Go variant prototype: "Marks"

#1
Hey, here's another shelved game prototype! :p

People have been posting theirs recently and it's really inspiring to see all the ideas!

This one's a kind of Go variant, inspired by the discussion in this thread: http://www.dinofarmgames.com/forum/index.php?threads/improving-go.2513/. Basically an implementation of the idea in post #15. I gave up on it because I was having trouble making the gameplay feel any more engaging than Go does. I tried adding some little twists (see below) but I don't think they completely worked. It would be great to hear if anyone has any other ideas for things to try.

The concept
In short, the idea was that if early-game Go is mostly about heuristic decisions to do with passive stones staking out territory, and late-game Go is more about capturing and calculation, then maybe we could reduce the need for calculation by extending the early game all the way to the end. Nice theory!

It required a radical change to the rules though, because in Go, claims on territory are made by the implicit threat of capture later. But if we have no late game where capturing will take place, the stones need to claim territory explicitly. This leads to...

The rules
Rule 1. Players alternate placing stones on vertexes of the board until all are occupied. But stones are never removed.

Rule 2. Stones are directional: each points in one of the 4 diagonal directions . Once placed, the direction is fixed. Each stone exerts a set total quantity of 'influence', diluted over all squares in the quadrant it points towards:
marks1.png
This red stone is influencing 4 squares, and placing 50 influence on each one.
marks2.png
The blue stone above is influencing 15 squares, but spreading its influence more thinly in proportion compared to the red one. (More below on why influence-per-stone is currently 200).

Rule 3. Having more influence than your opponent on a square means you own that square.
marks3.png
The UI displays the square in the owner's colour. In this game red has come in nearer the corner and claimed some of blue's squares by being more committed to them than blue was in the previous picture. The 25 red influence now beats the 13 blue, leaving 11 net for red (decimals maintained but not shown).

Rule 4. At the end of the match, the winner is the player who owns the most squares. And that's it!

Big problems
For me at this point, the gameplay felt bad - even worse than Go :cry:

In the absence of any wise senseis, you have no idea really what to play and there are so many identical-looking options. So far so Go! But unlike Go, there is never any of that nice sense of a discrete state space (however large and daunting) which is a big part of the appeal of tacticsy games. This game felt so intangible, like you were forever trying to grasp ambiguous ownership of squares which then got grasped away again by the other player; nothing ever felt clean or stable.

Another really bad, perhaps even insoluble, problem is how numerical it is. You need to have a really good memory of how 100 divides by different numbers, so you know what your stones are going to do, and then you've got to visualise how much influence you actually need to take control of opponents' squares. I tried to get the UI to show some of this but it's not great.

The game in this basic form is probably even harder, and even less approachable than Go, which is saying something. Perhaps that's a good thing though haha!

So how about some some rules changes which might give players concrete short and medium arcs to think about, rather than those only emerging through an understanding of deep long-arc strategy.

Idea #1: Locking in squares
Clearly by the end of a match we would want our owned squares to be ones where we had 'wasted' the minimum amount of influence to own them, and we'd want our opponent to 'waste' more influence than necessary over-committing to get their squares. (Since both players will be placing an equal number of stones, and therefore an equal total of influence spread across the board in the whole match, the winner will be the one who wastes least.) But one of the annoying things about the game is that every time you claim a nice group of squares, your opponent's very next stone often takes most of them straight back again! You feel like you've 'paid' to get ownership, but soon you have to 'pay' again, and then again, and again. It's like trying to hold onto a wet bar of soap.

So I thought how about if we could make such a strong claim on a square that it would prevent further claiming back by our opponent. We would overpay in the short term, but then not have to waste any more influnece there, and any influence our opponent placed there as a side-effect of their moves would have no effect and be wasted. This suggested:

Rule 5: If a player reaches net +100 influence on a square they take permanent ownership of it:
marks4.png
In the above picture we've come back in again and put a stop to red's shenanigans by claiming at least the corner square FOR GOOD! The UI no longer displays a number on it and fills the square in completely. (Incidentally the general UI convention for displaying influence in this game so it'd be intuitively graspable when assessing the whole board situation was really difficult. It's another place I'd love to hear if anyone has better ideas...)

In practice I originally had each stone placing 100 influence, but it seemed that with these numbers the mechanic of locking in squares didn't really come into play. (That was in matches against the dumb AI which just makes random moves, so perhaps it would work better in human v human?) Anyway I upped it to 200 per stone for matches against the random AI and I think it played better, with plenty of opportunities for locking in squares. This is an open question about this design. 100/100 is definitely more elegant IMO.

Here's near the end of that match where there are quite a few locked-in squares.
marks5.png

Idea #2: Bonus squares
This was a further attempt to give players some kind of explicit short-term goal. The board is generated with some squares which are worth more than 1 in the final tally-up. Taking control of these should give the players something to fight over in the early game, reducing the need to have expert-level understanding of long-arc strategy before you can even do anything sensible at a tactical level.
marks6.png
The squares with darker background are worth 5 and if I was playing first I might decide to straight away lock in the 2 squares in this corner and guarantee the 5 bonus.

I played one match against a friend with bonus squares and indeed we found that trying to claim them dominated play. Whether that's because we just had no idea of any better strategies, or whether bonus tiles are actually a good idea for this game, who knows! One thing I know is that they're OP against the random-move AI, for obvious reasons. Incidentally playing against a random-mover isn't completely pointless, since if you give it a few squares headstart it's not a trivial matter to catch up from there!

Idea #3: Walls
I was starting to lose momentum on the project at this point, so this was the last thing I tried, but maybe it's the best? Scattered around the board are barriers which block the 'spray' of influence emanating from a stone. Each stone still places the same quantity of influence, but it's 'focused'. Like so:
marks7.png
This is the move preview showing how red influence will be spread out if a red stone is played here. The walls are being used to screen off parts of the stone's normal quadrant of influence, effectively focusing it into a narrow area of tiles which wouldn't have been possible with just the normal open board. I imagine strategy would also involve finding areas 'sheltered' behind walls which were easier to defend?

I think this is a really promising thing from a gameplay point of view. Also, since the walls are randomly generated each match it addresses another big problem of RNG-less games where standard openings and playbooks can outweigh any kind of creative skill in determining the winner.

Conclusion
Thanks for reading if you made it this far! :coffee:

If anyone has ideas for how to improve this game please post them!!! I gave up on it due to approximately equal quantities of:
  • running out of ideas
  • feeling that the need for on-the-fly mathematical calculations was undesirable, and maybe unavoidable
  • not having enough people willing to playtest against, and no AI-writing skill
  • the temptation to work on designing something new (so much more fun than programming!)
But I'm not so sick of the prototype I wouldn't put in new ideas if there were any!

Oh yes there's a short video in community games on the discord too!