Saturday, March 3, 2012

Mailbag 3: Starmada Ship Design

Got several hits recently for the search term "starmada ship design".  Initially I kind of ignored them, but upon further reflection I realized that there's actually a lot to say on this topic.  There are several schools of thought, a lot of crunchy math to be done, and plenty of personal experience to draw on.  Plus, if I queue up a bunch of posts about Admiralty Edition, maybe Nova Edition will come out, by the Umbrella Principle (looking at starting an AE Simplest Campaign for the same reason, actually.  Well, that and the fact that I'm kinda jonesing for 'mada...).

So much for motivation.  While reflecting on this topic in the shower, I realized that there are essentially three types of 'traits' in Starmada: AE (well, kind of five, but two are degenerate).  The three interesting ones are Reliable traits, Swingy traits, Tactical traits.  The degenerate types are Awful and Awesome traits.

Swingy traits vs Reliable traits are fairly clear.  A swingy trait can perform strongly, but relies on luck to do so.  Examples of swingy traits include Fire-Linked, Variable RoF (and similar), and high damage ratings.  On average, you'll get exactly what you paid for by expectation, but the actual results vary wildly across games.  Reliable traits include good Acc values (3+, 4+), high RoF, Fire Control (when used to power Directed Damage), and Extra Hull Damage.  These are traits that are always good, and work to mitigate the effects of randomness on the way that you play.  Often you pay a premium for this reduction of randomness, but not necessarily.  An illustrative contrast is between Extra Hull Damage from Core, and Catastrophic from Rules Annex.  Both traits work to increase hull hits per die of damage, but they do it in different ways.  EHD guarantees a hull hit on every die of damage, even if you roll a 1 against an Armor-Plated target.  That's about as reliable as it gets.  Catastrophic works a bit differently.  Whenever you roll a hull hit with a catastrophic weapon, the target instead takes 1d6 hull hits.  Scary, and can lead to instant obliteration of a hull 6 or smaller target from a single die of damage...  but not reliably (only 1 in 12 times, or 1 in 18 against armor-plated targets).  Catastrophic is the poster child for swingy traits.  Another example is Countermeasures versus Armor Plating.  Countermeasures provides a nice, reliable -1 to all weapons fire against you, unless they have Fire Control.  The effectiveness of this actually varies based on the accuracy of their weapons, but it's never completely ineffective.  Armor Plating negates damage rolls of 1.  If your dice like to roll 1s, Armor Plating's great; it can negate huge piles of damage (but you might want to use different dice for firing...).  On the other hand, it's entirely possible to play a game with a fleet where every ship is Armor Plated and have AP negate exactly 0 hits.  It does happen.

Example from play: Matt fielded Catastrophic weapons extensively during our play of Admiralty Edition.  Most of the time, they were scary but not terribly effective, but there was one battle (the Dreadnought Match) where they really came into their own and just tore Alex to pieces.  Their primary value most of the time was really as terror weapons; in trying to stay out of the Catastrophic arcs, you might go places on the map you otherwise wouldn't, or you might split your fleet, and you often ended up staring into his Piercing +2 secondary guns, which were reliably good against targets of all types.  Because you really didn't want to try your luck against Cata...

Tactical traits are a third point on the triangle.  These traits are great if you have a good ability to predict your opponent (to borrow from Sirlin, they're powered by Yomi), and a good way to screw yourself over otherwise.  Cloaking is the canonical example, but dual-mode weapons, slow-firing weapons, and Screens are also tactical.  You almost always pay a premium for tactical traits, which is why you have to use them well for them to be cost-effective.  However, they're a lot of fun, and used wisely can be really, really good.  Arguably the Range-Based Foo traits are also tactical, though the Inverted versions can be used at a profit pretty reliably by 'dancing' at long range.

Example from play: Ah, Cloaking.  I initially started experimenting with Cloaking because by Eldar conversion to Starmada was performing terribly; speed and short-ranged weapons were just not cutting it.  Cloaking, however, gave them a whole new dimension and sent me on a winning streak; I don't recall ever having lost a game with the Cloakdar, though I did draw at least one (I know I did during their first game).  Cloaking let me choose the time, place, and manner of engagement, and let me focus my fire on small parts of the enemy fleet while avoiding retribution from the rest.  Matt and Tim avoided cloaking just based on the amount of work involved, while Jared tried it once but badly mispredicted Matt's movement and ended up right under his guns.

Awful traits are pretty self-explanatory.  They're traits which are just...  bad.  Mathematically, provably bad.  There aren't a whole lot of these in the game, fortunately, since Cricket did a pretty good job with the balance.  The two that I know of, unequivocally, are Anti-Fighter Batteries from the Imperial Starmada Sourcebook and Halves Shields from the Rules Annex.  AFBs are far too expensive for the degree of protection from fighters that they provide.  They're nice, because they can fire in a 360-degree arc, but range 1, 5+ to hit, and RoF 1 are all bad traits for anti-fighter weapons.  You want to hit fighters far away, and you want to hit them a bunch of times.  Further, AFBs boost both your ORAT and your DRAT, whereas a normal anti-fighter weapon would just boost your ORAT.  Since ORATs are usually higher and pricing works on square roots, the DRAT cost of AFBs just isn't worth it.  Halves Shields is the other flavor of bad.  It's actually pretty good...  but Piercing +2, also from the Annex is as good or better against all target classes, for the same price.  Piercing +2 is better against Shields 3, and exactly the same effectiveness against all other shield ratings.  Thus, there is mathematically no reason to ever use Halves Shields if Piercing +2 is available (ie, not banned).

Awesome traits are things which are too damn good not to use if available.  They include traits which are mathematically super-efficient, as well as traits whose only effective counter is those traits themselves.  The contents of this category are naturally contentious.  Generally-conceded traits here are Starship Exclusive and Ammo.  Other candidates for this category include the various shieldbreakers from the Annex, strikers, flotillas, repeating, G-arc weapons, and increased hits.  Personally, I tend to think that Stealth might be a bit too good too, but there is at least a kind of counter for Stealth (namely Inverted Ranged-Based traits).

Example from play: Ammo is disgustingly good.  I don't think we ever played a game where only one side used Ammo and that side didn't win.  My ammo-based disablers crippled Alex in the first round, Jared used ammo'd polecats from Hammer and Claw in his first game and won (one of a very few first-game wins), my 100-point one-shot torpedo boats annihilated Matt's 300ish-point battleship in Planetary Assault, and Jared used a one-shot area-effect shieldbreaker against Matt's short-ranged fleet for a bunch of shield damage in the opening round, which I believe turned into a win.  The trick with Ammo is that it lets you concentrate firepower across time.  For the same SU cost, you can have one weapon with no ammo limits, or five such weapons, each of which can fire exactly once.  Thus, you can get 5 times the firepower in the first round of firing.  This is the so-called "Glitter Fleet" idea you hear about on the forums.  It might be a bit more expensive in terms of CRAT...  but 5x firepower is more than enough to make up for it, and usually to wipe an enemy fleet.  And if it doesn't, you cede.  For bonus points, put Slow-Firing on your one-shot weapons.  No effective penalty for you, but makes 'em cheaper.

So at this point you're probably thinking, "OK, that's nice that you've split traits into five categories, but largely non-useful for me, the ship designer."  Not so!  Your use of these traits determines in large part how your fleet will play.  A fleet running strictly reliable traits will perform consistently, while a fleet full of swingy traits will sometimes win big and sometimes lose terribly.  A fleet with too many tactical traits will overburden the player's ability to guess and micromanage, and will likely lose unless those traits synergize well (Cloaking + Slow-Firing, for example), but one with one 'gimmick' tactical trait can play well and with a distinct style.  Awful traits should be avoided, and awesome traits should either be used universally or banned.

To build a fleet which is mechanically strong, you must understand which traits (and which combinations of traits) are mechanically strong.  To build a fleet that you will enjoy playing, you must know both how you prefer your fleets to 'feel', and how various properties of your ships influence their qualitative performance.  The best way to learn these things is to play fleets which vary across the reliable / swingy / tactical scale and see which you like most.  I tend towards tactical / reliable, while Matt for example likes to use swingy main guns and reliable secondaries.  That sort of thing.

There's a lot more to say on this topic.  In particular, I'd like to make clear the math behind Repeating, varying Acc values, shield-breaking traits and shields in general, range and speed, and damage and degradation (including flotillas).  This may have to become a regular column (until Nova comes out...  guess I'll just start talking about Nova when that happens).

No comments:

Post a Comment