TBG-200X Design

(End of March 2007 update)

Player Distribution

Player preferences vary from high population density with much interaction, down to completely solitary exploration. This spectrum can be satisfied using a single starting point with infinite space to expand from there.

There is also a need for separate spatial areas, allowing subgroups to escape the starting area (to avoid enemies or simply crowding) without being easily hunted down by powerful forces from the starting area. This is achieved by the existence of a few very long range one-way wormholes, allowing players to choose irreversibly to move out of an area. It's unlikely established powerful players would pursue because they can't come back either.

In the open-ended very long run players may discover a sequence of wormholes allowing return to the starting point, but this can also be negated by finding more wormholes leading out of the ring.

Big Themes

The modular starships of TBG-1 remain a good mechanism allowing growth and reward while requiring some skill to operate, but they are too static: they encourage accumulation without needing active management. The ship should instead be an economic system, with inputs from ore collection and trade, and outputs in player actions.

Powerful starships are made difficult to manage by needing a range of different types of ore, without also becoming able to use their power to trivially harvest those ores. This works by having each star system produce very few different types of ore (usually one), but having different modules need different type of ore. The ores are relatively bulky so players can't stockpile enough for hundreds of turns: they must collect and/or trade for the supplies they need to make their ship function.

Modules have tech-levels as in TBG-1, except that level 1 ("improvised") is the new-starter "demo" level with the special property that all types of level 1 module use the same ore (the readily available X1). Second level modules need level 2 ores specific to their type: warp drives need A2, impulse drives need B2 and so on.

(Revised February 2007)

TBG-1's concepts of module reliability, repair and maintenance are all replaced by "damage", a single measure of how far a module is below perfect condition. Each module has "cards" representing its ability to function, from five when in perfect condition down to zero when completely broken. Damage is inversely related to cards, in that taking damage is the same as losing cards and repairing damage with maintenance is the same as gaining cards. /* All officer actions that use a module type have an action cost, which is met by the product of the officer's relevant skill and the amount of damage applied to the modules (as in "The engines can't take much more").*/

Maintenance is no longer an officer action but happens automatically when the ship has suitable ore for a module with least one turn of damage to repair (i.e. fewer than five cards). Each module of the type being maintained uses a unit of the appropriate ore and damage is reduced by one. If ore for some of the modules is not available then the modules with most damage are maintained first.

(end of February 2007 revision)
 

To collect ore costs an engineering impulse drive action. For all ores except X1 the player also needs a comet exploration secret showing the material available in comets in their star system. Each secret shows a specific amount of material available, decreasing as it's collected until exhausted. The type of ore a system contains depends on its spectral type so is detectable from a distance and tends to be clustered.

To discover a new comet exploration secret the player uses a science officer sensor action for prospecting, choosing the level (n) of the raw material they want to find. The level actually found, if any, is random between 2 and 2n (eg 2 = 2-4, 6 = 2-12) with a bias towards the middle of the range (equivalent to rolling two Dns and adding results). Finding a particular raw material type secret reduces the chance of finding more of the same type in the same system for a while.

For example, a player with impulse drives of tech levels 3 and 4 would need B3 and B4 ores to maintain them. In a system with a blue star of spectral type B they could explore for level 2 (2-4) with a 50% chance of finding B3 and 25% of finding B4, or go for level 3 (2-6) with a 22% chance of B3 and 33% of B4. 

(added February 2007)

Cards

Many game mechanisms that were separate in TBG-1 and earlier 200X versions are combined into the large single mechanism of cards. Cards represent the player's options, and deciding which to play and in what order make up much of the player's actions for each turn.

The types of card are: module (meaning allocate one use of that module to the next action, taking one damage by using up the card), action (conditionally do something such as fire weapons, mine ore, change range etc.), and specials (includes spell casting, options derived from political or religious office, or any other things that fit the card model).

Players use a suitable UI to drag the cards they wish to play into the order they want (their "deck"). The action cards are writeable and filled in with a suitable UI allowing conditionals as complex as the player wants. Any number of writeable action cards can be made and used, but they take up space in the deck, which has the drawback of delaying the effects of later cards.

Competitive player activities such as combat (and most peaceful ones) are resolved by their cards taking effect in an order that depends on their order in the decks and how quickly each deck is being executed.

The time taken to play all the cards is split into tiny phases called "ticks", and the time taken for a player to play a card is a number of ticks equal to the mass of their ship (modules + ore / n for some value of n) plus their current rad level (reflecting the health and therefore efficiency of their crew).

Executing a module card has no immediate effect. Executing an action card has an effect depending on the relevant module cards already played and not already used by an earlier action. For example, a fire weapons action at long range uses the long range weapon cards already played and not used, but leaves any short range weapon cards available for a later fire weapons action at short range.

(end of February 2007 revision)

Mortal Officers

Officers age each turn and eventually die. This supports open-endedness and gives considerable significance to medical officers and the related ship components by letting them prolong the lives of other officers.

Officers have an effective age, which increases slowly for the passage of time and more quickly with the stress and damage that the officer suffers. Stress from activities such as adventures can be reduced by use of a medical officer and life support.

Damage includes injuries, which are single increases in effective age that can be avoided by immediate use of a medical officer, and diseases which are smaller increases in effective age that are applied each turn until cured by a medical officer. Sick bays are used in both cases.

Exceptional rewards allow very slow aging of officers, but it's essential there are no rewards which allow officers to live forever, which would likely enable them to become so powerful that they could keep gaining these rewards.

New officers can be recruited as adventure prizes up to a total defined by the player's status ranking. With only one at the lower end of the scale and 10% steps, this allows the top 10% of players to have ten officers at a time. Allowing officers to develop skills by experience tends to fix any imbalance in the usefulness of different TBG-1 officer types.

Skills

(Start of February 2007 update)

The properties needed in a skill system include: low level characters must be able to function at all, each skill level increase must make a useful improvement to everything the skill affects, and high level skills must not make characters outrageously powerful. TBG-1 skills fail particularly in not being open-ended, but also in being mostly linear - until they hit the artificial maximum they improve too much at the top and not enough at the bottom.

Each officer gets as many officer action cards each turn as their skill level, but can only play one of them: the rest are just not used rather than carrying forward to next turn. This gives the asymptotic rewards of increasing skill level, as it's always worth having another card to choose from.

The majority of the officer action cards are simple bonuses in the area of the skill, such as doubling effectiveness this turn of one module up to level n (n most commonly low, and never higher than the officer's skill level).

There are smaller numbers of more powerful cards, such as hacking an opposing ship's computer and preventing the ship taking any actions for the first n ticks of a battle.

There are a very few very unusual cards, such as allowing a huge jump or disabling enemy shields for the whole turn. Although the game is intended to be fully automated, this mechanism could also be used to introduce story arc clues and events by inserting special/unique cards during the game.

The officer action cards overlap with spells in the TBG-1 sense, so it's best to merge at least that aspect of religion into the skill system, broadening it to reward not only practice of the skill but also the behaviours that would earn the favour of the Old Ones.

Status can also be merged with skill/favour, fixing the flaw that using status as the primary reward adds a bias to visible actions over any that are interesting but secret. This leads to a single concept of Experience Level, gained and maintained by earning Experience Points (xps), as a mechanism to reward a wide variety of interesting play with the ability to do other interesting things.

Experience

Officers start with Experience Level 1 and cannot fall below Level 1. Each turn they gain Experience Points for a wide range of things and their Level may be adjusted for their xps as follows:

If they have at least as many xps as their Level they lose xps equal to Level in maintaining their current position. If they have fewer xps than their Level then their Level is reduced by 1 (to a minimum of 1) and they do not lose any xps.

If they then have at least 5 times as many xps as their Level, they lose 3 times as many xps as their Level and their Level is then increased by 1.

Experience Points can be gained for single events such as an adventure or a battle, and for holding position, either office (political or religious) or in one of the ranking tables. Rewards in the tables relate to number of players, say 1 xp for being in the top half of the table, an extra 1 for being in the top 10%, and a third for being the top ranked officer.

The ranking tables differ from some TBG-1 examples in mostly avoiding accumulation of scores. For example, the ranking for number of stars explored as First Contact only includes stars the officer explored first in the last 50 turns, and the star's exploration status is also "reset" after no player has been there for 50 turns, allowing a new player to achieve First Contact there.

Officers can also lose xps for events not suited to their calling, such as Orbital Bombardment for Medical Officers.

(End of February 2007 update)

Following needs revising to match new system above

Officers have skills in many areas which influence their success in appropriate actions. Skills start at level one and can be increased in various ways, with a highly skilled officer being one of the most important assets in the game.

The TBG-1 problem of balancing usefulness of officers, and the need to support more than four classes of skill without adding exotic character classes (e.g. D&D Paladins) leads to abandoning rigid officer types completely. The need to let multiple officers provide benefit in complex activities, particularly combat, while not making these activities impractical for low level players, leads to abandoning the idea of starting with four officers. Instead the player starts with a single character, who they may think of as the captain or their personal character, and all actions that benefit from officers can be carried out at low efficiency without an officer.

Officers increase skills by gaining experience points (xps) in the skill. The main way to do this is by using the skill, which gains xps equal to the module damage cost of the action. For example, an officer with skill 5 making an action that costs 100 would do 20 damage to the appropriate modules and score 20 xps, while an officer with skill 10 making the same action would get only 10 damage and 10 xps. Each 100 xps increase skill level by 1.

All officers start with all skills being at least level one, making them usable and hence improvable. What steers this free choice of skill back towards a more realistic specialisation strategy is that every xp gained also provides a second xp in a skill chosen by the semi-random "pinball machine" system. This works by considering the skills as forming a tree, with a single point at the top, splitting into Engineering, Science, Medical and Weaponry (and others such as Diplomacy and Psionics?) at the next level, then into a more specialised group of skills such as Space Combat or Hacking (not worked out yet), and finally at the bottom of the tree the actual skills such as Jumping.

When an officer gains an xp the program chooses a somewhat related skill to award a second xp in by going up and down this skill tree. Movement starts with the skill actually practiced, initially goes up, and is then randomly up or down (50% each) until reaching a skill at the bottom of the tree again. The effect is that officers tend to get bonuses towards other skills related to the ones they practice, but can sometimes get a bonus in distant skills (e.g. a weaponry specialist may also be good at battlefield medicine).

(February 2007 update)

Crew Health

Crew are not directly represented as in TBG-1 as all emphasis at that level is on the officers. Ships are instead assumed to have a reasonable number of suitable crew and they are signficant only indirectly, in that their health affects the efficiency of all card-related actions.

This helps places the medical officers/modules in the centre of the players' thinking, as the crew health affects too much to be ignored. Unlike the medical "add-on" concepts such as hospital ships and bio-research, this puts the medical field on a par with the other unignorable fields such as engines.

Crew health is measured in "rads", including not only radiation but any other noxious leakages from high tech equipment (particularly weapons) not designed to meet sissy safety standards. Since playing a card takes as long as the sum of the ship's mass and its current rad total, every rad slows down every action and is worth avoiding.

Rads come from several sources including: using modules, combat damage, to be decided (i.e. it's a general tuning mechanism that can be used to flavour or balance activities by making them more or less harmful to the crew).

Each module has a rad cost to use one of its cards. Medical modules cost zero, engineering and science cost one, weaponry costs a variable and generally larger amount. Being hit by weapons fire inflicts additional rads on the target ship, an amount related to the type of weapon used (i.e. some are intentionally "dirty" to wear down enemy crew and paralyse the ship).

Rads mostly (90%?) carry over from each turn to the next and can be removed in two main ways (ignoring spells and other exotica).

Life support neutralises rads that arrive in the same turn after the life support card is played, at the rate of one per tech level. This does not require an action card or an officer: the life support just scrubs out some incoming rads automatically.

Sick Bays can be used with an action and a medical officer to reduce the rads already accumulated by three times their tech level.

(end of February 2007 update)

Movement

Everything in a star system is effectively in the same place, and nothing happens anywhere other than at stars, so movement is between stars (hidden stars such as Dyson Spheres or Black Holes make this harder to see but the principle is the same). Movement in most cases requires a navigation action (science officer using sensors) and a jump action (engineering officer using warp drives).

Wormholes and Stargates provide special cases where no actions are needed. In other cases the costs of the actions depend on the skill of the navigating officer (how often they have jumped to that star before) and of the jumping officer (something more interesting and balanced than how often they have jumped before?).

The cost of reaching a star is also influenced by its spectral type (easier to reach large bright giants, which tend to have fewer interesting contents). Players may establish a beacon at a star and assign it a key number, or capture beacons to destroy them or change the key number: jumping to a star is easier when the player can supply the right key number for its beacon. Navigation secrets also improve chances of jumping to the appropriate star.

Trade

Most TBG-1 trade is done with non-player planets, but most interesting trade is done between players. Trade needs to consist mainly of trade between players, becoming more a matter of diplomacy and cunning than of database maintenance.

The most interesting things for a starship captain to trade are what makes their ships work - the ores, which provide a wide range of items to buy and sell. Unlike TBG-1 modules, the resources are commoditised (for example, 10 A3s for third level warp drives) so can be traded semi-automatically by letting players maintain standing buy and sell orders that execute when they meet others with compatible offers.

Some module trade is also allowed, but only one module can safely be removed from a ship per turn, so module trade consists of offering at most one module for a reserve price and each other ship in the same system can bid for it.

Trade with NPCs is then reduced to the use of relatively uncommon NPC aliens trading one item each (with prices rising and falling on player buying and selling) in order to reward players for moving between these trading posts.

Automatic trading requires a market system which maximises buying and selling while making it difficult to make abusive transfers under the guise of legitimate trade. The market algorithm considers all the buy and sell orders for each type of ore in the same star system, using them both to calculate a price and to decide which orders are executed.

First the price range is calculated to maximise the volume of trade, by considering which buy and sell orders match at each possible price. For a price n, this includes all buy orders for n or less, and all sell orders for n or more. For example, sell 10 @ 5, sell 10 @ 6, buy 10 @ 8 and buy 5 @ 10 makes the range 6 to 8 (volume 15, at prices 5 and 9-10 the volume is only 10). The actual price used is the middle of the range, and the orders executed are all of the fewer of buyers and sellers, plus matching orders of the other sort favouring the higher price for buyers and the lower price for sellers. In this example the price is 7, all the buy orders execute and only 5 of the sell @ 6 order fails.

The ships at each star system form a single market, and all ships at star systems which include Star Gates also form a single market, but with a cost for trading through the Gates of +1 to the selling price and -1 to the buying price of each unit of ore. This extra cash goes to the companies owning the two Star Gates involved (one each). Modules cannot be traded through Gates.

(Added February 2007)

Combat

Combat consists of two ships at a range of short or long with either or both firing weapons at the other, either to steal modules/ore or just to do damage.

Range is first determined at the first attack that executes. (Note that because actions can be conditional many attacks will be conditional on the other ship having already fired (and the UI should warn conspicuously when a player orders an unconditional attack, perhaps by making the background red). For a more scary example, an attack can be conditional on the opponent having already played a weapon card, allowing "retaliation" to get the first strike because the enemy appears to be preparing to attack.)

Initial range is as long as it can be, the assumption is that either the attacker doesn't want to close and opens fire at long range, or if they do want to close then the defender will not want to let them and will be tipped off that the battle has effectively started already.

In practice this means that range is long if either ship has already played more sensor cards than the other has played cloaking cards, weighted by tech level in each case. Otherwise it's short (including the tied case, such as no cloaking or sensor cards played at all). This should be easier to explain than the apparently confusing TBG-1 system, and just means you should play cloaks if you want short and sensors if you want long.

The first attack is then resolved using any weapons already played and in range, against any applicable defences already played.

A ship may attempt to change range by playing an open/close range action. The impulse cards played but not used by each player are then compared and equal numbers (weighted by tech level) cancelled out. If the ship changing range has the surplus range does change and the surplus carries over for next range changing action if any.

Combat benefits from a range of interesting weapons and defences, but needing different types of ore for each type of weapon and defence would be impractical. So the weaponry modules are kept to only two types (defence and weapon) for ore use purposes, but each comes in a wide variety of sub-types with different properties.

In the following table damage for defence modules means how much they absorb, and for weapon modules means how much they inflict: both are per tech level. The "Rads" entry is what they do to their own ship. All weapons inflict one rad on the target for each point of damage that gets past the defences.

Each attacking weapon that's in range applies its hits to activated defences (ones with unused cards already played - they don't need explicit actions or officer involvement) which can affect it, in type order in the table above.

Excess hits (including everything from Quantum Tunnellers) is applied directly to the target modules as card losses. If the attack is against a specific module hits are applied to that module, if it's against a type of module the hits are applied randomly in amounts big enough to remove whole cards. Each card can take as many hits as its tech level to lose one card (= take one damage).

Marines and Commandos are used to try to capture modules: other weapons can't do this and are not used up by "capture" attack actions. Commando attacks are stopped completely if the defending ship has any active turrets, and both marine and commando attacks are stopped if the defending ship has any active shields.

Marine and Commando attacks are countered by defending active Guards, and any remaining damage they inflict is used to capture the target module if they can damage it down to no remaining cards at all.

(Updated March 2007)

Politics

The most interesting political historical period is the Roman transition from Republic to Empire, perhaps more familiar to a modern audience as the first few episodes of Star Wars. In an open-ended game the reverse transition must also be included, allowing the restoration of the Republic so another cycle can begin.

This implies that government must be player-defined, which is achieved by fungible rules. That is, there are not rules for Republic and rules for Empire with a big switch between them, but rules for elements of government and rules for the players to change those rules. For example, if each alien group in the Federation contributes combat pods to its defence, there should not be rules for offices such as Admiral of the Home Fleet or Governor of the Sirius Sector. Instead there should be rules for assigning or removing each alien's pods to an office, and rules for appointing a player to that office. Then players can choose between a pluralist system of checks and balances to prevent tyranny, followed by a set of changes to concentrate power in fewer and fewer offices until one players gains the Imperium.

The aliens-in-pods and favour exchange model simplifies alien tribute considerably, reducing it to a supply of influence each turn which the government distributes to its office holders. They can then use this to buy tax income, levies, secrets, experience points or anything else aliens can provide.

There is a cycle of Senatorial elections, one per turn, so with 24 types of alien each elects its Senator once every 24 turns (starting on turn 25 to allow an initial build-up of players' influence). Election is of the relevant officer of the player with the most votes from the players, weighted by their influence with the relevant aliens. Favours from those aliens on the turn of the election can add substantial bonuses. Each officer can only be one Senator, so officers already holding office are excluded from election to other offices.

Tax income available is one unit of influence from each alien group per turn for each pod it has in play. Each unit can be assigned independently, so players can construct offices concentrating the favours of specific groups (such as Governor of the Vikings), or massing the favours of a type of alien (such as Admiral of the Combined Fleet of the warlike aliens), or spread the benefits widely to each Senator, or any mixture.

All Senators have votes within the Senate, one per tech level of their electors. Senators are ranked in seniority from first to twenty fourth, with newly elected Senators joining at the bottom of the ranking.

The Senate includes four committees, of sizes three, five, seven and nine. Senators are put into committees each turn in order of seniority, from smallest committee to largest. So the top three form the smallest committee (The Three), fourth to eighth form The Five, ninth to fifteenth The Seven, and the rest are The Nine.

Each committee has the power of making a proposal to the rest of the Senate once every twelve turns, spread evenly through the 24 turn electoral cycle and starting with The Three. A proposal can include any of:

The twelve proposals are evenly spaced through the 24 turn election cycle, with one every three turns. Within that three turns the cycle is:

Each alien group has a loyalty and a fear rating, defining its reasons for being in the Federation. Aliens are always members of the Federation, but when neither their loyalty nor their fear level is ten or above the group is in rebellion. When an alien group is in rebellion it pays no influence to the government and its Senator loses their votes in Senate and Committee and drops one place in seniority each turn.

Loyalty and fear are affected by some proposals being passed (see above) and by players interacting with that group's pods. In general, helping them can increase or decrease their loyalty according to the player's political aims. Their fear is increased mainly by players destroying their pods, allowing rebellions to be put down harshly, but that also decreases their loyalty. Fear decreases over time naturally, as well as through player actions.

(End of March 2007 update)

Secrets

There are several different types of secrets, implemented as game-objects which can be created by specific events, shared with other players and finally eliminated by special actions or more commonly by being published, which in many cases has in-game effects. The types of secret are: exploration, political, archaeological, navigation, adventure, religious, doom.

Exploration secrets describe non-obvious temporary assets such as useful ore deposits in comets. They are created by using the Explore action in the right place, and allow the owner to take actions exploiting the secret resources.

Political secrets are created by "bad" actions such as subverting alien governments and smuggling contraband. When published they affect the victim alien race's attitude to the culprit and the Federation as a whole. They can be eliminated before publication by an espionage adventure at the site of the original crime, finding and removing or discrediting the evidence. Players may find these secrets much more useful for blackmail than for actually publishing.

Archaeological secrets are produced in specific places and extend in an open-ended sequence, of which each can only be discovered by a player who has the previous secret in the sequence (or it has been published). Publishing a secret gains status related to how many levels it is above the previously highest secret published in that sequence.

Navigation secrets are specific to each star and make it easier to jump to that star. Publication makes it easier for all players to jump to that star. For open-endedness, the secret and its effect are removed at a random time by shifts in the hyperspace currents.

Adventure secrets are both treasure and enablers for adventures and allow for multi-stage adventures within the same star system. That is, initially the place offers only an adventure with the first secret as prize, but to players who have that secret (by finding it, by transfer from another player or by publication) it offers a second adventure. This may contain the double adventure prize or another secret leading to a third adventure, and so on.

Doom secrets appear as adventure prizes or as gifts from aliens in advance of periodic major invasion threats. This happens at intervals of one to two hundred turns and represents a severe threat to most players from a large fleet of big high-tech hostile ships coming into Known Space from outside. The secrets may give clues on current position of the attacking fleet, starting when it's still ten to twenty turns out. Alternatively the secret may reveal a deathstar-style weakness allowing one of the invading ships to be easily defeated by anyone with the secret. The owning player may publish the secret (putting the information into everyone's results), or may find it more profitable to sell it to powerful players who have more to lose if taken by surprise.

Money

Ideally the game would rely on players to implement their own concept of money, but this is probably too difficult in practice. The second best choice would be to use status as money, but this is so unrealistic in cases such as First Contacts that it would be too annoying. The current plan is therefore to fall back on game-defined money such as TBG-1's energy.

Spare idea: it would be possible to implement treasury bills/bonds as interest paying debt of the Federation government. This is probably bad as it competes with player-run joint stock companies for small investments, but is less interesting.

Another spare idea: use as money a special form of energy (anti-matter?) which has bad effects in large amounts. This is an alternative way to inhibit dreadnoughts, but also a way to prevent stockpiling of cash rather than pursuit of interesting investments.

Adventures

Most star systems offer adventures which an officer of the appropriate type can do as an action. The adventure is divided into episodes which either cost something and provide adventure points, or give a reward and cost adventure points. Each episode can be taken or skipped at a small point cost, so the overall aim is to trade the costs for the benefits via the currency of adventure points (which do not carry over between adventures).

The costs and benefits include a wide range of items useful in the game as a whole and some items useful only in adventures. Players may conduct adventures to gain large prizes at the cost of items they already own, or forgo the main prizes to accumulate more minor items for future adventures.

Accepting risk of injury or disease to the officer is a cost with a large gain of points, available even to starting players with no other items to spend in adventures.

(Start of March 2007 update)

Aliens

Extending TBG-1 aliens to a large/infinite universe causes several plausibility problems. They need to be generic since there are so many species that their characteristics have to be generated. They can't have any reasonable way to join in monetary trade with the players at First Contact since they have no common economy. The lack of realism in the actions and abilities of their ships which confused or irritated TBG-1 players is much worse in the generally more consistent and sophisticated background of TBG-200X.

The card based combat system also complicates alien ships unacceptably, by requiring much more intelligent orders from the AI than the largely random orders used by TBG-1 aliens. And the political system of TBG-200X which includes aliens serving players leads into the quicksands both of fleet actions and of players looting their own political subordinates.

Fixing these problems involves a large change in the nature of the universe and provides several other benefits. The starting point is to make complex life very rare: so apart from bacteria there are only two species in known space, humans of Earth and The Others.

The Others have visited Earth many times in the past, and often took small groups of humans into space and/or gave them the ability to go on their own. Over the centuries The Others have withdrawn from the space near Earth and at the start of the game there are none detectable. They may be involved in long term plotlines, but anything like that is best kept secret for now.

The humans who left Earth have gradually lost their independent ships in the sense that players have ships, and instead all live in standardised habitation pods. Pods can survives indefinitely in any star system, but can also be incorporated into players' ships as modules with unusual powers.

These humans are conveniently called "aliens" as they are distinct from Earth humans in the time of the players, and retain elements of their original cultures. For example, the Vikings are good at combat, the Polynesians good at exploration, the Romans good at engineering and so on. This model keeps the number of "alien" groups finite and makes them immediately meaningful to the players.

Each alien group is coherent in having a single attitude to each player, as a measure of how much they like that player, usually called "influence". This influence is the mechanism of trade and co-operation between players and aliens.

When a player's ship is in the same star system as an alien pod the aliens will ask for a favour of some sort and, if the player grants it, increase that player's influence with the alien group. Similarly, if the player already has enough influence, the aliens will offer a favour, reducing the player's influence if they take it.

Common examples of favours asked by aliens are officer actions, usually supported by some combination of module cards. E.g. "Capture renegade aliens: needs weaponry officer action and 3 levels of marines" or "Cure alien plague: needs medical officer action and 6 levels of sick bay".

Common examples of favours offered by aliens are secrets, such as the location of ores or salvagable modules. A special case of favour is the offer to attach their pod to the player's ship and travel with them. This ensures the opportunity for favour exchange every turn, but also lets the pod act as a module which does not need ore for maintenance.

Combat or other damage to pods while in the player's ship is bad for their influence, as is damaging their group's pods in enemy ships.

Aliens have base tech levels related to the technological status of their ancestral groups on Earth.

Some aliens have strong historical antipathies to other aliens and change favour with a player in response to the player gaining and losing influence with their enemies. For example, gaining +2 with the Athenians can automatically give -1 with the Spartans.

An alien race's attitude to the Federation is influenced by its dealings with players and by their specific actions, such as subverting the alien government to make it more or less favourable (by, for example, assassinating unfavourable politicians). When this attitude reaches a high enough level the aliens apply to join the Federation at the lowest level of membership. If accepted, their votes are allocated to the player to whom they have the best attitude. If their attitude to the Federation later declines (for example if the Federation allows an enemy race to also join) they may unilaterally leave.

Aliens are grouped around Earth in a way that tends to keep them having encounters with players. Specifically the number of pods relates to the current number of players and when they move they go to stars distributed randomly around the average distance from Earth of the players (ignoring outliers such as ones who went through a trans-galactic wormhole).

Sample alien stereotypes for further refinement (reducing Occidental bias and filling in all the specials would be good):

Locations

One of the least open-ended parts of TBG-1 is exploration, which was almost immediately exhausted in the early turns by the small and static nature of the board making it easily mappable. The much larger TBG-200X board reduces that problem a little, but the real fix is to make the board change over time so that it can't be permanently mapped. The "few aliens" model allows this by eliminating all the TBG-1 location types that relate to a settled universe (homeworld, colony, factory, prison etc.) Instead, the alien pods take the role of many location types and themselves move around.

This table includes locations added in the earlier TBG-2 version which was playtested for a while several years ago.

Religion

With officer action cards taking up the cinematic effects which were spells in TBG-1, the emphasis for TBG-2 religion moves to prophecy: it allows players to know the future. This may be an advantage, but should also sometimes allow players to feel trapped by their own destinies.

The alien pods include temples of the Great Old Ones, who may be individuals of the known non-human species. Players' officers can use these temples (with the help of the aliens in them in return for influence) to gain information about the future. Whether this comes literally from godlike but real creatures, or is a form of psionic power, or something else, is not clear.

As the Icelandic Sagas say, all men are lucky or doomed. But in TBG-2 terms this varies randomly from turn to turn. Most turns are neutral for luck, with 10% lucky, 10% unlucky, 1% blessed with great luck, and 1% doomed. The exact effects if luck are various, but the key things here are that a player's luck is known to the adjudication program in advance, and that a player can choose what risks to take much more efficiently when they know whether the turn is going to be lucky or doomed.

The officer action "Random Prophecy" gives them a secret which includes the luck state of a randomly chosen ship on a turn in the near future. It also gives them three mana, the unit of religious favour. Mana can be used to buy the officer action "Directed Prophecy", which lets the player choose which ship's luck and turn to forsee.

Mana is of four types, one for each Great Old One. Each officer can only use mana of the appropriate sort.

While a player has at least 25 mana they may declare themself a prophet of the appropriate Old One. If they end any turn with less than 25 mana they lose their prophet status. Being a prophet costs three mana per turn to maintain and gives the following benefits:

Old system:

Unlike TBG-1 players, TBG-2 players cannot pursue all religions at the same time. Favour is gained for actions which suit some deities and simultaneously lost with the deities who don't like those actions. Favour is relatively more valuble than in TBG-1 as it's harder to get, especially in large amounts.

Actions' religious significance is associated with the second generation deities, the four offspring of the original Great Old Ones who each combine the attributes of two of them. They are Mighty/Merciful, Mighty/Fierce, Wise/Merciful and Wise/Fierce. An action gains favour in the two attributes it represents and loses it in the other two.

Any player may declare themselves a prophet of any one of the Great Old Ones, or a follower of another prophet player (not both, and not a follower of themself). Prophets gain bonus favour related to the favour gains of their followers (a pure bonus, not subtracted from the followers' own favour) and pay a cost each turn to claim the status of prophet. The numbers are chosen so that it's only worth being a prophet if one has at least two or three followers.

(End of March 2007 update)

Fleets

Power-sorting and ship alliances are expanded from the TBG-1 form to allow multiple ships to operate as a fleet. Each ship can designate another as its leader rather than ally, and each ship can be the leader of any number of other ships. A leader and all its followers at the same star form a fleet. The followers of a common leader also form a fleet even when the leader is not present.

Players may choose a leader ship freely, except that the leader can reject a follower to prevent an enemy from hiding inside the fleet that wants to fight them. Alien ships contributed to the Federation Navy are assigned a leader by the player government (and they chase the leader each turn so fleets persist). Independent alien ships choose the strongest ship of their own race present.

Power sorting becomes a two-stage process, first sorting fleets into pairs by total power and then sorting the ships within each fleet by their individual ranking setting to meet a ship of the meeting fleet. The default ranking setting is a measure of melee combat strength, so freighters and fighter carriers tend to sort lower, but players may choose any settings. (For example, battleships can hide behind a skirmishing line of sacrificial light scouts, or a weak player ship may 'lead' a strong fleet of aliens from the back.)

Lower ranked ships of the larger fleet in excess of the number of ships in the smaller fleet do not meet anyone. In this sense the TBG-1 alliance system is extended to allow screening of multiple ships, but only when the enemy fleet is outnumbered. Each ship may only directly attack the one ship it is paired against, but all ships may deploy their fighters defensively to protect any other ships in their fleet (is this really a good idea?).

Stargate Railway Web

The most interesting period of financial history is the 19th century railway building era, allowing a layer of creative fraud and co-operation to be built on top of the economic activity of building a rail network. The analogy here is the Stargate Web, but the open-endedness requirement means the system differs from historical railways in being cyclic: the Web must grow and then collapse to allow new growth.

Co-operation is encouraged by making Stargates very expensive and providing support for joint stock companies to let players combine their capital to meet the costs. Growth is encouraged by rewarding the controlling player of each company for making it larger. Collapse is encouraged by making larger companies increasingly less efficient until the owners voluntarily liquidate them. There is thus a tension between the controlling player's goals and those of the minority shareholders.

The profit for a Stargate comes when a player chooses to use it to jump to another Stargate, which costs less than jumping in the normal way and provides certain success rather than needing navigation skill, beacon stars or luck to avoid a misjump. The tendency for collapse is created by reducing the profit when both source and destination Stargates are owned by the same company (it allows NPC corruption, say).

The profit for growth comes in allowing the company to issue new shares to its controller at a fixed par price to pay for new Stargates or taking over other companies. Since the company will often have retained profits the market value of shares will be higher than par and the controller can resell these new shares for profit, as long as the company remains reasonably profitable and promises to pay dividends at some point.

The scope for fraud is greatly increased by takeovers, particularly when controllers of both companies have minority holdings in the other and can manipulate both companies' cash reserves to create a net flow of wealth to themselves at the merger. Minority shareholders may tolerate this risk when they have too little available capital to control companies themselves if the alternative is stockpiling cash with no profit at all. Other motives include encouraging growth of the Stargate Web in the area where they wish to trade and explore, or wanting to take advantage of profits caused by their own future travel plans.

Most games involving share dealing and takeovers get stuck in liquidity problems when no player is willing to match other players' trades. This is avoided by requiring that the controller is also market maker, setting a price and spread each turn that others may buy or sell at.

(Start of March 2007 update, how it actually works)

A player can create a new company at any time by buying at least one of its shares for $100 and naming the company. These shares are created on demand and the cash goes into the company's capital fund. The player becomes chairman of the new company.

The company chairman can buy new shares from the company for $100 at any time, putting the cash into its capital fund. This is notionally to enable the company to expand, but may also be used by the chairman to cream off additional profit and tactically in takeover battles.

When the company has at least $1000 in its capital fund it can spend this to build a new star gate at the system where its chairman starts the turn. Each share issued must always be backed by either $100 in the capital fund or a tenth of a star gate.

Any player may use a pair of star gates for movement, paying a fee to the company owning each. This fee goes into the profit fund of each company. If both star gates are owned by the same company the user pays the same amount but the company only gets one fee, i.e. half the money paid is lost to NPC corruption.

If the company already owns at least one stargate and has cash in its capital fund, the chairman may sell shares back to it for $100 each from the capital fund.

At the end of each turn the player with the most shares in a company becomes the chairman. The chairman can set a dividend rate per share each turn, paid to the owner of each share at the end of the turn from the company's profit fund.

The chairman acts as the market maker for their company's shares, offering to buy and/or sell them on every turn, by choosing a single dealing price. All other players may choose separate prices to buy and/or sell any number of shares. Trading is resolved if possible by all buyers who offered more than the chairman's price buying shares from the chairman, and all sellers who asked less than the chairman's price selling to the chairman.

Dealing is simultaneous, so the chairman may use cash from shares sold to buy others in the same turn, and may sell shares bought in the same turn. If all buys above the dealing price and all sells below it cannot be satisfied (because the chairman has either too little cash or too few shares), then the price is automatically adjusted up or down to give the maximum number of share trades.

As well as building new stargates, the capital fund is used to take over other companies. Companies cannot normally own shares in other companies, so the takeover process is a single all-or-nothing event which either merges the companies completely or has no effect.

Taking over a target company costs $100 for each share it has in circulation (i.e. $1000 for each of its stargates plus the amount in its own capital fund), plus an amount from the attacking company's profit fund chosen by the chairman. The offer is made on one turn and the target company's shareholders vote to accept or reject it next turn, one vote per share held after dealings on the turn of the vote.

If a takeover is rejected there is no effect (other than the attacking company was unable to use the takeover cash for anything else on the turn of the vote). If accepted then all shares in the target company are bought for the offered amount, and the target company is eliminated with all its assets merged into the attacking company (star gates, capital fund and profit fund).

The chairman may liquidate a company at the end of a turn. This converts its star gates to cash at $1000 each and distributes all its assets to the shareholders (rounding down to the nearest $1 per share). The company has no further existence.

(End of March 2007 update)

User Interface

The player controls their position from a web browser, but extensive support for third party add-ons and the distributed nature of the game's core net services means this is not a single monolithic page. It is generated as needed by combining many pieces.

The order entry forms presented cannot be static because they vary with the orders themselves. For example, when an officer is assigned to an adventure the browser must display an order form for that adventure. It is much better for this to happen within the browser than to need a refresh from the server for each change.

These two requirements are met by organising the player's results/order page as a configurable set of frames. A top-level officer assignment frame contains javascript which rewrites other frames to offer appropriate options for the current assignment. Results frames can be associated with arbitrary net services allowing players to combine "raw" results from the adjudication system with third party tools for mapping, combat simulation, database searches etc.

Javascript Emulation

As far as possible any complex order entry frame should include the javascript equivalent of the related adjudication code, allowing modelling of orders' effects before submitting them. This shouldn't be carried to extremes for cases where large amounts of hypothetical orders from other players would need to be supplied to produce results. The only place this is worthwhile is likely to be the combat simulator/adjudication service.

Phased Implementation

Implementing the game as a series of increasingly complex sub-games simplifies the design process by allowing prototyping, and allows the distributed development team to practice on smaller subprojects while producing playable games early.

Phase 1 - simple multi-player game of trade and auctions

Each player starts with a ship of level 1 modules and a supply of X1 ore. There is no movement in the full sense, but players can choose on each turn whether to be at Earth (allowing trade and bidding) or at one of the ore collection locations (gaining some ore of the chosen type).

When at Earth, players trade ore using the full trade rules. They may also bid for a new module each turn, which is not part of the normal rules at all. The module is of random type and level and the bids are in terms of damage to a specific module type using the damage rules, though not including any form of officer skills.

The aim of the game, apart from testing internal ship economics, maintenance and trade, is left to the players, though acquiring the highest total or individual module factors are the obvious goals.

Phase 2 - add adventures

Replace the auctions with a partial implementation of adventures. Use damage costs instead of officer woundings and gain of modules as the big prizes instead of officer skill increases.

Original plan left in for reference, probably not feasible without using sash tools

Phase 1 - Ship modelling

This stage is a toy rather than a game as such, since each player's position is entirely independent: there are no deadlined turns or other interactions between players. The phase allows modelling of a ship's internal economy, both in local javascript for one turn, and via the adjudicating server over many turns. The major goals are to find a good model and to establish a template for basic game scripts.

The only script in phase 1 is: tbg2/ship/display

tbg2/ship/display is a published net service accessible by web or direct command ("telnet style"). It includes the logic to run one turn of a ship's internal economy, with components producing and consuming according to their environment (star type, raw material supply and on/off status). It handles each request from clients in one of two ways:

If the request contains no turn number it's treated as a new web request with no previous context. A default initial ship is generated and the appropriate web page generated.

If the request has a turn number and 'command=webpage' in it, it's assumed to be a form post in which the player has set some commands. These commands are used to initialise a ship and a suitable web page is generated.

Phase 1.5 - Eco-ship modelling with expandable infrastructure

Phase 1 took various shortcuts to simplify coding. Phase 1.5 adds the missing support to make the phase 1 script part of an expandable game, while not changing its basic functionality.

The new scripts are: tbg2/adjudicate, tbg2/ship/adjudicate, tbg2/ship/database.

The existing script tbg2/ship/display is extended to handle requests with a turn number and 'command=adjudicate'. A ship is initialised from the other commands in the request, a turn is run and the results sent to the client as text (several key=value\n pairs). This script also starts saving orders submitted to it in per-player files in tbg2/ship/orders/n where n represents player number and order format is the form data.

tbg2/ship/database is a published service which handles one of three types of request. It's a type of service that would be secured in at least two ways in the complete system.

The 'read n' command requests data on ship n. The script checks for the existence of a file on that ship and generates a starting position shipfile if there isn't one. In either case it returns the ship's details as key=value\n pairs.

The 'write n' command is followed by a ship defined by key=value\n pairs. It updates the file for that ship with the new data supplied.

The 'list' command is answered with a list of existing ship files, in text as one number per line.

tbg2/ship/adjudicate runs a turn for each ship. It does this by looping over a list of ship numbers from tbg2/ship/database, getting details on each ship from tbg2/ship/database, orders from the tbg2/ship/orders/n file and making a suitable adjudication request to tbg2/ship/display. It then submits updated ship details back to tbg2/ship/database with a write command.

tbg2/adjudicate is run at the turn deadline and calls tbg2/ship/adjudicate. Over time this script grows to call other adjudication scripts for each phase in the turn sequence.

Phase 2 - Trade

Phase 2 makes the game multi-player by allowing trade of resources between players.

New scripts are: tbg2/trade/display, tbg2/trade/adjudicate

tbg2/trade/display shows the buy and sell orders of all players and allows entry of new orders for one player (defined by the player=n form or ? parameter). New buy and sell orders submitted to it are stored in tbg2/trade/orders/n files for each player number n.

tbg2/trade/adjudicate is called from the main tbg2/adjudicate script at each deadline and resolves buy and sell orders from tbg2/trade/orders as fairly as possible, reading and writing ship data with the tbg2/ship/database service.

Another old combat system left in for reference

The TBG-1 combat system has some flaws which are addressed here. Very long battles are possible but usually accidental or contrived for abusive asset transfer: battles shouldn't be longer than a few rounds. Damage and looting motives are tangled into one system so, for example, it's difficult to do much damage to enemy modules which influence the battle itself because that would allow too much loot to be taken. The two goals should be more separate, and damage to current battle effectiveness should be uncoupled from permanent damage. There's a distinction between the usefulness of engineering (only impulse drives) and science and weaponry (two areas each). Warp factor should be useful in retreat and in blocking enemy retreat by subspace jamming. There should be a medical combat role so all officers and modules are significant in battle. Defensive ships should not need to maintain huge arsenals, it should be possible for occasional warriors to exert themselves more than regular attackers can risk.

Combat is resolved using battle orders from each player. These consist of target preferences, and the levels of effort to be put into various aspects of the battle by the officers. These effort levels amount to how much damage to apply to modules: unlike TBG-1 there are multi-turn combat decisions in whether to exhaust the ship in one battle or fight less hard but conserve resources to fight again next turn.
 

Combat Order	Officer	Module	Effect
Retreat	Engineering	Warp Drive	Battle lasts only one round
Prolong	Engineering	Warp Drive	Battle lasts two rounds
Close	Engineering	Impulse Drive	Set range to Short in second round
Evade	Engineering	Impulse Drive	Set range to Long in second round
Recon	Science	Sensor	Set range to Long in first round
Ambush	Science	Cloak	Set range to Short in first round
Battle Stations	Medical	Life Support	Brace life support to minimise overall casualties
M*A*S*H	Medical	Sick Bay	Deploy medical crew and equipment to minimise casualties in one module type
Shield	Weaponry	Shield	Use shields to defend
Beam	Weaponry	Beam	Use beams to attack
Fighter Strike	Weaponry	Fighter	Use fighters to attack
Fighter Interception	Weaponry	Fighter	Use fighters to defend
Marine Boarding	Weaponry	Marine	Use marines to attack
Marine Defence	Weaponry	Marine	Use marines to defend

Combat consists mainly of comparing the scores of the two ships in activities from the table above. The score is the effort (module damage) expended multiplied by the relevant skill of the officer making the action. The activities for each module type are a distinct skill, eg practice at evasion will also make an engineer better at closing, but not at retreating. Modules function in combat even without a specific officer action to operate them, in this case an otherwise-invisible NCO with minimum (and never improving) skill level of one is assumed.

Modules include integral crew, so damage to modules includes casualties. This gives medical resources a combat role in reducing damage to modules in the sense or preserving crew.

Each officer can take only one action in each combat round.

Combat orders (where effort means module damage to take plus an optional officer) are:
Recon or Ambush, and effort
First Round module type target
First Round Shield effort
First Round Intercept Fighter effort
First Round Strike Fighter effort
First Round Beam effort
First Round Boarding effort
First Round Marine Defence effort
Retreat or Prolong, and effort
Evade or Close, and effort
Second Round orders as for First Round, can all be different.

Combat Sequence

In each subphase below where weapons "engage" each other, this means they reduce the enemy weapon's score by half their own, to a minimum of zero. For example, 20 vs 12 reduces to 14 vs 2. Exceptionally, beams and shields are reduced only during the battle: they recover at the end. All other reductions are added to damage, up to the maximum of five turns damage to each module.
 

Phase	Details
Initial Range Phase (determine whether first round is at long or short range)	If both players order Recon, range is long, if both order Ambush it's short. If players give different orders, the higher scorer gets their choice.
First round (Details depend on range)	Long Range	Short Range
Fighter Subphase (fighters duel and winners target enemy ship)	Each player's strike fighters engage enemy interception fighters. Surviving strike fighters engage enemy shields.	All fighters engage regardless of whether they're on strike or intercept missions. Surviving fighters engage enemy shields.
Beam Subphase (beams are effective only at short range)	Beams do nothing.	Beams engage enemy shields.
Damage Subphase (attack one module type if enemy has no remaining shields, damage reduced by defending medical actions)	Surviving strike fighters engage enemy modules.	Surviving fighters and beams engage enemy modules.
Boarding Subphase (try to loot defenceless ship)	No boarding	If there are no surviving enemy fighters or shields, boarding marines engage enemy defending marines. If boarders score higher than defenders, they capture loot modules related to their margin of victory.
Retreat Phase (Jump out of combat or jam enemy warp field to prevent escape)	If both players order Retreat, the battle ends, if both order Prolong, there is a second round. If players give different orders, the higher scorer gets their choice.
Range Change Phase (set range for second round if any)	If both players order Evade, range for second round is long, if both order Close, it's short. If players give different orders, the higher scorer gets their choice.
Second round (surviving forces continue)	Subphases as for first round. Scores are re-used allowing for losses in the first round, target module types can be different.

Combat is fought in rounds at one of four ranges: Long, Medium, Short or Contact. Different weapons are usable at different ranges and interact differently with the defences. Each weapon and defence is usable only once, attack and defence mutually annihilate with the least general defences removed first (first fighter, ECM or screen, then deflector, then armour).

A ship can use attacks and defence in each round up to its production capacity, and in total up to five times its production capacity (except for armour, which can be used in any amount available). Each use consumes the weapon or defence plus an additional mission cost listed in the table below. If the ship can't pay the cost, it can't use the item.

Most types of defence are ship-wide, but armour is per-module and can't be moved to other modules after production. The general pattern of combat is therefore for attacks to overcome cheap/specific defences first, then deflectors, then the armour of each target module in turn. When a module with no remaining armour takes a hit it powers down with loss of its internal resources and targetting moves on to the next module.

Loot, a single module, can be taken only at Contact range and only by Marine weapons.

There are two overall types of battle: the Fly-by attack usually intended to damage the enemy, and the Intercept attack usually intended to take loot. The range pattern of rounds for a fly-by is one each of: Long, Medium, Short, Medium, Long. The range pattern for an Intercept is two each of: Long, Medium, Short, then rounds at Contact range until one side retreats or both sides run out of usable weapons.

Players choose their preferred range and this may cause early rounds to be skipped acording to sensor-cloaking interactions similar to TBG-1. In the fly-by case it also allows the battle to end earlier if a ship has the cloak advantage and range preference to disengage early.

The effect of impulse differences on range is not to change it directly as in TBG-1, but to either insert extra rounds or skip rounds, according to impulse strength and preferred range of each ship.

The effectiveness of ship components is reduced for the current battle by more than the amount actually damaged. The fraction of initial components remaining is cubed for current effectiveness, for example 90% becomes 72.9%, 80% becomes 51.2%, 70% becomes 34.3%, 10% becomes 0.1%.

Obsolete but good skill roll system, leave in for thinking about

Many actions needs a "skill roll", either to determine success/failure (e.g. hyperspace jump to another star), or to determine the level of success (e.g. mining comets for a variable amount of raw materials).

A success/failure skill roll has a difficulty level and notionally consists of rolling this many D2s (or tossing coins marked 1 and 2) needing all of them to show 1s. A level of success roll consists of rolling a series of D2s, ending with the first 2, and summing the 1s for a total showing level of success.

For a particular action, the number of rolls an officer can make is their skill level in the appropriate field. The best of these rolls is the one that counts. For example, an officer with skill 10 attempts a success/failure task of difficulty 3: they have up to 10 chances to roll 3 D2s, needing at least one set of 3 to be entirely 1s for success. For another example, the same officer rolling for level of success would start rolling a series 10 times, continuing each series until rolling a 2: the longest series of 1s is their score.