A solvable game is a game whose winner can be predicted from the start, even if both players play optimally.

Do you think Small Earth falls into this category? Do you think some boards are solvable and others aren’t? If so, what percent of boards do you think are solvable? In a sense, do you think move order could determine the outcome of a game from the board alone?

My theory is that Small Earth would be solvable if it had no fog, but the fog conceals info from the player that would have normally have won before the game started, in a way that nothing can be predicted with 100% accuracy. One could argue that player A could determine the optimal moves for player B each turn, so they could in fact predict every move and play as if there were no fog. But what if player B recognizes this and plays suboptimally, to undermine player A’s strategy? Would player A then play suboptimally to counter this counter? Would player B then play optimally, hoping player A would play suboptimally? You can never be 100% sure which strategy the other player will go for, and that is why I believe in a fog game, Small Earth is not solvable.

I am very interested to hear your thoughts on this!

No boards are solvable from the start, because unless there is a significant income advantage you cannot guarantee that you'll be able to defend two territories of a bonus at once if your opponent has a chokepoint.
(For instance: https://www.warzone.com/MultiPlayer?GameID=23156027)
That being said, if the picks of your opponents are known from the start, for instance if you get 1 and 4, certain boards become solvable immediately, though those are rare. The only solvable games are ones where there is a significant income and positional advantage and one side has enough income/troops to counter all possible attacks. However from my experience, what determines the winner of a game on se is hardly ever calculation, but instead tricking your opponent and creating positional threats.

I think it is possible to create an ai that can play the template almost perfectly, or with something like a classical chess engine in which it assigns numerical advantages, but even then the position will only be solved once one side has a net income advantage.

Edited 7/17/2020 01:34:37

You just used this thread to flex that you beat Ozijs! :P

fine ill pick one that I lost, k?

Lol no I get your point. But I think there are boards in which you would never have to defend a double border? Or at least ways to avoid having to defend a double border?

There are positions in which you can indeed do that, especially if you attack the chokepoint before your opponent.
https://www.warzone.com/MultiPlayer?GameID=23222615
This game demonstrates both that and the above point, and I lost. Happy?

Some boards have picks that if gotten are impossible to counter. So yes some boards are "solvable."

Most aren't.

I found a solvable map: https://www.warzone.com/SinglePlayer?PreviewMap=38593

@Orannis: Regarding playing "optimally", optimal WarLight play requires you to use a random number generator in some cases. I have also calculated how to set the RNG up for a certain restricted example in my guide here: https://docs.google.com/document/d/1NyhCpIQKShAbWGXicO_ph9whV_UyMS-3a7_re0C7R8Y/edit#heading=h.y6pv3soufztp. So even if two players choose to play a certain situation out differently with 1 players choice causing him to lose the game it's still possible that both players have played optimally.

I'd like to formulate your question differently:
"Is it possible that I can send my opponent a screenshot of the board as I see it, including telling him my moves and there still is nothing he can do about it?"

Although this question sounds kinda "academic", the truth is that thousands of WarLight players solve this riddle on a daily basis since this is when you should press the surrender button if on the losing side. Practically you might still hope that your opponent throws away his win.

About whether this could be true for the start you haven't stated thegame settings (auto / manual picks, wastelands?). If believe such games should be exceptionally rare from the start. If it's with manual picks, the loser had to come into this bad situation despite picking optimally so you could also state your question differently again: "Are there small earth games which are totally dependant on picking luck?"

the surrender tutorial is one sided

There are indeed boards that are losing from the start, such as this one: https://www.warzone.com/MultiPlayer?GameID=23272883

You could have hit North Africa full force, Kenghis Ghan. Or you could have outpredicted your opponent in Europe. This game wasn't lost from the start.

Maybe, still not good picks.

Edited 7/21/2020 14:53:28

I dont think the gameplay is completely solvable, but the picks are. I am convinced that there is a best pick set on almost any board. Often it even favours one of the two pick cycles

Edited 7/21/2020 15:02:01

You could have hit North Africa full force, Kenghis Ghan. Or you could have outpredicted your opponent in Europe. This game wasn't lost from the start.

Not lost, but losing.

yeah I studied the KG game extensively. Trying to see if I could construct a moveset for Green that was unbeatable. I can't. All of them devolve into 50-50s.