Attacker Vs. Defender (Who has advantage?)

I won't go into the numbers, there are plenty of posts explaining the way attacks work. There is still a random factor, and whether your opponent has +luck or not. But what I will say, is that I have found reloading over and over and doing the exact same attack sequence will result in the exact same results. Try it once. Once you moved 2 spaces, you changed the sequence, and new numbers were generated for the battle. Normally, accept for Arcean, there is no difference between attacking or being attacked.

I do understand there is a random factor. Also if what you say is true, the numbers are exactly the same when reloading and performing the same sequence, then that is a bad implementation of a random generator (or was that intended?)

Does luck really tip the balance of a sure combat victory? I've read the wiki's and I'm not convinced that it does. I can see it playing a large roll when the statistics are within 30%. Here's another battle that baffles me.

Me (Torian): My luck is 0%
Fleet of 2
1/9/1 x2
5/1/1 x2
HP:48 x2
vs.
Thalan
19/0/0 x1
24/0/0 x1
HP:31 x1

I have initiated combat with the above setup (several times) and every single time the opponent wins. Of course I'm not going to go through the math, but I'm sure it seems obvious who should win the battle. Each time the Thalan ship survived with 17-20HP. I've even run this with both simulators linked on the wiki, and they both say I have over a 98% chance of winning. How is it that luck could tip this combat so many different times? If it really is due to luck, then there should be some way to see the AI's luck value (using spies and such).

If anyone could possibly shed some light on what I'm missing, I'd be very grateful.

First question is, did you replay the movies of the battles and see the sequence of shots by each ship, damage done and so forth? I would think that would have given you the answers to your questions.

In your first example, seems you should have won, but were unlucky with the roll. It is possible that the random number generator makes the same roll when you restart and replay, so not as perfectly random as one would like...that's the only explanation I can think of for your result. Is there any chance that the ships of one side could have been within the area of influence of a nearby military starbase? Or that even moving a couple of squares could have moved the battle in or out of an AOI?

I must say, I've never seen a battle in which both sides lost all ships and the battlefield was cleared. Not ever.

In your second example, if I'm reading the data correctly, Krynn has in effect a 24-4-4 defense after stacking. (4 being the square root of 24 truncated to lower whole number.) So you are shooting 1 vs 24, then 9 vs 4, then 1 vs 4, all of this twice...the 1's have little chance, and the 9 to 4 rolls averaging out at 4 or 5 agains 2 or 3 so not much damage there, either. Meanwhile Krynn is shooting 19 vs 7 (5 plus two times the square root of 1) which would average at about 9 or 10 vs 3 or 4. Krynn has the edge.

There is a nice paper in the Wiki about the math of ship weapons and defense and one of the conclusions is that it is usually best to focus your offense on one weapon instead of gaining the techs for multiple weapons, so 0-11-0 instead of the 1-9-1 ship that you described. The reverse is more or less true for defense, where it is desirable to have perfect square numbers and a spread defense if the AI's are not all using the same weapons. This is especially true in the early game, where a 1-1-1 defense gives you the effect of 3-3-3, a real bargain.

BTW, there is some evidence that defenses wear down under repeated attacks, which is something to think about.

Hi!

If you'd go through the math you'd see it's not so obvious who should win.

In DA you don't attack with 1 and 9 and 1, but with each weapon separate. In your case and assuming harpoons (attack 3):
- 1 beam,
- 3 attacks of 3 missile and
- 1 mass.

When combat round starts, Thalans are defending to each particular attack:
- 1 beam vs. 24 beam defenses, you hit and 24 is decreased by 1.
- 3 missile vs. sqrt(23)=4. Let's say you hit with 2, he rolls 3. 24 is decreased by damage done, so it is 21 now.
- 3 missile vs. sqrt(21)=4. You hit with 1, he rolls 2. 21 is decreased by damage done, so it is 20 now.
- 3 missile vs. sqrt(20)=4. You hit with 3, he rolls 1. 20 is decreased by damage done, so it is 19 now, his HP is decreased by 2 to 22 now.
- 1 mass vs. sqrt(19)=4 beam defenses, you miss.

Now your next ship fires at 19 defenses by the same procedure.

Next he fires back, but you have very weak defenses vs. his beam attack (assuming he has 6 phasors with attack 3 each):
- 3 beam vs. your 5 defenses. He rolls 2, you roll 3, so your defenses are now 3.
- 3 beam vs. your 3 defenses. He rolls 3, you roll 2. Your defenses are now 1, you lose 1 HP on the first ship, 23 now.
- 3 beam vs. your 1 defenses. He rolls 1, you roll 1, so your defenses are now 0.
- 3 beam vs. your 0 defenses. He rolls 1, you lose 1 HP on the first ship, 20 now.
- 3 beam vs. your 0 defenses. He rolls 3, you lose 3 HP on the first ship, 18 now.
- 3 beam vs. your 0 defenses. He rolls 0, you lose 0 HP on the first ship, 18 now.
- there's one missing pont of his attack (6 phasors + weapons bonus for 19 beam attack), that isn't covered here. Supposedly it is added to the first shot of first weapon into so called supershot. I actually don't know.

Combat round over. Defenses replenish, next round starts, until all ships on one side are destroyed, or 50 rounds are over.

If you look just to the upper first round, you see you're losing hit points MUCH faster than he does. And when one of your ships is gone, your attack is practically null vs. his huge off-type defenses! Also, I'm not sure if defenses deplete for each weapon type attack, like I described. From DA (Beta) Combat System it's not exactly clear, but it seems logically.

BR, Iztok

Could someone from Stardock comment on this quote?

Thanks again for your excellent analysis Iztok. I am wondering Iztok whether you are happy now with the DA combat model or whether you actually prefer the DL model?

Hi!

I already had quite some discussions about DA combat with Wyndstar. Despite there IS a Rock-Paper-Scissors system in it, the AI unfortunately can't use it properly. At the extreme edges of DA combat model AI most of the time fails, and uses only the Rock part of it.

I personally don't have a strong preference to either combat model, but AI was designed to use DL. From the gameplay experience I'd rather see the DL combat model, with some small adjustment against end-game kills through large defenses.

P.S. Well, I haven't had much combat experience recently. From reply #2 it is clearly visible AI used highly defended ships. Maybe now it can use more than only Rock?

BR, Iztok

Of course it would be silly to put off-type lower weapons when I could focus on a weapon the enemy has little defense against. The 1's were due to a military Starbase I had (and was in range). But if all the examples above are true, then the simulators must be way out of date (or meant for a different version of GcivII?).



I completely looked over the calculations for wrong defense in DA. When reading the wiki it said that Dmax was calculated using the sqrt(defense) and that the dmg done is subtracted from Dmax (atleast that was my interpretation. But the DA changed it to subtracted from the original value in which a new sqrt() is calculated for the next weapon? In the case of "off type" defenses, when the weaponDmg < sqrt(maxDefenseValue). The defender is at a sure advantage. Basically I needed a weaponDmg >= 5 to even have a chance to win.

What's interesting is in the case of 10 weapons (3dmg each) vs 25 defense (off type) the result would likely the be same as if the defender had 25 (on type) defense.

e.g.
off-type
3dmg vs sqrt(25) where attacker averages 1dmg and defender will average 2
3dmg vs sqrt(24) ~ 1dmg vs 1.5 defense (on average) this will continue until finished
Result is no dmg done on average (with the ending defense being sqrt(15))

correct-type
3dmg vs 25 where attacker averages 1dmg and defender will average 12.5
3dmg vs 24 ~ 1dmg vs 12 defense (on average) (each round .5 is subtracted from avg)
Result is no dmg done on average (with the ending defense being 15)

If this is correct, then wow! Time to invest in some major defenses.


Thank you both for the clarification. The Thalans ending HP does suggest exactly what was stated by Iztok.

As a side note. Perhaps some mention should be made on the wiki about the simulators?

I'm also curious how the order of attacking and defending ships is chosen. If this could somehow be influenced, that could seriously turn the tide of a battle. I've looked all over the wiki and even these forums with no luck. From what I have read, the order of the ships in the fleet list window is the order they will attack/defend? This does seem to be true from observation.

Hi!

Unfortunatelly that's true. All curently available combat sims are for DL battles, not DA.


Welcome in the club! Please not this only works until you can mount enough defenses on a single ship to block ~75% of attack power of opponents fleet. This timeframe lasts until AIs get (close) to the end of weapons tree. You really should make use of that advantage until it lasts (read: conquer as much AIs as you can), because when it runs out, you need to become "picky" what to fight with what.

BR, Iztok

That is correct, the formula is total attack / (total defense + hit points) . The higher the result of the formula, the higher the priority for enemies to attack such a ship. But as you can see, the higher the attack rating (compared to defense), the higher the priority, so it's difficult to create a fleet in which you have some uber defense ship that takes all the hits and a couple of attack ships that do the damage.

Hi!

Err, not exactly. In your case you still roll 3 vs 4 in most shots, so there's a reasonable chance you'll score some hits. But rolling 3 vs. 24, there's only a chance you'd score a hit if he rolls 3 or less. That's only 1/8 shots to have A CHANCE to score a hit. Defenses in DA really rock!   

BR, Iztok

Hello there...
My head spins. Interesting stuff. But this is my experience:

I never use defenses of any kind; I prefer to use the space for weapons and engines. I usually just research one type of weapon (missiles). I tend to play on medium (sometimes large), at crippling.

What usually seems to happen to me is that if I attack first, and my total fleet strike power is as great as theirs (or bigger), I win. If they attack first, and their strike power is as great as mine (or greater), they win.

First strike always seems to help (I usually play Yor or Korath); sufficeintly that sometimnes I win with a smaller strike power than them - if I strike first.

Is this just some flaw in my brain making me perceive combat this way? I don't look deeply into the numbers, but combat always seems to follow these basic rules for me.

I guess I'll have to try adding some defenses & see what my mileage is!

Cheers

This isn't quite the case in DA, as stated above, due to "First Strike" requiring the "Super Warrior" ability (Acreans/Custom Race). Only with this ability does the "First Strike" rule apply, otherwise combat is attack/defense is computed simultaneously.



This is good to know. Now I see why my HUGE ships are always the last targets Ah well at least I can attempt to calculate a good fleet combo now. Btw, where did you come across this equation?

Indeed my assumption was of course based on purely the avg value being used without possibility of randomness. The first case does have a chance, just quite slim, whereas the second case the attacker is toast.

All this makes my initial tech development strategy quite a bit different.



Has nobody else seen a battle that both ships lost? I will say that I've seen some very mighty strange things happen. Once I saw a Thalan ship attack a mining starbase without any modules which it killed, but his ship disappeared as well (without blowing up).

In either case, I will say that I've only seen ships lost on both sides when all combat sequence battles are turned off.

I can't say that I have ever seen 2 ships completely destroy each other, sounds like a bug. Though mutual distruction is something that some players would like to see in the combat system, I don't think it is intended to happen. While defense is king early to mid game, what Iztok says is true. Once the AI can put enough attack on a ship to blast through all your defense, the strategy switches from all-defense to all-attack. And it is here where the lack of mutual distruction seems a bit boggling.

Probably on the wiki or some kind of post on this forum.

I am wondering whether you guys create big fleets with some big ships (capital ships) and some escorts (frigates/destroyers) or just same type of ship fleets? I think it would be more 'realistic' to have the former.

Hi!

... but it shouldn't, at least in DA. The only advantage for attacker in DA I know is: in case the 50 combat rounds limit is reached and both ships are totaly equal, attacker wins.

BR, Iztok

Hi!

Realistic yes, but game doesn't work this way. When firepower of fleets became large, game punishes usege of fleets. So I always try to use the biggest hull possible, using it as a singleton most-defense ship. I use fleets of most-defense ships only to kill opponent's all-attack fleets faster, to diminish attrition to my ships.

Smaller hulls (in my case this would be medium hull in late mid-game ) just die in battles. I rather use them for patrolling, and killing un- or low-armed ships. I can afford maintaning them, because I usually build small amount of quality ships, not quantity of them.

BR, Iztok

Thanks Iztok, shame that the game was designed this way. I hope they fix in the next expansion. It would be nice if mixed fleets were the best way to go.

Mixed fleets may never be the way to go. Attrition is too high for a mixed fleet, and even with the new HP modules I haven't found a super good way to keep smaller ships alive. The best argument for mixed fleets in TA is that because they increased the upkeep for all ships so much, it is not financially feasible to have too many capital ships roaming around.

In practice, it just means that in TA I baby a very small fleet of capital ships, and they are charged with doing all of my heavy lifting. Transport interceptor type ships have more reason to be tiny now instead of medium, just to reduce upkeep.

~ Wyndstar

While the math adds up, the random factor does seem to play more of a role when the total HP of the two ships/fleets are the same. I have seen this happen more often than not especially in late game combat.
Also very mixed fleets of one major weapon type per ship and mixed defense seem to win more often. IE 3 ship fleet 1 ship all gun, 1 ship all missile, 1 ship all beam with each ship having defense of 2/2/2.
Also side note when combat screen is not used their does seem to be a few times when both the attacker and the defender are destroyed.

Hi!

"Invisible ships" bug strikes again? When you'll see that event next time, please click the space both ships/fleets occupied before combat. If you'll find ships there, obviously you found a bug.

BR, Iztok