Does anyone tell me if this behaviour is a bug or a feature of warlight: If you attack a territory (assume a zero luck/60% defendor losses/70% attacker losses) with 10 troops from another with 10 troops, then the warlight algorithm will define that the 10 attack troops lose .7*10 = 7 troops. The defending territory will lose .6*10troops = 7 troops. So, on the map before the attack there was a 11 troops attacker territory and a 10 troops defendor territory and after the attack in both territories remain 4 troops. So far so good. Now one might think that the same result should be, when these 10 troops attack from say 2 territories each 5 troops. The defendor again loses the same amount of troops. But the attacker loses ALL his troops: in the first attack he loses 10 *.7 troops = 7 troops. So he loses all his troops, the defendor just 5 * .6 troops = 3 troops, leaving 7 in the defendors territory. In the second attack 7 * .7 troops or most likely all attacking troops are lost.

Is this nonlinear behaviour of warlight a wanted feature? And if yes, then why is it?

the problem is that your looking at the situation backwards, when you attack with five, the defender looses .6*5, regardless of how many defending troops there are.

http://warlight.net/ViewHelp.aspx?s=Attacks

this explains it better

I think the problem is, that the defender always defends with all he has in his territory. In example above the 10 troops. If a strong troop of 10 attacks him, he naturally defends with 10 troops. If 2 attacks in a row with each 5 troops are done, then he defends first time with 10 troops, second time with whats left, in this case 7 troops. So in total he defends with 17 troops.

So, during one turn, never attack multiple times the same territory with small troops if you have the choice to attack from one territory with the same total amount of troops.

yeah, because your two territories don't attack simultaneously.

its not a problem, its meant to be that way

"So, during one turn, never attack multiple times the same territory with small troops if you have the choice to attack from one territory with the same total amount of troops. "

only if you know they won't attack out of it and expect/want to kill all their armies.

There are many times, they will have 10 armies someplace in a theoretical defensive possition, and i will place say 3-4 armies in each area bordering it.. and use the other theoretic 1-4 armies leftover in my back area as stalling moves.. that way presuming he attacks, if he splits and send 5 each way, he loses both times, with 4*.7 and 4*.7 vs 5*.6 and 5*.6 which is a 5.6 to 6 lose ratio with less armies on my side used.. or he attacks all out, in which case my dual attack can capture said location from the area he didn't assault.

thats how i handle it too perrin, but I will usualy load one side as kind of a guess as to which side he will attack, usually a 3 and 7 split....you can figure out odds on that yourself, its not worth me doing it lol

it really depends on situation, on which situation leads into further bonuses better, and whether or not you think that person will do that, or intentionally not do that so as not to be predictable although that is more often a situation of seperate bonuses, *quebec vs E.US/Finland Vs W.russia/California vs Mexico/etc.*

but back to the original thought by mosq, this is intentional as i understand it.. picture individual people, if 10 guys aer sitting there, it's better to attack them with 10 guys now, then with 5 guys now, then wait a minute and 5 guys later.. Ideally it would be nice to be able to put in flanking bonuses of a sort, but idk if that would really be idealistic and would make the game just more complicated, shifting the distance from noob-vet to a range idk if fizzer would like..

It's a feature. It makes sense from a war standpoint as well. 300 Spartan soldiers held off a small pass for a long time against an enormous Persian army despite said Persian army having guys that were allegedly immortal (but died anyway) and also having a giant executioner dude with weapons where hands should be and chains running from nipple to nipple. If the Persians had used a priority card and came in with a big stack it would have been over easy, but they fiddled around with some delays and then sent in a bunch of 1 and 2 attacks late in the turn and didn't get the job done.

In terms of what is best tactically, it really depends. If you're looking to power through the defense in a specific spot, then you want to load up entirely on one territory. If you're looking to defend a few places, and then crack back later in the turn after they've attacked you in hopes of capturing their attacking spot, then it all depends. Sometimes you want to evenly distribute, sometimes you'll want to heavily distribute to a few places and lightly to others and sometimes you might even want to go all or nothing on a specific spot. It's based on your income versus their income, how many armies are already on the spots, and what you expect your opponent to do.