I will say that at theoretically perfect play, this is bad, it basically just gives your opponent information for no reason and helps them deduce what is the optimal play from their position more easily. We are not perfect robots though, so I can see your point about humans not being able to fully take advantage of such a move, and could easily mislead them into a making a mistake they otherwise wouldn't. I will say, I thought about case 1 more in depth and realized that if you are the East Russia player, you can logically deduce that the ghost move into East Russia is a bait move. I say this because, if your opponent was actually planning to counter East Russia, they would have no incentive to reveal to you they are in Central Russia, because it decreases the probability that you will attempt to capture East Russia. Essentially, I believe case 1 and 3 mostly just reveal to your opponent one of your spots, and that you don't know one of their spots, I think this could be taken advantage of. Case 2 definitely intrigues me the most, because this seems far more likely to influence your opponent's play in a way that's advantageous for you. Revealing SEA makes it pretty likely that your opponent is more likely to move the attack in Shanghai to last order, since simply rearranging the order of your moves doesn't hurt you as much as revealing information. You could of course, always make the argument of "But what if your opponent knows you are thinking this way, and as a result will play in a way that you don't expect?", I would argue that you can predict this line of thinking and play in a way they don't expect, then they could predict your new line of thinki... I'm sure you get the idea. It essentially just becomes rock-paper-scissors but with extra steps.

I will say, this is a very interesting idea Rufus has come up with, and thank him for sharing this with the community. I would be curious to see other more complicated examples of this tactic being used that could have a stronger case for being useful, because obviously the examples Rufus showed are mainly to just show the basic ideas of this concept. In theoretically perfect play, I stand by my conclusion that this is never an optimal move to make, but I do think with more complicated examples, it will be much harder for a human to make sense of what is happening and make the best move as a result. I'm excited to see this expanded upon more.

I feel like this game is totally a good use case.

https://www.warzone.com/MultiPlayer?GameID=36592360

Player Beep wanted a nice Russia triple pick, but unfortunately lost the first pick.
Now Beep needs to distract their opponent from the Russia combo, and they know how their opponent will expand, and that they will likely take Tomsk in turn 2, as that is just how to naturally expand.
By giving away the info of West Russia pick on turn 1, they will steer the attention of Central Russia towards it, and can take Tomsk themselves in turn 3.

Now I get that this scenario is a little staged, but it's not too far from reality, I had situations like this.

Something really funny occurred to me right now:

What if from now on forwards we the strat community started to use Transfer-Only all the time when we do conditional delays? The mindblowing reason for this is by doing that opponent doesn't exactly know if we have the territory where the delay was aimed if it gets triggered. That way you can try to fake that you came from a territory that you don't have. Once again, if condition is triggered and the move wouldn't work either way.

We can change the meta together!

Likewise, you can make attack-only moves to territories that you already own...

You can also when you full deploy and attack your opponent and say you have an attack of 10 armies then, just enter 13 armies into your attack and make it seem like you have 3 income more than you have ;)
The problem is, nobody (expect us who won't get tricked by it) checks these details anyways.

About the faking income it has been done waaaay before, so that's nothing new to some of the players.

Yes, sure, sorry I used your thread as a metagaming-main.
Anyways, people who play army cap next CL should start thinking about using this, as it can be especially useful there to confuse whether you have armies somewhere in the fog.

It's fine, Beep. Always exciting to hear some ideas about this, so let's keep going.

Speaking about army cap. Juan once came up with an idea that if you really really want to get rid of some of your income (like you can't even get rid all of it in any moveset) then you can use Autocommit. But this is an extreme case and we don't have such templates.

Edited 12/31/2023 12:23:13

https://www.warzone.com/MultiPlayer?GameID=13075796

Beep Beep you are an AI, we all know that :)

lol nice game? off-topic much?

@giantfrog the “just good for making an average move” was referred to the final randomisation process after discarding dominated moves. So basically if we play a 3 player rock/paper/scissor tournement you, me and AI, and I play flawed, you will like win the tournement exploiting my weaknesses, and the AI score second by playing flawless.
So my point is that in a predict game like wz is, the moves by Rufus may influence the game in practise, which in theory (between AIs) would not.
Ofc WZ is way more complex than R/P/S, and I agree that an AI would be hard to beat in a tournement.

Edited 12/31/2023 19:14:29