.  They would be better at that.  Once I have space superiority, my Defense fleets will be better at blockading planets to make sure the enemy never has the opportunity to recreate a like fleet ever again.

I DON'T need repairing or upgrade-repair loopholes because the backbone of my fleets never need to be repaired!

Roxlimn, you keep using your experience on low difficulty levels to prove that you are right.  Fair enough - these are clearly strategies that you have used a lot and with success.  The problem is that you then try to use that same experience to prove that other posters in this discussion are wrong.  You offer no evidence for why these methods would work on Suicidal other than that they work on Tough - sorry, but there isn't really much correlation.  When opposing fleets get over 1,000 attack, your defense fleets will melt.  When your Good defenses are facing not Psyonic Beams but Doom Rays because the Korath have reached the end of the beam weapon tree before you can finish researching shields due to Suicidal's research and economic bonuses for the AI, and they are packing more of them into a hull than you could due to the miniaturization bonus, the rules that used to work stop working.  That is simply too much firepower for any defensive design with the game rules to deal with.  Now as support ships to pick off stray weaklings, starbases, transports or orbital defenders defense ships could still have some utility.  You could even use defense ships for other purposes, such as to get the maximum diplomatic impact out of the SCC, or even, temporarily, as the backbone of your fleet within precise timing windows - before the AI is able to research the highest levels of weapons and logistics. 

You can make an argument that if you play well, you should end the game as a contest before this point - and the best players usually do.  But even there, to my knowledge none of of them consistently use defense ships in combat.  Considering how well the upgrade trick does work, basically any role that can be filled by a defense ship with the possible exception of clearing orbital defenders can be equally filled by an attack ship, with the bonus that the attack ships will still be useful late game.  You also say that you use both defense ships and a main attack fleet composed of attack ships - that means that you've had to research both lines, which is a massive investment of time and resources, giving the AIs time to build their fleets (again, particularly on Suicidal).  If you had just researched the weapon branch, you could have started the conquest much sooner.