Anyway, to details of fixing the skill system:
Between uses two approaches in its various magic schools: each
consists either of a single open-ended skill (eg teleport ever
further), or of an unlimited number of closed subskills (eg
healing an infinite range of "diseases" up to completely). The
second approach fits TBG better, with separate subskills for most
activities.
So instead of getting better (once) at all engineering tasks by fixing a warp drive, you only get better at fixing warp drives - but keep getting better indefinitly. This fits well with open-ended tech levels as any amount of bonuses on fixing are useful for sufficiently advanced modules.
It fits less well into activities such as ore collecting, where the benefits might be asymptotic or based on competition with other players (eg the higher skilled officer might go first, rather than the one on the higher power ship).
The current non-activity skill points adapt fairly easily to this system: academy study gives a bonus to all that officer's activity skills but limited to only a fraction of their unmodified level, and adventures give large pluses to single skills (eg stopping an alien warp drive from blowing up (at the last second) might give +5 to fixing warp drives generally).
Giving a skill increase on each success would be bad, encouraging too much fiddling with easy cases and accelerating progress for anyone who happens to do well early on. Giving an increase for each failure is appealing, but also wrong as it encourages too much trying practically impossible actions just for the increase.
The neatest seems to be giving increases for alternating success and failure (ie your skill goes up on success if it was even, and on failure if it was odd). This probably discourages configuration of ships to make tasks consistently too easy or too hard, unless the player intends to sacrifice long-term skill gain for short-term success at everything by emphasising module factors.
=====================
Moving on to a much more general consideration of skills, actions, life and death, it may not be possible to make all of a huge range of subskills both interesting and open-ended. So it would be good to have a way to include the cyclic solution of officer death as a backup to genuine open-endedness of all the subskills
This was discussed a lot here some years ago, with the biggest problem being the very sharp discontinuity from being completely alive (and probably highly skilled) to being completely dead (and probably replaced by a new ensign with no skills). Various solutions such as replacing with a semi-skilled new officer or having a deputy with some scope for independent action all failed to fit into the TBG-1 model of one officer with one action per turn.
Turning to the original TBG-2, as an expanded and opened out version of TBG-1, officers had 2-5 actions per turn based on their rank, which was mostly a measure of skill. Combining this idea with death leads to a neater scheme that should work better than either.
Keep to one action per officer but allow multiple officers of each type, and kill off the old ones. The recruitment rules that make it work are:
1) New recruits can't be more than half as skilled as any of the existing officers of that type.
2) New recruits (apart from new players' starting characters) always start with significantly non-zero skills.
So initially you get one officer of each type with one action as in TBG-1, and can't recruit because all available new officers have non-zero skills. After playing for a while and getting your original officers' skills up, you find candidates in the hiring halls that you can take on as Second Officers. After further time the Second Officer skills get high enough to be able to take on Third Officers. At some point the old officers die off and everyone moves up a rank.
Other nice effects include:
1) Officers age towards death not purely in turns passing but in stress, eg 1 point for a resting turn, 2 points for an action, 5-10 for severe injury, 1 per turn for a disease. This gives the underworked medical officer(s) more to do in reducing the effects of injuries (using sickbay factors) and healing diseases (using life support factors).
2) There's much more choice and room for good or bad play in allocating tasks between officers of the same type, mixing the requirements for success with the need to train up the lower ranks before the top ones expire.
3) There may be room for temporary assignments, eg moving into a new sector where you have no skills for local navigation or healing the local aliens' plagues, it may be best to sack the lowest ranking officers and take on temps with local skills.
4) Officers can be transferable (as long as the incoming ones meet recruitment rules), like modules, allowing more scope for creative trading.
5) Things like jump movement which were previously separate from officers can now be turned into officer actions as there are enough spare ones to go round. Jumping, and particularly avoiding misjumps, then fits into the general skill system.
==============
Running numbers for officer lifespan, skill growth per turn and number of officers per ship, the doubling requirement doesn't work: it makes the new recruits be unsatisfyingly highly skilled. So scrap that.
The new plan is to take the central idea from Otterman Empire - number of units depends on the player's status or fame - and refocus the game around that. I won't go into all the details covered on the newsgroup where it was discussed in depth, but the basics applied to a TBG setting are:
1) Many actions and positions give status points, particularly successes and anything impressive. Status points decay as a percentage each turn.
2) Rank players by status, and give them more officers for higher ranking, so new starters get the usual 4 and players get a new officer for each 10% up the ranking they go, so there's 13 for being in the top 10%. Officers aren't lost directly by falling down the status rankings, they just can't be replaced when they die.
3) Getting new officers from hiring halls or promoting crew is too easy, they should be acquired as adventure prizes, only if the ship has the status to support them. Players with a spare slot can then choose the type of officer they get next by doing the right sort of adventures, giving another level of load- balancing (eg avoid medical officers if you have enough doctors already).
4) Using status as the main reward and goal, rather than cash or big hardware, allows much better control of players. They can be rewarded for doing interesting things like making first contact with new aliens, or piracy against other player ships, rather than for building mountains of cash to improve their scores.
5) Other areas of the game open out with status, including religious and political offices bringing status as well as power (or even ceremonial offices which give only status), and in-game support for player alliances becomes practical.
6) Unlike cash and hardware, status can plausibly decay each turn, so success doesn't entirely accumulate, but must be renewed in the long run by doing additional interesting things. Status gives more officers, which give more opportunities for gaining status, but the players still have to make the most of those opportunities.
7) The torpedo solution to cash mountains never worked well because players rate useless cash mountains more highly than useful weapons because of the scoring system. Putting scoring onto status and including the major prize of more officers in it should make players willing to turn excess cash into status, whether it's called conspicuous consumption, charitable donations or patronising the arts.
Everything comes out right.
================
It should be meaningful to meet the weaponry officer who is famous for presiding over the mutual destruction of two super-dreadnoughts, or the engineer who made the longest (successful) jump ever recorded.
Status can be recurring to officers (not to ships, as that would allow runaway growth). For example, chaining a dybuk could put an officer in line for a medal, if the appropriate prophet/politician decides to grant it, which would give them a big one-off status plus a smaller status income for life.
there needs to be some safety net like the TBG-1 warp drive of last resort hack. It's not quite enough to give players an extra officer if they fall below 4, as with sufficiently bad play they might have just three high level doctors, busily keeping each other alive for too long.
It's probably enough to give such players a special dumb officer who can't gain or use skills, but can attempt actions at skill==0 until replaced by a real one.
