Sunday, July 24, 2011

A Farewell to Starships

Jared, Eric Faust, Matt Glisson, and I played another game of Starmada last night using ships from the Imperial Starmada Sourcebook only.  Glisson and Jared had Imperials, while Efaust and I had Arcturans; 1200 points per side, straight-up 'to half VP' around a small planet with four moons.  We played this particular fleet matchup in response to a post here, which claimed that the Arcturans were underpowered compared to the Imperials.  Considering that in previous games, the Imperials had not pulled their weight, I found this surprising, and so we set out to test this hypothesis.  Upon a reading of both fleets, though, things do look skewed towards the Imperials - while the Arcturans do have 4+ weapons (with the Imperials stuck at 5+), they lack Fire Control and most of their 4+s are slow-firing.  Likewise, the Imperials all have Countermeasures and Fire Control, giving them a distinct advantage at long range (their 5+s were at 5+ to hit at long, while our 4+s were at 6+ to hit at long).  Finally, their close-range anti-fighter weapons were simply superior to ours in volume of fire, as they had a comparable number of mounts in any given arc, but with RoF3 compared to our 1.  Our lasers had Anti-fighter, but all of their ships had Fire Control for an equivalent effect.  The only things we had going for us were Armor Plating, larger hull sizes, and higher shields; it looks like the Arcturans have to get in under the Imperial long-range advantage to really have a shot.  So that's what we tried to do.

The Imperials deployed a few ships in the moons, but kept most of their fleet out at long range on the right edge of the map.  We pushed into the moons and destroyed their ships there, scoring about 290 VP of the 600 we needed to win, at a cost of 55 VP scored for them by destroying one of our independent fighter flights, and some damage to our carrier.  We were than at an impasse, though - they had no incentive to advance and fight us in the rocks, because they would lose their long-range advantage, and we had no incentive to run out into their guns.  We did anyways, though, because sitting around plotting movement orders of 0 until somebody breaks is terribly uninteresting.

We advanced through the cover of the moons to as close as we could get, but the carrier was out of cover for one turn and was mauled badly; it wasn't destroyed, but it lacked sufficient engines to actually make it to the front in time for the assault, and was worth enough points that we couldn't risk leaving out in the open for fire support, so it hid for the rest of the game.  Our fighters crippled one of their cruisers, but were then obliterated by their close defense cannons, and then we charged.  Emergency thrust was deployed to debatable effect; while the boosts did help close the gap from long to medium, the subsequent (higher than expected) engine damage was problematic later.  We had awful luck with our firing, with only one of our four ships hitting with any weapons.  In response, we took around 30 points of damage, though after damage allocation all of our ships were still functional as a result of armor plating.  However, between an unusually high numbers of 2s rolled and the emergency thrust penalty, we were dead in the water, running on Engines 2 for most of our ships, and our Slow-Firing weapons were shot for the next turn.  As a result, we ceded, as the sheer volume of close-range CDC fire they could bring to bear on our sitting ducks next turn would have been almost-certainly sufficient to win the game for them.

Conclusion: Hypothesis confirmed.

Secondary conclusion: In Starmada, ship design > luck > tactics.  Good ship design can protect you from luck (with 3+ or 2+ ACC, high rates of fire, Fire Control, things like that), and as long as you don't do something terribly stupid tactically, good ships can generally annihilate mediocre ships and win the game by VP.  With less good ship designs, where you start seeing 5+ ACCs and low rates of fire, luck trumps tactics - you can play as well as you like, but the dice will still keep you from winning.  Tactics come in a distant third as far as determining factors of victory go, in our experience.  When 'tactical decisions' do arise, the optimal course is typically to maintain range if you have an advantage over the enemy at range (as a result of ship design), or to hide behind terrain from the enemy's long range guns if you don't (as a result of your own ship designs).  This results in the kind of deadlocks that we keep seeing - short-ranged fleets take and hold the blocking terrain, then do nothing, while the longer-ranged fleet sits back and does nothing from the start.  There are both really dissatisfying outcomes.  And sure, there are counters for this type of thing - ship design counters include cloaking, stealth, and tons of engines, while tactical counters include evasive action and emergency thrust as means by which a short-range fleet can try to take the offensive and do its thing.  In our case, though, Evasive Action would have been no good - we'd've given them -1 to shooting, but since we would've ended up in their medium range bands and they had Fire Control, we just end up inflicting a -1 to our shooting and halving our speeds for no benefit.

So I guess that's a systemic criticism of Starmada for us; from the very beginning, from my first game against Alex, short-ranged fleets have perished to even slightly longer-ranged guns.  And there is very little we can do about it, and it makes things less than fun.  The current state of the ship design metagame here is "ships with as much forward firepower with range 15 and Inverted Ranged-Based Something as possible", because that's what wins.  And it's kind of a boring place to be in; I try to liven things up with short-ranged cloakers, but everybody knows that I do that, and I'm pretty easily countered by fighters (can't bring fighter cover with me, or it gives away my position, and can't fire anti-fighter weapons before enemy fighters get a chance to maul me after I decloak.  I've been using Point Defense, and it helps, but it's not nearly enough to save me if somebody decides they're going to launch a cloud of strikers to surround their ships and wait for me to come out of cloak).

In short, I think I'm done with Starmada for the most part.  Jared seems to agree.  It's been fun, and it's been really interesting to watch the ship design metagame evolve (from "fairly innocent" to "oh god so many strikers" to "all guns forward"), but...  yeah.

On a happier note, Glisson suggested StarGrunt II yesterday once we were done deconstructing our problems with Starmada.  I've been reading it, and it looks interesting...  still has the potential to be shafted by dice, but I really like their stance on point values for units (namely that there aren't any), and unbalanced forces / secret objectives.  I'm hoping we'll be able to muster a game of it before the summer's out.


  1. It SHOULD (and you have far more Starmada 'time in helm' than I do) come down to a Rock, Paper, Scissors game where Rock is Long Range, Paper is Carrier / Fighter, and Scissors is Short Range. If your fleet is dependent on Long Range weapons, a Fighter based fleet should beat you. To beat the Fighter based fleet a Short Range fleet is needed, and to beat a Short Range fleet a Long Range fleet is required. Of course no one should field a fleet 100% dependent on these three facets, leaving some room for variation and the outcome in question. Tactically, knowing which of the three you are playing, Rock, Paper, or Scissors, determines your tactics and you have to play them that way. If you find that Rock beat Paper AND Scissors then either the game system is failing or there is something you've missed in design, tactics, or just luck.

    1. In our observation, the categories you're using (Long Range, Short Range, Fighters) are much too general. Fighter-heavy fleets lose to anyone with dedicated, effective anti-fighter weapons, be they short range (as we typically think of point defense weapons) or long range, as employed by BeowulfJB on the Starmada forums. Long-range anti-fighter weapons are strictly more effective than short-range ones. Strikers, on the other hand, are only countered effectively by launching your own, less expensive interceptor strikers; any other option (except perhaps the Point Defense trait) is simply not cost-effective against them.

      These are just examples, but they illustrate the flaws in your model and the lack of tactical choices; if you're up against an optimized Defense 2 independent superfighter force and you didn't bring Anti-Fighter Piercing weapons and cheap dogfighters, you lose with high probability. If you're up against a striker saturation fleet with engine hacks (because engines don't effect the pricing of vessels with no weapons, you can put huge piles of them on dedicated carriers for free) and you didn't bring enough interceptors, you lose if they're playing at all intelligently. If you're up against repeater / increased hits 3+ inverted range mods cheese and you're neither superfighters nor strikers, then you're fighting against a terrific numeric superefficiency and you will very likely lose before you get into medium range unless you're using both Countermeasures and Stealth. Effectively, you have to design with your opponent's design in mind, while they're doing the same. The most important lateral thinking problem is not tactical, but in design.

      The problem is not that we missed things in design; it's that we found everything, and that the ship design space is larger than the tactical search space. Only makes sense that eventually we found design options for which there is no good tactical counter. The system fails under the kind of extreme optimization we were doing; I know for a fact that one of us was running Monte Carlo simulations to determine the effectiveness of various weapon configurations. I was running expected value math and generally finding weird edge cases in the rules (engines on carriers, the weird carronade-thrust interaction, and such). We put a lot of strain on the system, and eventually it did break.