Hi, designer of Tumbleweed here. First of all, I'm glad to see the game discussed here! But, more to the point, Tumbleweed is an extremely simple game in terms of mechanics. Unfortunately, I'm a pretty bad explainer! I agree with Carroll that the ruleset given at igGC is simpler. If anyone has ideas for further improvement, clarifications, feel free to hit me up!
The mechanism is not complicated, it's just surprisingly difficult to explain with words. Actually playing the game makes it click.
Definitely a game that would be welcome here and benefit from tournament play, but it's lengthy for a modern abstract, usually 100-150 moves on the smaller board (not a bad thing, just unusual nowadays).
There are also some videos on Alek Erickson's channel, showing the game being played in Ludii. It might give you some idea about the mechanics, tactics and strategies:
@metzgerism, I know there has been some casual play on hex7. But I firmly believe hex8 to be the least size where some strategic nuances really get to shine through.
I take advantage of the annoucement of this promising new game to draw attention to the existence of not very edand however very good games created some years ago. I especially think of Ayu, which is an exceptionally deep and original game , with unique tactics and unfortunately very few players. Ayu may however be played at the excellent mindsports site.
When it comes to Tumbleweed's depth, here are the facts:
In terms of combinatorial complexity, size 8 Tumbleweed can be assumed to be above 10^100 and size 11 to be above 10^250. For comparison, 11x11 Hex is 10^98, Chess is 10^123, Amazons are 10^212 and 19x19 Go is 10^360. This lower bounds arevery conservative, and we can't know for sure until stronger players emerge. The real question is effective complexity, which stems from effective game length: how many moves does it take for the result to become obvious? Stronger players will avoid blunders, but they will have better foresight, so this could go either way.
In terms of strategy and tactics, there are some easily conveyable principles that can be taught to, or even observed by, beginners. Many of them sound very natural, almost like a story. At the same time, there is much tension between those principles, and applying them often isn't easy: for instance, in the opening there is a very clear dilemma between spreading out as wide as possible and keeping your pieces secure. While the player base is still fairly small (24 players on players on igGC, plus no more than 10 others on MindSports, Ludii and TTS), a fairly robust skill ladder is quickly being established.
So, to sum it up, Tumbleweed has high combinatorial complexity, learnable strategic basics and a lot to reveal later on. This is probably enough to establish strategic depth, so I agree with @Clark94*. That being said, I don't feel comfortable comparing it to games it has little to do with (like hex or chess), and I think it's definitely more "shallow" than Go.
\_\_\___
\* Of course, this assumes that there aren't some pathological game-breaking strategies! There don't seem to be any so far, but you never know... That being said, any potential flaws could probably be fixed with simply changing the opening protocol accordingly.
There are some notable changes compared to the igGC version:
The board is rotated 60 degrees
The colors are blue&red instead of red&white
The pie is reversed: you set up the second mover, then the first mover
There is no pass button and no automatic influence scoring: the players are trusted to count the score by themselves and resign accordingly
Overall the MindSports system (both the page at large and this specific implementation) isn't awfully convenient and I treat Tumbleweed being there as a badge of honor more than anything else, but there are some competent players who use it.
I add that we find at mindsports many very good abstract games which cannot be played elsewhere. They are carefully selected by Christian Freeling; and of course his own ceations may be played at the site.
I would definitely love to see Tumbleweed here. It has my vote for best Combinatorial game of 2020, which does not quite do it justice. I'm playing on MindSports right now (wish it had a better mobile interface and better notifications).
coldsalmon, Zapala, myself (spartacu5), Clark94, Metzgerism, Liberte, and others.
Tonight on IGGC I will be playing my 3rd round Tumbleweed tournament match against "atari", who is one of the top-ranked players and a dan-level Go player.
Just wanted to add my voice to support the implementation of Tumbleweed on LG.
Tumbleweed really gives me a feeling that it's a very deep game. It might be the most interesting game I've discovered since I discovered go more than a decade ago. I would be delighted to play it at a turn-based pace with the crazy LB brains.
I would also like to see it implemented here... it's been difficult for me to play live on iggc (work, family, etc), but async play has been relatively easy.
Here is an instructional video about Life in Tumbleweed. Life is very different than in Go because groups can get eaten away at the edges. However, it is similar in the sense that collectives of stones help each other live - as in strength in numbers.
Tumbleweed will be played in this year's Mind Sports Olympiad! This tournament is open to anyone, so it's a good chance to get into the game while we wait for it to hopefully get implemented here one day. Here is the registration.
Just voicing my support as well for Tumbleweed on Little Golem. I believe it's the best abstract to come out in a long time, and I think its popularity could grow if it were added here! :)
Tumbleweed is currently in Alpha at boardgamearena. If you are also a user of that website, please let me know your handle and I can invite you to the group for testing the implementation. Thanks!
William Fraser: I'm not sure if there is a way to get around that (I tried to invite you, but I'm not sure if you were allowed to join). In any case, I'm happy to play some games against you before the tournament starts so you can reach 100 elo!
I just received my copy of Abstract Games magazine in the mail. There is a short article about Little Golem. Richard says in the article that Tumbleweed is "slated to be next".
The Mindsports Olympiad Training tournament is happening tomorrow at 7:00pm CEST on BGA. If you happen to be below BGA’s 100 Elo requirement but would like to join anyways, let me know your BGA username so I can invite you.
I think the implementation is correct - even if it is not the same as the original. Original rules: first player places two stones (red and then white). The second player plays as the red player or does a 'swap' and chooses the white stones. So the first move is made by the opponent with the red stone. In any case, the player with the red stone starts the game.
The implementation on LG is the same, but the players' colours don't change - but the stones are swapped. So the player either swipes his turn or makes a swap (the stones on the board are swapped) and the opponent makes the first move.
That swapping of stones on the board allows you to mark who is white and black in the game, and that doesn't change in any game implemented on LG. Similarly, swap is implemented in other games on LG (hex, havannah).