Request for (fleet?) feature

There's one thing about GC1 that has always annoyed me. I'd stick two or more defensive ships on a frontier planet for defense during a war, and they'd beat back the masses. All ships would take some damage, but with the damage spread between all of them, they could keep up with the damage doing repairs.

Until one of them leveled. All of a sudden, because that ship has a higher defensive value, it would catch all the attacks, eventually being beaten down and destroyed, at which point things go back to normal until the next ship levels, and so forth and so on.

At first, I was thinking of a fleet function, something to the effect of "designated defender" and a matching "designated attacker," though it occurs to me that that would lead to too much micromanagement for this use. While those might still be useful functions, I think it might be better just to change the way a default attacker/defender is chosen. Ideally, it would be "pick whichever ship has the best chances of survival" since someone has posted a formula for this that works for CG1, though I can see where there might be problems where attacker chooses ship A, defender chooses ship B, attacker switches to ship C because it's better against ship B, and defender switches to ship D because it's better against ship C, and then the attacker switches back to ship A (decisions made by the combat algorithm, not the players, that would be FAR too much micromanagement). Maybe a simple rule (use highest of total offense/defense as appropriate * current hit points), or "if, given best ship A, if ship B has at least 90% as much defense and at least 20% more HP as ship A, use it instead, though I can see the "best" solution working if at any time you've come back to a pair of ships you already suggested, that's the pair you go with (favors the defender, but I'm willing to accept that tradeoff, being a defensive player

)

Locked Post

Advanced Search

Watch this post

Do not email me updates for this post Email me updates for this post

Successfully updated karma reason!

There is a fleet mechanic coming in, today's dev journal br CariElf talks a bit about a cool sounding screen which will display battles. There isn't much information yet about how fleets are going to work though. I can't tell you anything at all when it comes right down to it.

Successfully updated karma reason!

though I can see where there might be problems where attacker chooses ship A, defender chooses ship B, attacker switches to ship C because it's better against ship B, and defender switches to ship D because it's better against ship C, and then the attacker switches back to ship A ...

Rock-Paper-Scissors has no best pure strategy. The highest score that can be had is achieved by playing a mixed strategy of 1/3rd chance of playing any of the three moves each turn by both players. If playing multiple rounds, it is also prudent to take advantage of any perceptible pattern in enemy choice history, while making your own as random as possible. Another way of looking at this is to say we have friendly ship list { A, B, C, ... } and enemy ship list {a, b, c, ... }. They can also be thought of as weapons or weapon groups. We can draw up a cross-reference table like so:

-- A B C ...
a
b [outcome score]
c
...

Then extract only the set of outcomes with the highest value to you, and choose randomly between them. Or assign probability of choosing a given attack according to it's outcome score.

The main task is then choosing a scoring system. If we define "winning" as the process of reducing enemy threat of damaging our weapons to zero before the enemy does that to us, then we can develop an appropriate score system. And the scores could be weighted to take into account any attack/defense modifiers like lead ship/escourts protecting their charge, heroic units, ship size, racial bonus, etc.

Then each firing produces a change in the scoring table, which results in a new table for the next round. The objective is then to search for a sequence of table updates (chess positions) that result in a winning outcome. As an example of this kind of look-ahead paying off, it might be possible to take out a single enemy gun this round by shooting at that small ship, but if firepower was directed instead at a sleightly larger ship then it would take out three enemy guns next round.

Successfully updated karma reason!