playstrategy.org

Weekly Challenge 6: Sewing Stand Off

#1

Weekly Challenge 6: Sewing Stand Off

Play the most number of moves in an Oware game vs @PS-Greedy-One-Move without any captures by either player, then go onto win the game.

Adjudication Criteria:
1. Most moves (plys) without a capture from the start of the game
2. Fewest moves (plys) to win the game
3. Earliest Entry

Closing Date: Monday 24th April 1200 UTC+0

#2

I usually do a test game to show people what we expect from a weekly challenge entry. This week I found my game going rather well and got to 80 moves without a capture, but then remembered I had to also win the game and I had almost run out of time! This led to some panic moves and I was well behind once we started scoring but managed to pull it back to win by the narrowest of margins 25-23.

playstrategy.org/giB8ohwG

80 moves (160 plys) without capture
107 moves (213 plys) to win the game

Good luck all, I think this is beatable but will require some patience to do so!

#4

I don't even mange to win vs Gereedy one Move without restrictions

#5

I'm willing to give it a shot at least, maybe I'll learn something about Oware :)

#6

@HawaiiBlue Here's a short example game playing semi-greedily say (toward the end, just going for big captures)

9 ply without captures, then 7 ply to win.

playstrategy.org/RNYij0vvcvY6

I think the bot being greedy means you probably know what its next move will be in some cases. If there's a capture, it'll be the biggest capture.

I should over the 80 move game a bit closer and see if I can understand it :)

Edits (more attempts):

9 move/18 ply no capture, 9 move/17 ply to win.

playstrategy.org/eIENK0eU3nx4

16 move (31 ply) no cap, 12 moves/24 ply to win [slow progress on the no caps]

playstrategy.org/qPr6uoLZcbAF

#7

Here’s 28 moves (56 ply) no captures, and then 30 moves (60 ply) to win.

playstrategy.org/cH3Ryda3

I didn’t really play much Oware before and a few things I’m finding useful to know are for instance when counting out moves to split it up a bit to make it easier to count

E.g. if I have 5 seeds in a pit, three in front of it on my side, to think of 5=3+2 so I know it lands on the second pit on the far side (that’s quicker for me than counting ahead space by space).

Things like 12 in a pit lands in the pit in front, which will add two there (leaves the original pit always empty).

That sort of stuff helps to plan things out a bit.

Then a couple of small observations, for this challenge specifically:

- It seems useful to build up the first pit until it’s 12 or more so it’s guaranteed to be empty after moving it. It’s annoying with any less than that, because if your opponent ends up with a 1 in their nearest pit, or 2 two pits away they’ll immediately capture.

- A lot of moves are kind of forced, like constantly emptying one pit, when the opponent has moves on their side than can land in it on their next move. Sometimes it also seems forced to empty one pit before another, since when emptying a larger pit it could turn a 4 into a 5 or something which could hit an otherwise safe pit of one or two.

That’s just some things I’ve been thinking about to try and build up to this 80 move daunting target :P

#10

Somewhat new idea, following the original example game, try to make a huge pile on the second pit (B/b) and use that to basically reset the board and cycle though maybe not perfectly. It probably involves waiting a lot by repeatedly clearing the first pit (A/a) and eventually clearing some of the last few pits to again play some waiting moves, moving single seeds and such.

47 moves (93 ply) no captures, 20 moves (39 ply) to win

playstrategy.org/XtjxdJpW6fhb

Reconnecting