I already posted a similar remark in a previous beta stage, but here it goes again:

Why on earth (or any planet ) would I custom build a ship with defense systems ? Look at this:

I've advanced in reasearch equivalently both in offense and defense. Still, I can easily pull up a ship featuring a +19 offense (in one area) when I could hardly pull a +8 defense : offense gets way too much compared to defense.

There are at least three other major reasons to reverse this trend.

But first I suppose:
- That the combat system is still the same as before (i.e. attacker rolls a random die from 0 to attack, defender rolls a random die from 0 to defense ; if attacker result is greater than defender then attacker damages defender) (BTW, where did I read this ???)
- That the old way is not carried out (i.e. when you have more defense than offense then half your defense is added to your offense) ; I don't see how this could be carried out with the new system anyway.

Now the arguments:

- there are three weapon branches and defense being reactive (i.e. you don't know what you're going to face) should be much more effective than offense. i.e: If I can make a ship 2 attack / 0 defense or 1 attack / 1 defense (choice of one), I get 2 chances out of 3 that my defense will be wasted just because I'll meet an opponent with an offense that doesn't match my defense. Say I build a (2,0,0)/(0,0,0) ship : meeting any other similar ship is a 50/50 proposition. Meeting a (1,0,0)/(0,0,1) is cheesy (that ship defense is worthless in that situation). Meeting a (1,0,0)/(1,0,0) ship is probably still a win (large offense is better that large defense). So no use placing any defense at all.

- Even a large defense value is not going to win the fight ; for this, you require a good offense value. Defense will only buy you more combat rounds. Hence there can be no harm in making defense more (much) effective than offense.

- Also note that the effects of going from (for exemple) 10 to 15 in offense are greater than those of going from 10 to 15 in defense: the attacker will still be dealing on the average more damage to the defender in a round ; I think (this would have to be backed by some computations) that 1 point of attack is worth 2 points of defense in that regard.


So, to make things about equal between offense and defense, the following conditions should apply:
- you can put 6 points of defense where you can put 1 point of attack (space wise and maybe cost wise).
- you should have a tech level for a 2n defense when you have a tech level for a n attack, considering you have spent the same amount of research in both fields.
- you should be able to research one level of defense in at least two fields (probably all three fields actually) in the time you take to research one level of one kind of attack ; otherwise your ships will still end up missing defense wise.

Then, with a level of say n=2, you could design the following ships:
(A) (2,0,0)/(0,0,0)
( (1,0,0)/(2,2,2)
(C) (1,0,0)/(4,2,0)
(D) (0,0,2)/(0,0,0)
(E) (0,0,1)/(2,2,2)
and now A and B and D and E match each other, C is better than A and B but only because it is specially designed to counter their kind of weapons, but would clearly lose to D or E. On the overall, all these ships are on par.

All these ships would take about the same amount of research, use the same space and cost the same.

In addition, I believe it would be good if a ship with two weapon system of value n was not as good as a ship with one weapon system of value 2n. For exemple, the first ship should probably not do an attack with a rating of 2n but two attacks with a rating of n to reflect that it is technologically inferior.
Still, this poses some problems to represent the ship's attack values.

Even if that scheme is still flawed, it corrects many problems, mainly that the only viable route today is:
- ignore all defense
- concentrate your research on one weapon field
- build ships (n,0,0)/(0,0,0)
You'll still end up much better than any opponent having maximized his defense against your weapon of choice.

I could run a spreadsheet simulation to support my suggestions.

Yves