Attack calculations

Recent

Hottest

Forum Index » General Discussion
1 2 3 ► Go Go to page...

Author

Message

Jun 16, 2009 15:35

Subject: Attack calculations

draka

Messages: 38,
Joined: Jun 16, 2009,
Offline

Can anyone tell me exactly how the attack calculations are made I can normally predict to a 90% accuracy but today I got one that makes no sense to me

My Titan big tank was on 7 after one attack then got hit by a kharilean light on a mountain and got -4, I have attacked with a sapien buggy with 10 against a Titan big tank with 1 as a second attacker and got
Nothing

To me this makes no sense can anyone tell me exactly how these calculations are made ?

Thx draka

This message was edited 1 time. Last update was at Feb 01, 2012 15:25

Jun 17, 2009 00:54

Subject: Re:Attack calculations

lorfje

Messages: 28,
Joined: Jun 12, 2009,
Offline

The maximum attack strength an Underling could bring to bear on your Plasma Tank is 3 + 2 + 3 + 3 (3 basic + 2 mountain bonus + 3 gang-up bonus + 3 resurface bonus). That is the same attack strength as the Plasma has against itself. As for attacking a Plasma with a Marauder, it is 4 attack against 14 defence and that difference is far too large and you will never get even one point of damage to the Plasma.

Jun 17, 2009 01:18

Subject: Attack calculations

rolando

Messages: 157,
Joined: Jun 10, 2009,
Location: California, USA
Offline

Also, on what terrain was your Plasma Tank? If you had it on a Forest terrain the Plasma Tank receives -3 in defense.

Jun 17, 2009 05:55

Subject: Re:Attack calculations

zCRP

Messages: 21,
Joined: Jun 12, 2009,
Location: Milan, Italy
Offline

what is the general rule for attacks? Is it simply attack - defense?
For example, Marauder has a basic 4 against heayv. If I attack with gang-up bonus I get another 3, and if I attack against Plasma in forest the defender has 14-3=11. Thus attacker has 7 against 11 for defender. Does this mean that attacker will lose 4 points ? Is there some "white noise" added? Do defenders have some advantage against attackers like in Risk?

Jun 17, 2009 08:36

Subject: Re:Attack calculations

jeffrey

Messages: 13,
Joined: Jun 03, 2009,
Offline

There is quite a lot of random factor involved. I don't believe that the defender or attacker has any inherent bonus.

Jun 17, 2009 09:14

Subject: Re:Attack calculations

qgcbuni

Messages: 2,
Joined: Jun 17, 2009,
Offline

Yes, admins, what is the formula? Nobody seems to know it.

Jun 17, 2009 16:48

Subject: Re:Attack calculations

jeffrey

Messages: 13,
Joined: Jun 03, 2009,
Offline

The actual math behind it is crazy complex and the random factor makes it impossible to predict the damage output exactly. I'll try to summarize it as I understand it...

The defending unit's defense modified by the terrain is subtracted from the attacking unit's raw attack modified by the terrain. (raw attack + terrain modifier) - (raw defense + terrain modifier)

Then bonuses are added to this number (GangUp bonus, resurfacing bonus)

This gives you sort of a base value. Then comes the random factor. At this point, starting with 0, a chance to deal +1 damage is rolled. The aforementioned base value determines how likely it is that you'll get this damage from the roll. The number of times this chance is rolled depends on the health of the attacking unit.

In short, the greater the health of the attacking unit, the more chances it gets to inflict damage.

So that's how it works... there are other minor variables as well, but this is the basic idea.

Hope that helps.

Jun 17, 2009 17:21

Subject: Attack calculations

lion37

Messages: 112,
Joined: Jun 13, 2009,
Offline

Adding any element of randomness to the game is a big mistake. It will allow people to hack into the client program (or create a fake client) that will make sure they always "roll" the favorable outcome.

I know this sounds a little technical, but the programmers in the Uniwar team will confirm I am correct. If you dont remove all randomness from the game now it will create big problems for you down the road.

Jun 17, 2009 22:06

Subject: Attack calculations

Anonymous

Lion37, that's ridiculous.

Jun 18, 2009 01:42

Subject: Attack calculations

lorfje

Messages: 28,
Joined: Jun 12, 2009,
Offline

Anonymous wrote: Lion37, that's ridiculous.

Not quite. I have had one or two times that I needed to replay a turn because my connection with the server went bad in-game. Probably you can force this issue to happen by moving out of wifi-range.
So, the game should communicate the state of the game to the uniwar server more frequently, and wait for a connection to be reestablished before play can resume (temporarily halting the game timer).

Jun 18, 2009 02:40

Subject: Re:Attack calculations

rolando

Messages: 157,
Joined: Jun 10, 2009,
Location: California, USA
Offline

Before everyone starts trying to beat the system by losing their connection to the server I will let you all know that though there is a random element to how much damage your units deal and receive the same amount of damage will occur in the same turn if you tried to replay it between those two units. The server communicates this to the client when it becomes your turn, so if you were to replay that same turn making the same moves the result will be the same every time. Once you have a new turn the outcome changes.

Here's an example:
In your turn you use a 10HP Marine to attack a 10HP enemy Marine and both lose 5 points. So hypothetically if you could replay that move over and over during that same turn with those same units the outcome will still be that both lose 5 points. But, if you were to wait to attack until your next turn and then you used the 10HP Marine to attack the 10HP enemy Marine the outcome may come out differently. A different outcome may also occur if you were to end your turn and your enemy decided to use their 10HP Marine to attack your 10HP Marine, but then the outcome would always be the same during your enemy's turn until it becomes the next person's turn.

Did I lose anyone?

This message was edited 1 time. Last update was at Jun 18, 2009 04:01

Jun 18, 2009 03:15

Subject: Re:Attack calculations

lorfje

Messages: 28,
Joined: Jun 12, 2009,
Offline

Interesting info, but it remains a weak point that you can, in effect, undo a whole turn as long as you did not finish it. Bad luck, bad strategy, nasty surprises from your opponent, there are many reasons you might wish you had played differently. The temptation of such an easy cheat will sometimes be too much for players and that would be unfortunate for the appeal of the game.

Jun 18, 2009 04:39

Subject: Re:Attack calculations

_jnc_

Messages: 60,
Joined: Jun 15, 2009,
Location: Switzerland
Offline

Thanks Rolando for the clarification!

While I think it is manageable to calculate the damage you are doing by attacking; I find it hard to understand how much damage the attacker takes.
Let me give you an example: If you attack an Eclipse with a Swarmer (just an example, I'm not saying it is usually a good move

) your swarmer will inflict around 3 damage and loose around 8 of your precious little bees. If you do the same thing with a Speeder, you will also inflict around 3 damage but only loose maybe 3 Speeders. Both attackers have 5 stats against heavy ground units, and the Eclipse always 9 defense.
BUT: Obviously the defending unit causes different damage to the attackers due to its different attack stats against aerial (12) and heavy (3). Does anyone see the formula behind that?

Just wondering as I had some serious losses due to miscalculation

This message was edited 1 time. Last update was at Jun 18, 2009 04:41

Jun 18, 2009 04:54

Subject: Attack calculations

lion37

Messages: 112,
Joined: Jun 13, 2009,
Offline

If the random calculation are really all done in the server (which I doubt) then that means you cannot play your turn offline and then just upload the result once your connection resumes. Not good.

If the random calculation actually take place in the client software then it is always theoretically possible to hack the client and make it work in your favor. This will be especially true once we get more clients (such as a web browser client, etc). Also not good.

This is a technical issue and I know exactly what I am talking about. Please consider eliminating all randomness from the game or you will have problems down the read.

Jun 18, 2009 10:45

Subject: Re:Attack calculations

Robin

Messages: 22,
Joined: Jun 02, 2009,
Offline

lorfje wrote: Interesting info, but it remains a weak point that you can, in effect, undo a whole turn as long as you did not finish it.

Each of your actions are saved as soon you perform them, so you cannot redo moves that you've already done.

lion73 wrote: If the random calculation are really all done in the server (which I doubt) then that means you cannot play your turn offline and then just upload the result once your connection resumes. Not good.

If the random calculation actually take place in the client software then it is always theoretically possible to hack the client and make it work in your favor. This will be especially true once we get more clients (such as a web browser client, etc). Also not good.

The random calculations are done on the client, but are seeded from the server. This is why the results are always the same.

Edit: quoted the wrong person

This message was edited 1 time. Last update was at Jun 18, 2009 10:57

Forum Index » General Discussion
1 2 3 ► Go Go to page...