Defense Solutions

OK, there has been a lively discussion as to the worth, value and cost of defense under the new system since DA was released 3 weeks ago. See: WWW Link and WWW Link and WWW Link.

First off, let me say (and I'm sure I'll reiterate again) that I am not really concerned with how using defense on your ships effects players - and I will try to explain this further when I discuss the problems. This game is not multiplayer (I am not advocating multiplayer), so I will not play against any of you ever. I play my games against AI opponents. How they use and adapt using defense is all that really matters to me. I believe, and will try to explain why, under the current system the AI will never be able to use defense effectively. This mandates a change, in my opinion, because defense really should matter.

Let me take you back for a moment to the days of the Dread Lords combat system. Over close to a hundred games my experience was something like this:

There was an important balance between the three weapon types and the three defense types. Defense against the proper weapons type was very effective. You would still take (some) damage, but you could seriously decrease the amount of damage you took in a battle by having ships equipped with the right defense. Larger hulls got better as the game went on, because a larger hull could pack more weapons and armor, enough armor to ignore all of the smaller hull ships even (because defense could be so effective). Defense was expensive, and if your ships were fighting one of the two weapon types they were not equipped for, you would take a lot of damage and it would cost your empire plenty.

You often wanted to build ships with two weapon types, so that if the enemy used one kind of armor, you could still do enough damage to win with your other weapons. If you were using only one weapon type and were attacked by someone who had armor against your weapons... the game could become very tense. You would put everything into technology to try and quickly rush up a different weapons tree while your already built (and largely ineffective because of the enemy armor) force tried to hold off the enemy long enough for your new ships to roll off the assembly line and start doing damage. I will never forget some of my clashes with Drath Dreadnaughts (Drath used to have a 50% defense boost) - that I was never able to damage in time. I lost games. Early on, I lost a lot of games when I didn't pay attention to what defense my enemy was using.

Part of preparing for war including checking enemy fleets to see both what kind of weapons and defense they were using, and building a fleet to counter. Even with minaturization maxed and huge hulls, you just couldn't build a ship that excelled in all three kinds of weapons and all three kinds of armor simultaneously. Ships were designed for specific situations and against specific enemies. Amazingly - good job stardock - the AI was very good at adapting to your specific rock/paper/scissors strategy. Several weeks or months into a war, the AI would predictably - and mercilessly, start kicking out ships that best matched the fleet you used against it. It was a good system that always kept you on your toes, and kept combat interesting.

I liked using defense, I liked gaining experience. Fleets of medium fighters could take on 5-10 times their number of similar medium fighters if you had armor against the enemy and they didn't have armor against you. Defense made that much difference. There were forum debates about whether the cost was justified for defense, but you always took the time to find out what weapons and defense your enemies were using, even if you stuck to all attack ships yourself.

Changes were made to the system, that for one made tiny ships relevant again in DA. Defense now goes away with each point that it blocks until the end of a round. Every ship in a fleet will concentrate its firepower on what it feels is the largest threat until it is destroyed, and then move on to the next largest threat and so on. If a ship with defense survives the attacks of the enemy fleet, the defense is restored to full power at the beginning of the next round. Battles are limited to 50 rounds. There were other important changes as well - extra firepower from one ship would be applied to the next and so on, but those changes don't effect the current problems as drastically.

This new system brought in the interesting change where suddenly large packs of tiny ships could take down capital ships that formerly they could never scratch the paint on. Depth of choice and strategy between hull size and logistics value was added. However, defense became less valuable, because rather than trying to stop the weapons from each ship in a fleet, defense was now tasked with stopping the weapons from an entire fleet simultaneously. At the same time defense got a slight boost in dependability, the "rolling a 0" problem of the old system was gone. 10 points of defense will stop 10 points of attack (really 20 points, law of averages, you can read about this elsewhere). The changes, however, have led to a net negative to defensive value.

There are really two fundamentaly seperate (although related) problems to how defense currently works in DA.

Problem #1: Defense is not effective enough against matching weapon types.

And this is a BIG problem. Weapons ultimately outpace defense, and under the current system they even easily outpace matching defenses. Because defenses now have to face an entire fleet's worth of attack, they are swept aside without a thought.

Because defense is so effortlessly overcome (be it with better tech, miniturization, logistics, hull size, etc.), even matching defense, it ceases to matter at all. I no longer need to check what defenses my opponents are using, I know my fleets will wipe them out in the first round. Because I know I can do this, I know my opponenet can do this, so I don't (typically) waste resources on defense myself.

In law one thing we like to do is use what is called the "but for" test to establish causation as an element in a cause of actoin. In this situation, it works like this "but for the AIs use of proper defense, I would have won." Until that statement is true, defense doesn't matter. Put another way, if an AI can't beat you by using DEFENSE, then you don't need to pay attention to it as a strategy at all. Remember, under the old system the AI could handily beat you with defense, which added all of that rock/paper/scissors strategy I was reminiscing about.

If defense doesn't matter, then weapon types don't matter. If weapon types don't matter, you have lost a lot of depth. Now there is no point in taking the time to research two weapon trees, you don't need to worry about shooting around defense. In fact, researching two weapon types puts you at a DISADVANTAGE (because for instance 4 pts of defense will become six points, 4 pts against your primary weapon and 2 pts against your secondary weapon). But offensive focus beats everything always under the current system (exasperated by the weapon power carrying over to the next ship - a rule change I really love). You will do more (if only marginally) damage if your opponent uses ANY defense by just sticking to one type of gun. And matching armor doesn't slow you down.

This means that you can just pick whether you want to destroy every empire in the game with doom rays or black hole eruptors or whatever when you start, and no one will ever knock you off your strategy. There is no point in finding out what the enemy uses, you know you can beat it. AND you know you are cost effective (see below).

Perhaps offense should have been made more powerful. But if even at the hardest difficulty level the AI can't BEAT you with (matching) defense, then ultimately defense and weapon types are meaningless. On suicidal I SHOULD LOSE GAMES for not paying attention to what my enemies are fielding in terms of weapons and defense - while just focusing on one weapon type myself.

For the last three weeks I have been kicking out monster beam ships and nothing can stop me. I can give the enemy the best defenses against my weapons, it still can't stop me. It is not because of my brillance or skill, it is because the game design is currently set up so that NOTHING can stop this strategy. If both sides use it, it just comes down to who can produce more industrial units. This is incredibly boring - and again, is much more shallow combat gameplay than what existed just a month ago.

--end problem 1--

Problem #2: Defense costs too much with its current value.

Yes, there are instances where defense can be worth the current cost, but it is rare. The AI will try to build ships that strike a reasonable balance between offense and defense. Small defensive values are just a waste of money, because they are quickly swept aside and the ship does not survive to another round.

Concerning the AI, it will dutifully start to churn out more expensive ships that will kill much less bc per bc than your offensive ships. This is even IF it is using the right defense against you. Trading a high level defensive tech can slow an AIs ship production to a crawl, and the few ships it does turn out every week are easily swept aside. Even when the Krynn build huge hulls loaded with defense, you just fleet up and blow them away. Thousands of bc in production are swept away with (typically) only a few hundred bc in production lost - if you even take any casualties (thanks again 1 hp rule).

--end problem 2--

OK, you may or may not agree with me that these two are problems. Please, again, keep in mind that all of this is in the context of what the AI does, and how it handles a game. Perhaps Problem #1 or #2 are not problems for YOU when you are out-thinking the AI. Good job, have a cookie! What I want to explore is what sort of solutions we as a player community can come up with to help the AI with these problems.

Two things I want to get rid of right off the bat as "solutions". The first is just modding the game. I like playing on the metaverse, and I like playing with the default game. This is a system wide problem that effects the game as a whole. A mod fix is only a fix for a very few number of players. Mods are a great solution to some minor problems or if you really want to heavily customize the game. They are not a satisfying solution to basic game design problems.

The second "solution" I want to discard right away is to just tweak the AI ship building algorithims. This takes two forms, some people suggesting that if defense is useless, just have the AI always build attack ships. Alternatively, people think the solution is to just have the AI become much, much more clever in its ship design to be able to "figure out" combinations of defense that will actually be effective. First off, taking away defense choices from the AI doesn't fix Problem #1, defense is still useless, its just that the AI is no longer hampered by it. Second, as has been commented on before, there is a VERY fine line currently between defense being useless and making you invincibile. Trying to code an algorithim that can predictably find that line with all the variables in a galaxy at any given game state is such a mammoth task as to approach impossible, and at the very least would not be an exercise in good time management by the development team. This is not a problem to just chalk up to "the AI is just dumb" and move on.

I want solutions that work, but that are also simple to be implemented by the development team. So with that loooong preamble aside, here are my two proposed solutions for fixing defense:

Solution #1:Cut all defense module values to 1/3 of their current cost.

At half of current price matching defense is about equal pound for pound with pure offense, but remember, we want defense to slightly exceed offense when you use the right kind so that if you go against matching defense you will lose. If defense is simply of the same value as weapons, you can just look at the industry graph and figure out (mostly) who will win a war no matter the mix of weapons and defense. You are still not motivated to start paying attention to defense if it is simply of equal value.

Why this works: Under the new combat system, when you buy defense, you are currently buying hit points. These are hitpoints which can (in a few situations) regenerate every round of combat. With very high offense values, defense rarely gets a chance to regenerate. Because of this, we need to look at the cost of the hps we are buying.
Defense is currently not only more expensive on a hp for hp basis than what you buy with a core hull, it is more expensive than the core hull PLUS WEAPONS. (wow)

Offense values rise to destroy ships with greater logistics, better miniturization, getting farther down the weapons tech tree, or with larger hulls. That is a lot of ways to overcome that measly little defense, and most players use combinations of these factors to send their fleet attack values very high (3k or more per fleet at the end of the tech tree). As long as we are dealing with large fleet attack values, we are not going to have much of a chance for our defense to regenerate.

Defensive ships already have lower offensive value. The trick here, to get defense to outpace attack if it matches, is to make defense cost less pound for pound for weapons. If both fleets are going to mutually destroy each other in the first round of combat, make the pure offensive ships more expensive (they can destroy more, after all). If offense is more expensive, then it will pay to invest in defense, even if it is just to "fill space".

Effects of this solution: Well, I think this can have a very good effect. If defensive ships are cheaper, then the AI is not at a disadvantage when it builds defense. Trading a defense ship for a pure offensive ship becomes a net positive for the defending empire. Remember, under the current system offense can ultimately outpace any defense, so the question you want the player to weigh is "at what cost?" I CAN destroy your heavily armored dreadnaught with guns, but is that efficient? Better to try and take you down with missles - and maybe use some defense myself!

Remember, these cheap hitpoints you are buying quickly lose their effectiveness against two weapon types. But we need to outpace matching weapon value, so they really do need to be very cheap.

-- end solution 1--

Solution #2: Quadruple all hit point values for ships and starbases, except for cargo hulls. Change the anamolies that add +5 hit points to +20 hit points.

I thought long and hard about what specific value to suggest for an across the board increase in hit points. I often end up with games where ships pack roughly 4 times their hp in weapon damage. For instance, in another thread I posted a screenshot of ships whose base stats were 574 attack/124 hp. This ratio can get a lot higher for tiny hulls. Of course, this will change slightly depending on racial stats, galactic wonders, miniturization level and number of military starbases. If you pack four times your hp in damage, you will on average do twice your hp in damage per round. The high weapon damage vs. hp values also leads to offense ships being very able to get xp, because it becomes very simple to make ships/fleets that will wipe out the other side in the first round of combat... so your ship lives.

Why this works: Defense does have an advantage in that it recharges at the end of a round. If a ship takes damage but lives, it gets a bunch of "free hp" back from the defense. Under this system, defense doesn't need to be repriced, the triple cost should be justified. If you x4 hitpoints then on average, against a matching defensive fleet you will kill less than twice your number in enemy ships per round (all sizes being about equal). If you kill say 1.5 ships per round vs. two, that makes a big difference because ship #2s defense will recharge, rather than being lost in the first round. Small changes in losses in the first round have cumulative effects in later rounds. I didn't do a full analysis at all sizes and all weapon values - but from the graphs I did do this appears to be the point (at x4 hp) where (matching) defense again overtakes offense.

You will still take damage as the defender under this system, so attrition matters. Also, you CAN be overcome if the enemy is willing to spend enough, or - lets not forget - is using weapons that your defense isn't matched to. But this does let us get to the sweet spot where defense can matter again.

Effects of this solution: Well, more battles will get to that 50 round limit. Watching full battles will take a lot longer. There will be more ties - or that is to say battles where both fleets have at least SOME surviving ships. This might be cool though, imagine two well armored capital ships slugging it out against each other, and after a full battle both survive (but with damage). Do you want to use your next move to battle again or to try and limp away and repair?

Planetary defense will become SLIGHTLY more effective because the increased repair % while orbiting a planet will have a larger impact when there are more hps on a ship.
Also, all ships might be more surviveable. It might often take more than one "Move" to destroy a fleet, although not necessarily more than one turn. This might lead to clutter and higher ship maintenance costs - although I think super dominator corvettes have largely the same effect and the game already manages those.

--end solution 2--

Whew, you finally made it to the end of this post (almost). What a trooper. You must care about this as much as I do, or be really bored

I am very interested if anyone else has other solutions, what they are, and how you think they will effect overall game balance and why. I'm also interested in feedback on either of my solutions. Or, alternatively, maybe you think my assumptions about defense vs. the AI is wrong and you want to prove that the AI can beat you with defense under the current system. Ultimately, I'm interested in a discussion about defense and fixes (if needed at all, feel free to try and convince me otherwise) that focus on what the AI does in a game, not on how a player can be clever with defense to beat the AI. I love this game, and want this thread to be something constructive.

Thanks for listening!