Ship design strategies
From the Galactic Civilizations Wiki
Ship design strategies can be documented here. It is highly recommended you read the Ship_combat section before you start designing your warships.
Contents |
[edit] Optimal Ship Size
Does size really matter?
The quick and dirty answer is that it depends. There are two schools of thought on shipbuilding. The first, is to go small, but build in high quantity. The second is to build fewer, larger ships. The differences depend on the strategy you most prefer. Smaller ships get fewer hit points, and can generally fit smaller numbers of weapons, armors and specialized modules. However, the cost of producing these smaller ships can mean that a player can produce them much faster than a frigate. Another advantage of smaller ships is that you can put more of them into any one fleet. Because Gal Civ adds all of a given fleet's ship's attack values together, this can mean that a bunch of small ships can take down a slightly smaller fleet of large ships.
Let us think of it this way. 4 fighters with an attack value of 4 beam and 6 hit points are in a group against 2 medium sized ships with 6 beam attack and 20 hit points. When the battle commences, the four little ships take out one of the big ships fairly quickly (4 attack x 4 ships). Granted, the algorithm is not that simple, but you get the idea. The big ship returns fire, and maybe manages to kill one of the little ships. Still, with three little ships remaining, the attack value is still twice what the remaining big ship can manage. Thus, David brings down Goliath. Add in to this the fact that you can continually produce smaller ships at a faster rate than an equally situated opponent can produce bigger ships, and you can see how small ships can win the day.
This does not mean, however, that smaller is better. Bigger ships provide flexibility, and by boosting the higher hit points, can suck up a lot of damage, making them more likely to survive battles and accumulate experience, which greatly affects ship performance. Further, a combination of big and little ships can mean that the big ship survives, gains experience, and also benefits from having the extra "oomph" that smaller ships added into a fleet can provide. Further, there is just something cool about seeing a dreadnought fly up next to an enemy planet and proceed to decimate that planet's defenses.
Does this mean that a mass amount of medium and small sized ships would be the most cost efficient for most games?
No, it is risky to put very expensive weapons or armors on the small ship because they die easily. The bigger ships are also more logistically efficient. 4 small fighters take twelve logistics points, while 2 medium ships take 8 points. This means 4 small fighters are more likely to find themselves fighting against 3 medium ships instead of two!--Dabei 01:08, 13 March 2006 (UTC)
[edit] Best Weapon Technology
Beam weapons, mass drivers, missiles... all of the above?
Generally you should put on the weapons that has best damage/size ratio. However, some weapons like nano ripper are significantly more expensive. There is a rock/paper/scissors arrangement of the weapon types. Though many players prefer high offense to overpower defense, it pays to select weapons that attack the opponent's weak points. If your target empire's navy is dominated by ships with low mass driver defense, use at least some mass drivers to attack them.
A second consideration is specialization. Trying to raise all three weapon areas might lead to mastery of none. Consider having one weapon technology be ahead of the others, so that overall firepower is enhanced. This is especially true early on, when the second tier weapons are twice as strong as first tier weapons.
However, other schools of thought suggest that maintaining variety can be of real value. Some find it very effective to have all weapon techs available at about the same level, which can be easily achieved by grabbing all necessary techs from other civs. Thus equipped you can always react to AI ships, no matter their defense. With a mastery of one weapon tree you more or less only achieve to be able to pack more of them into one ship, something you can simulate with miniaturization research.
According to this thinking, mastery in one and beginner in the two others will probably mean your demise when the AI equips their ships with the corresponding defense, which it WILL do on bright or higher intelligence level.
[edit] Further analysis of weapons
- The size of beam weapons scales the most with hull size, missiles the least and mass drivers are in between. So generally, the damage/size ratio of missiles improves as you use bigger ship hulls and miniaturization.
- Mass drivers tend to be the cheapest, followed by Beam weapons, and finally Missiles.
- Beam weapons are the cheapest to research to their highest tech-level. It costs 49,400 research points to get the best beam weapon, while it takes 74,735 research points to get the best Missile weapon. Mass drivers take 67,000 research points.
- Subspace Blasters - Subspace Blasters, the top Beam weapon, is worse than the preceding weapons Disruptor I, II and III.
- Quantum drivers I, II and III are worse in 99% of situations than the preceding tech Graviton Driver IV. This is due to the poor damage to size ratio for Quantum drivers across a wide range of hull sizes and miniaturization bonuses. In almost all cases, Graviton drivers IV are better as you can inflict as much if not more damage given the same hull space use.
It's impossible to determine which weapon types are the most efficient without considering the defenses they're up against. For example: Missiles might have a better damage/size ratio, but Point Defenses might be equally efficient and cancel the advantage of Missiles.
An additional analysis has been done using spreadsheet data about weapons' properties and research costs. The following graphs are likely valid for any GC-2 game above version 1.1. up to current 1.5 DL and 1.6 DA. The simulated ship was filled-up only with weapons (fractional parts also used), whose sizes were corrected with stated sizemod. General impression is missiles are the best weapon type to use throughout the whole game. Despite using the last mass-tech weapon Black Hole Generator makes ship slightly more affordabe than ship with Blackhole Eruptor missile, is there the Positonic Torpedo II, that beats all other late-game weapons in firepower per price ratio.
Please note those spikes in graphs are mostly generated by evil-only psionic weapons and Nano Ripper. The highest is from Psionic Beam, that has so big firepower (attack 12, 8 space), that it completely unbalances the game. So if you want an easy win, go evil and use Psionic Beam.
[edit] Weapons and Probability
One point not touched on above has to do with probability theory. Per the manual, combat involves the summing of all of the weapon points in a fleet by type (all beam, all missile, all gun), then determining a "damage roll" based on a random whole number generated between one and that total. This is then reduced by a similar roll for the defense (see below), and any leftover points are applied as damage. This has two major consequences for fleets with multiple weapons types:
- All other things being equal, having the same number of weapons points in all three categories produces slightly more damage than in only two, and two are slightly more powerful than one, on average. That is, in fleets with 12 total offensive points, a 4/4/4 fleet does more damage than a 6/6/0 fleet, which does more than a 12/0/0 fleet.
This is because each weapon type "rolls" separately. The minimum damage that each weapon type can do is 1 point. The minimum damage total for a two-weapon-type ship is thus 2, and for a three-weapon-type ship is 3. The "roll" in the previous example runs between 3-12 or 2-12, rather than 1-12. The average damage is thus 7.5, 7.0, and 6.5 respectively. True, the rolls are separate, and are defended against separately, but the average damage is still increased by 0.5 points for two weapon types, and 1 point for three types.
[In the later versions of the game, damage varies in the range [0,max] rather than [1,max], and therefore the above logic would not generally hold. However, if the player has luck, this changes the range from [0,max] to [luck*max,max], and this kind of logic may not only hold, but be more powerful].
How does this help? Not much, unless you are very early in the game, when fleets are small and combat factors and hitpoints are relatively low. Later in the game, the slight difference won't be worth very much. Unfortunately, early in the game is when you probably can't afford to research all the weapons techs.
- Beyond this, having multiple weapons types changes the probability distribution for damage totals. This is again because of the separate rolls. Drawing a number from a small distribution (say 1-4) two or three times will give you the same average value as drawing once from a distribution across the same range of values (2-8 or 3-12), but the probability of drawing each number is different. This is a fact commonly exploited in table-top strategy games and RPGs (the infamous 2d6 roll).
In our previous example, drawing once from 2-12 or twice from 1-6 gives the same average value, 7. However, in the first case, all resulting numbers are equally likely, with a chance of .09091 (1/11) for each; in the second, the odds are quite different. If we think of each draw as a role of dice, there are six sides to each die, each of which has a 1/6 chance of turning up. However, treating each die as a separate roll, there are 36 possible rolls, with six ways the dice can come up with a total of 7 (1-6, 2-5, 3-4, 4-3, 5-2, 6-1), but only one way (6-6) to roll a 12. Thus, the chance of rolling a 7 is .1667 (6/36), while the chance of rolling a 12 is .0278 (1/36).
The upshot is that the DISTRIBUTION of total damage becomes more narrow with more weapons types. A 12/0/0 fleet can expect 50% of its shots to do between 4 and 8 points of damage; a 6/6/0 fleet can expect just under 50% of its salvos to do between 6 and 8 points of damage (note the increase in average damage, as pointed out above. A 4/4/4 fleet can expect just under 50% of its salvos to do either 7 or 8 points of damage. That is precision!
What does this mean for YOUR fleet? That depends... the interactions of weapons and defenses can be quite chaotic in the game. Against a defensively balanced fleet (roughly equal numbers of the three defensive commodities), the outcome of having more than one type of weapon (in approximately equal quantities) is based largely on the ratio of fleet offensive rating to individual defender's defensive ratings. With more diversity of weapons types, the offense's rolls will converge on the average value... if this average is significantly less than one half (say, one-third) the average defending ship's defenses (totalled for each category, including the squareroot contributions of each type), few of your salvos will actually score damage. If the offense's total is significantly more than half the sum of the all of the defender's defenses, the offensive fleet is likely to score at least some damage on most salvos.
However, trying to calculate these ratios out ahead of time is really a lot of trouble. But it is something to bear in mind. Also, bear in mind that, since weapons are totalled by fleet, you can still have specialist beam, missile, or gun ships, and simply recombine them in different mixes as you feel is appropriate. However, it'll still cost you to get all that technology.
Edit: It should be noted that several posts in the GalCiv2 forum by Stardock programmers have specified that there are 0 damage "rolls" possible - the "1 point minimum" stated above isn't right.
[edit] Come the Defenses
The above comments are based on defenders lacking any defenses; spreading weapons becomes a liability against ships with defenses. Let's look again at the example of the 4/4/4 ship vs. the 12/0/0 against a defender with 4/0/0 defense. You would think that having weapons that are up against sub-optimal defenses would produce more damage, but let's do the calculations. The 4/4/4 would have two 4 point shots against the square root of the lone, suboptimal defense factor. These shots would average 1.0 damage (avg. of 1-4 = 2.5, minus average of 1-2 = 1.5) each. The third attack would average no damage (avg. 1-4 = 2.5, minus avg. 1-4 = 2.5). The total average attack comes out to 2.0 damage. In the case of the 12/0/0 ship, it attacks against the 4 defense (avg. 1-12 = 6.5, minus avg. 1-4 = 2.5), averaging 4.0 damage, two points higher than the total for the the spread attacks. But, from the above discussion, we would have expected the attacker to do on average one point MORE damage by spreading out his weapons. So what happened?
What happened is that each weapon attacked separately, giving the defenses three chances to act, rather than just one. This works in a manner analogous to offense: each additional weapons technology used in the attack will produce a defense with a minimum "roll" of 1. Also, the expected defense points (the sum of the defender's rolls) increased because each defense technology gets its own roll. Even though you are going up against suboptimal defenses, they are still square-root effective, which is better than nothing.
So diverse weapons drops your total damage for each weapon type you employ, proportional to the defense factors of the defending ship. Since having multiple weapons only bumps your average damage by 0.5 points per additional class of weapon, you actually lose a MINIMUM of 0.5 damage point per weapons class above one against a defended ship.
Take a 90/0/0 ship and a 30/30/30 ship vs. a 9/0/0 defense ship. The 30/30/30 will have two 13.5 point attacks(avg. 1-30=15.5 minus avg. 1-3=2.0) and one 10.5 point attack (avg. 1-30=15.5 minus avg. 1-9=5.0) for a total of 37 points, while the 90/0/0 will have one 40 point attack (avg. 1-90=45.5 minus avg. 1-9=5.0). You lose three points for having three weapons types. Obviously, if the defender had guessed wrong about which weapon you would be using for your single-weapon whip, the unbalanced attacker would do even more extra damage (5 points vs. 3).
Other notes:
- Because defenses still defend against weapons other than the one they were designed to counter, having ANY defenses at all automatically produces this effect.
- Naturally, having multiple weapons will narrow the distribution of defender's rolls in a similar manner to what it does for the attacker's, resulting in more predictable outcomes for combat overall.
- As defenses on ships become better, having more types of weapons becomes more debilitating. If your average damage is less than the the sum of the square roots of the two highest defense ratings plus the lowest rating, you won't be generating enough damage to even make it through his defense. At this point, concentrating on one weapons type is the only way to "burn through" the defenses.
- The above averages are good approximations, but not quite exact. A ship with weapons 4/4/4 attacking a ship with defense 4/0/0 actually does 0.625 beam damage, 1.125 mass driver damage, and 1.125 missile damage on average per round of combat, for a total of 2.875 damage. (The reader can easily verify this himself by summing up all possible combinations of weapons rolls and defense rolls, i.e., beams roll 1 & defense rolls 1 = 0 damage, beams roll 2 & defense roll 1 = 1 damage, etc.) A ship with weapons 12/0/0 against the same 4/0/0 defense would do 4.208 damage per round of combat. However, this doesn't change the conclusions of this section, it's only a slight tweak to the numbers. Similarly, doing the exact calculation for the example with 30/30/30 weapons attacking 9/0/0 defense, this does 10.944 beams damage, and 13.544 damage of each of the other two weapons types, for a total of 38.033. The 90/0/0 weapons against 9/0/0 defense totals 40.648 damage.
- For another thing, defenses, especially if you research them later in the game, take much less time than the corresponding weapons techs. Sometimes you can get half a dozen defense techs researched in 1 turn. this makes it so 4/4/4 defense can be achieved in a few rounds, while it may take many more to reach 4/4/4 attack.
In summary, spreading weapons type will therefore only be numerically advantageous in the opening stages of the game, when few ships have defenses.
[edit] Defense Technology
Shields, armor, point defense... or just pile on more weapons?
The answer here is that once again, it depends. The biggest consideration here are your opponent's capabilities. Notice that by the formula described in the defense section, the total defense against any given attack is calculated as optimal def + sqrt(subopt defense1) + sqrt(subopt defense2), where the optimal defense is shields vs beams, armor vs mass drivers and point defense vs. missiles. As a result, installing a module with defense 1, regardless of its type, will have the same benefit against every type of attack. Additionally, having def value of 1 against each missile, mass drivers and beams is equivalent to having a def of 3 against any attack. This method of "spreading" defense begins to degrade as more modules are installed, however. But we digress. The real question is - does defense make a difference? From my experience, this depends entirely on whether you are fighting a lot of smaller ships or a few bigger ships, and also on how advanced your enemy's weapon capabilities are. Here are some observations I've made so far:
Defense is important in early game, but becomes less so in mid and late game. This is because as weapon tech and miniaturization advances, it becomes easier and easier to kill ships in one shot (since HP of ships remains fairly low even for larger hulls).
FALSE: Maybe you can dispatch tiny vessels with one shot, but:
a.) Experienced crews have gained more HP
b.) Large ships have HP amounts of 50+
c.) One shot kills are no longer as important, as the defender is guaranteed at least one shot.
The highest shot I received in a late game session was 13 dmg. One large ship of my fleet can dispatch singlehandedly 7 small ships of the Iconians, which are as technologically developed as I am. Why ? Because I fitted the ship with two Barriers and the smaller ships cant scratch it with their Beam weapons. Sorry, but I fail to see how 50+ is low Hitpoints for a ship, and thats not even a Huge ship.
Defense allows three large ships to destroy a fleet of 7 tiny ships with relatively low damage outputs.
Defense is not nearly as important against large, heavily armed ships unless you can raise the appropriate defense value really high.
In general speed is preferable to defense once your ship's total attack value is 18+. Why 18? because that is the default number of hp for a medium ship, and because two ships with an attack of 18 or more would be expected to kill an unshielded medium ship with no retaliation. Furthermore, given an offensive and a defensive tech of the same level, the shield will take up slightly more space than the weapon module while providing equal or lesser defense (usually lesser as tech advances).
- Note: As of 1.3, defensive modules take less space than similar era weapons.
As a hypothetical example, let's simulate a battle between two fleets of 3 medium ships each. Fleet 1 will have ships with an att of 8 and a def of 0. Fleet 2 will have ships with att 5 and def 3 each, and for the sake of argument we will assume that the 3 def is of the correct type. We will assume that each round, the average damage is inflicted, and the shield also perform on "average"
Assuming fleet 1 gets first strike: Round 1: Fleet 1 inflicts 13.5 avg damage, 4.5 of which is negated by shields, for a total of 9 dmg. Fleet 2 retaliates for 9 dmg, none of which is negated, so thus far we are even.
Round 2: Fleet 1 once again inflicts 9 dmg, which destroys one of the ships in fleet 2. Fleet 2 retaliates for 6 dmg, so fleet 1 still has 3 ships in it.
Round 3. Fleet 1 yet again inflicts 9 dmg. Fleet 2 inflicts 6 more, which destroys a ship. Now fleet 1 has 2 pristine ships, fleet 2 has 2 ships, one of which sustained 9 dmg.
Round 4. Both fleets do 6 dmg to each other.
Round 5. Fleet 1 does 6 more dmg, fleet 2 loses one more ship. Fleet 2 retaliates for 3 dmg.
... Round 9. Fleet 1 inflicts 6 more damage, destroying the last ship in fleet 2. Fleet 1 survives with one pristine ship and one ship that has sustained 15 dmg (and has 3 hp left).
Moral of the story - first strike determined the winner in this case. In theory two identical ships that have att + def values being equal will inflict the same average damage. However, a ship with a high attack value could get "lucky" and destroy the defended ship outright, whereas the ship with high def could not do that.
- Note: There is no longer a first strike. Defending side now gets a guaranteed shot, even if the ship is destroyed that round.
Defense does make a difference, however, when one large ship goes up against 4 or 5 small ships. For instance, let's consider a fleet of 4 tiny ships, each with an attack of 8, going against a huge ship with an attack of 16 and a defense of 8. On paper, the fleet of 4 tiny ships will actually have an advantage over the huge ship - it will be listed as att 32 def 0 fleet going against an att 16 def 8 fleet. However, that 32 attack does not come in one large shot, it comes in several small shots. As a result, each tiny ship will on average inflict 4.5 - 4 = 0.5 damage to the huge ship, while the huge ship will on average do 8.5 - 0 = 8.5 dmg per shot. As you can see, this fight will almost certainly be a no-contest win for the huge ship despite the apparent attack superiority of 4 tinys.
To conclude, defense is useful if you expect to be hit by a lot of low-damage shots, and not nearly as useful against large concentrated attacks. In general, it is also more useful to have first strike rather than higher defense.
- Note: If your racial bonus for defense is higher than weapons, it may make defenses more efficient.
[edit] Armor Points - Perfect Squares - Even Distribution
It appears to be optimal to have the armor on your ships be a perfect square. Because "non-optimal armors only use square-root", building 2 or 3 points of armor is silly. Having 1 or 4 points is better, because the 4 points is effectively 2 when used non-optimally. In the early game, I found that adding a single point of armor to my small ships is a big benefit.
Also, it pays to have 1 point of each. That gives you 3 effective points of defense. So, it pays to spread your armor tech spending evenly. It pays to focus your weapons tech spending; I like lasers because they're the most damage/space.
Note: Should rounding rules take place after the square root (eg: sqrt(3) = 1.7 rounds to 2), the actual optimal defensive numbers would be defined by the equation X^2-X+1 where X is the target you want to round up to.
[edit] Military Starbases and Ship Design
It's worth taking a look at the effects of a military Starbase on your ship designs, as these beauties can really swing the space battles in your favour.
If you have a military Starbase with a defense-assist module, this adds +x defense to any of your ships within its radius. However, this only applies if the ship already has at least +1 defense (in any category). This makes it extremely worthwhile to add at least 1 defense to all ships that you'll be using within range of such a starbase. The attack-assist works in a similar manner, in that you must have at least +1 attack (in any category) for the bonus to apply.
Because this is a flat bonus and not a percentile one, the effect on small ships is much more significant than on larger ones. As fsk+ notes below, adding +1 armour is very effective in the early game - combining it with a starbase defense assist module will make even a small or medium ship a really tough nut to crack. I tend to focus my starbases on +defense, add just +1 defense to my ships, and then use the rest of the ship space for adding weapons, engines and utility modules.
[edit] Ship design with new combat rules of Dark Avatar
There are actually only two feasible ways to build ships in DA: all-attack one and most-defenses one.
All-attack means only best weapons and engines (0-2) in biggest hull available. You'll be losing some of those ships, but by "in case of mutual destruction the strongest ships survives" rule (and this design will probably be the strongest), not so many you'd expect.
Most-defense means your single ship has to have at least 75% of defenses as your opponent's fleet has attack (ofc. higher is better, and defenses have to match his weapons' type). If you can match that, your ships will take in combat only little damage, but will destroy lots of opponent's fleets. However timeframe of usefullnes for this design type is limited to early and mid game. With opponent's very high weapons and logistic tech your main fighter should be all-weapons design.
