Most ignore vector movement, and fighters travel faster than capital ships (ignoring the fact both are in the same medium - spacecraft are not a direct analogues of water-based ships and aerial fighters). I'm not interested in realism - I mean, sci fi is an invitation to be inventive and use handwavium.
So why not be creative? Why do we get 101 variations of the same game?
Another is the relative boredom of space terrain (or lack thereof). Most tables are rather bare and featureless - with perhaps the odd sty foam planet to add interest. I've got two issues with this - one, it seems rather silly that two ships would fight over empty space. Presuming some sort of FTL as the norm, why would they bother? Most hyperspace systems in games and literature are not easily interdictable. If ships must use jump gates (EVE Online, The Lost Fleet) then there is a chokepoint. But really, why would not fights take place where they matter? I.e. near planets, in mineral-rich asteroid belts, near starbases etc? Secondly, more damningly, a bland open space table is boring from a gameplay perspective. Without terrain to dodge in and around, the tactical possibilities tend to be "go closer" or "move further away." It's not surprising many space games end up as a dice-chugging brawl in the middle. Contrast the effect limited space terrain has compared to a crowded Infinity table. There are far fewer decision points.
The upcoming Dropfleet Commander is primarily over planets. I'd like to see where it goes with its hi, medium and low orbit altitude bands. In addition, fights revolve around landing troops on the planet, adding more gameplay choices through victory conditions like orbital bombardment and ground combat. Furthermore, the detection rules add a different dimension to combat. Whilst I am hesitant to proclaim it as the messiah, it at least seems an interesting step away from the same-old-same-old. (If they don't dumb it down too much)
So where is this leading? Well, this train of thought about spaceships was triggered by the PC game Dreadnought, also yet to be released.
Basically, you have huge capital ships flying entirely through atmosphere. It's rather like aeronef with a sci fi bent. Whilst it does not exactly promote vector movement, and is indeed pretty much WW2-with-flying-sci fi battleships it does a few interesting things, and there is an impressive sense of scale often absent in the depths of space.
Distinct classes - every ship has a role
Many space games tend to simply have ship classes that are smaller versions of the bigger ones. Battleship = the cool, meaningful "hero" class ship
Cruiser = merely a smaller battleship with less guns and hitpoints
Escort = a tiny battleship that acts as ablative armour for bigger ships and gets swatted down in the first turn.
In contrast, Dreadnought deliberately copies RPGs and MMOs to give ships flavour.
Dreadnaughts - have armour and broadside firepower. A "tank" capable of dealing and taking hits.
Destroyers - can be customized with energy management aka "jack of all trades"
Corvette - has a cloak, hit-and-run, fast strikes "assassin" with tactical warps
Artillery cruiser - high power, long range, slow cannon "sniper"
Tactical cruiser - weaker weapons, but can repair others "healer"
I'm not saying other space games don't differentiate between ships and classes, but I am suggesting space games seldom sit down and put major effort into designing differentiating and mutually complimentary ship classes which are equally valuable to the team. Rather than simply making small ships attractive by virtue of throw-away cheapness in a points system.
Tactical Warping - aka emphasizing a single sci fi feature
Although I have seen this before in tabletop games (Battleshift?) using short hyperspace jumps to teleport and leap directly forward short distances to engage/disengage adds an interesting flavour to combat and gives it more sci fi flavour. Making smaller ships better at this would give them a distinct advantage. Many sci fi games of all genres start with a basic premise (i.e. hard sci fi platoon games are usually Vietnam-in-space) and follow the basic tactics and feel of that genre. The "sci fi" is just a few cool toys and special rules tacked on that do not actually alter the fundamental gameplay that much. However, picking a single sci fi technology and making gameplay revolve around it will change the core nature of the game significantly from historical games from which it will inenvitably be derived.
To sum up...
I like space games that feel different than the usual "WW2-in-space" or have an interesting twist on the space theme, such as the atmospheric combat in Dreadnought and Dropfleet Commander. Giving all ships a distinct and meaningful role, instead of making small ones simply weaker copies of larger ones, will lead to a more interesting game. Picking and emphasizing a sci fi feature (such as tactical warping/teleporting) and making it a core game feature will further give a game a distinct flavour. Finally, terrain is important, if for no other reason than the tactical interest and decision points it adds.
