Watch your opponent's deployment habits. Most people have a bit of pattern they stick to.

Once you see what it is, you can try to figure out a countermove. If your opponent spreads out his armies to cover all the bases, concentrate all your forces in one place. If your opponent likes to create a large force and barge through, make a few small attacks here and there, and one or two might get through.

Think like a retard.

This has won me a few games. I usually go for the worse strategically but less predictable option, but that loses a lot of games too. Knowing when to be unpredictable and when to be more predictable, but strategic is important. How to identify this? Depends on the position you are compared to your opponent.

Know the value of taking each territory over the other. In a 1v1 you can normally guess which one they will defend depending how important the territory is, and also how desperate their situation is. If you go into these situations with a stack, it normally changes this up as well.

Its partly knowing the playstyle of your enemy
and partly knowing the current situation. With that information you can usually predict fairly well.

If the player is completely unknown to you, it`ll be more dependend on:
1. Your knowledge of the map. There are usually territories that are attacked more frequently
2. The skill-level of the player you are facing. This can be based on his 1v1 ladder rating. If your enemy doesn`t have one, the best indicator will be the winrate.

Inconceivable

Know the value of taking each territory over the other. In a 1v1 you can normally guess which one they will defend depending how important the territory is

bullshit because wifom

Not always: sometimes you and your opponent have a very different view of the map (due to fog), and then it's not _inconceivable_ that you might know your opponent's priorities without him knowing yours. But otherwise, yeah.

I dunno what wifom (edit: but the penny dropped after I sent the comment :) ) is, but I think that calling advice from current No.1 in the world "bullshit" is daring at least :)

@Lolowut: Without context I had no chance to understand. Moreover, it looked like a typo at first sight ("wine" instead of "whine"), I am sorry for being so suspicious to you...

Thanks all for your responses. They really helped me to evaluate how to handle such situations in future (and no, I am not about to rolling dices).

Edited 4/7/2016 20:43:57

WIFOM explained:

You're a cop, and you're chasing a criminal. The criminal happens upon a fork in the road, there is a well-lit road and a dark road. The criminal takes a road. You're now at the fork in the road. Here's where it becomes circular logic:

The criminal has an advantage if he picks the dark road. There's a chance you could take the dark road and the criminal could still get away because the lack of lighting could throw you off of chasing him. However the criminal knows this, and he knows that you know this as well. So since he knows that you know the dark road is an advantage, it's quite obvious to pick the well-lit road! However, he could've thought this through and picked the dark road, thinking that you would think he would pick the well-lit road instead.

This thought process can continue on and on. The point here is MOST multi-bordering situations are this exact situation. There's a territory that gives a positional advantage AND breaks the bonus, and there's a territory that's less advantageous but still breaks the bonus. There is no right answer on how to handle this. You can say: Well study your opponent! They have tendencies. However, this could be the statistical anomaly game in which they do something against their tendencies. The information in the game is not useless, but the opponent has access to the same information.

Also: I would consider a game in which to win the game you need both positional advantage AND the bonus broken in order to yield an advantage not a WIFOM moment. If only 1 of the options gives you a direct advantage, then the circular logic game shouldn't be played.

Edited 4/7/2016 20:49:21

and no, I am not about to rolling dices

Then you are throwing away the only mathematical sound solution.

Thanks for thorough explanation Lolowut.

Then you are throwing away the only mathematical sound solution.

I do, and for good reason - I don't play against mathematical entity, but against human. And as said above, humans have tendencies. Obviously, there is a risk that opponent will trow his tendencies out of the window in any particular game, but that probability is lower than 50% at any given moment.

Edited 4/7/2016 21:07:34

And then there's the unexpected option of hiding in a bush and shooting the cop :P

Easiest way to be unpredictable is use an RNG.

If you choose which territory to attack randomly, your opponent cannot predict your attack--they can only get lucky.

Game theory would likely say that no choice is always better. That is, faced with the identical situation 100 times you should sometimes choose one and sometimes choose the other.

And then there's the unexpected option of hiding in a bush and shooting the cop :P

Very sound logic actually. What if, there is territories A, B, and C?

Territory A has a doubleborder on B and C. The opponent predicts that the player will strike territory B, since it is the most strategically viable. Because of this, the opponent plays an order delay card, creates a stack in Territory C, and on last order attacks Territory A.

But, Opponent predicted wrong. The player didn't attack at all! Instead, he just built his stack, and the opponent got stackwiped since they predicted incorrectly. A third option, like you suggest.

Maybe this only makes sense in my head, or I overthought it too much ;_;

Ox, I think you misinterpreted her message as more than just banter. We're usually talking about a situation in which the opponent's best move is to accurately guess which territory will be attacked, in order to defend, and the player's best move is to accurately guess which territory will be defended, in order to break the bonus.

Yes, you can expand the scenario and it can be more complicated, but the straight up: How do I know which territory to defend/attack if it's a straight up 50/50 guess?

The answer is: There's no RIGHT way to guess it.

And then there's the unexpected option of hiding in a bush and shooting the cop :P

I'll just take this as letting them take both territories but then revealing a stack of twenty you were amassing all game lon right next to him and going into his territories brutally :P

I don't agree, sometimes a territory in warlight is just so important you can't run the risk of losing it, even if you know the other player knows this. This is probably even more likely when you don't need to risk it to win.

Buns: sometimes you can exploit that though. If your opponent knows the territory is that important, they may "know" you will defend it, so they won't waste armies on what they believe is a futile attack. Then you can defend it for "free" or perhaps only a token defense, while you use your armies to your advantage elsewhere. A gamble, to be sure, but can be a big payoff.

I know that, but what I was saying was that why take that big gamble, when you have the advantage holding the territory even with a broken bonus. I know what you guys are on about, but from what I seen people only take that gamble when they're in a losing position anyway, so you know when attacking it's more likely of them to do it when they are at a disadvantage.