James Nathan: Yavalath and Manalath

Game: Yavalath
Designer: LUDI
Publisher: nestorgames
Players: 2-3
Time:  10 Minutes
Times played: 27 times.  (14 with a friend’s purchased copy, 6 times with my purchased copy, 7 times against the GDL in Dr. Browne’s LUDI Experiment III survey, ex post facto)

Game: Manalath
Designer: Dieter Stein and Néstor Romeral Andrés
Publisher: Included in nestorgames’ “Yavalath & Co” book
Players: 2
Time:  30 Minutes
Times played: 8 times with a purchased copy of Yavalath

If you’re just joining us, yesterday we discussed Cameron Browne’s PhD project where his computer program LUDI evolved and mutated an input set of games into new games (and provided them whimsical names). It’s quite fascinating and I encourage you to read it -or the thesis.

Yavalath’s rules are fairly simple.  On your turn, place a piece. If there are four in a row you win, but if there are three in a row (without having four) you lose.

That’s it.

I don’t know how LUDI would feel about three-player Yavalath, but that’s a thing.  Here’s a sample midgame. Black has lost; red and white continue to play.


If I have my history correct, and I may not, when LUDI came up with Yavalath, there were no published games that followed the n-in-a-row wins, n-1 loses, pattern.  I’ve played Yavalath a number of times, and, as has been my experience with abstracts, it plays quite differently depending upon the opponent.  From my usual gaming partners, to more casual folks, to the LUDI survey AIs, the games were their own.  In each, however, the game concludes with the losing condition having been met, as the cadence of the game often devolves into a series of defensive moves by the losing player where they are forced to create 3 in a row in successive attempts to block a winning move by the opponent.

The three-player version alters this dynamic slightly, but regardless, I like the game quite a bit.  I don’t say this pejoratively, but it reminds me of an “app” game – it has a certain 3-minute game addictive quality to it.  It’s not the deepest of abstracts, but it fits comfortably in its niche, and I find it hard to put away. “Just one more game…”


When I e-mailed Cameron to ask him some questions about LUDI, he made sure I was aware of Manalath (which I had not been), and said that he found it much more interesting.

Oh?  Here I was enjoying Yavalath in my ignorant bliss, but forever there is now Manalath.

Manalath follows the n wins, n-1 loses structure, but with some clever twists.  First, we’re switching to groups rather than strictly in a line. Second, we’re creating a maximum group size (some moves will now be off limits, not just ill-advised).  Third, well, we’ll get to that.

So in Manalath, the numbers are 4, 5, and 6: four in a group loses; five in a group wins; 6+ in a group is not allowed.  One wrinkle, is that we’re only going to check victory/loss conditions for each player at the end of their turn. Second wrinkle, (the “Third” from above), is that, well, you can place either color on your turn – that’s not to say you can switch what color is yours, rather, you are simply allowed to play your color (e.g. black), or your opponents (e.g. white) (and now it should make more sense why we need to only check victory/loss at the end of your own turn).


When I said earlier that there were no published n/n-1 games at Yavalath’s evolution, that’s a nuance to allow for Dieter having a draft of Manalath, but not a fully fleshed game (again, if I have my history correct, and I may not).

In a sense, without the LUDI background of Yavalath, Manalath feels like the finished product of something that Yavalath was the proof of concept for.  This is no longer a somewhat time-passing puzzle, and it is now an engaging back are forth – each player writing puzzles for the other, and needing no more and no less than every tool Dieter has given them to build and escape various “check” states. (The time frame is also perfect for me – substantial enough that I feel like my neurons have gotten a work out, but I also have no taste for longer arc abstract games.)


Currently, Manalath is available in a book of other games playable with a Yavalth set, “Yavalath & Co”, and while there isn’t a stand alone published version of Manalth, Dieter recently shared a photo of a beautiful wooden board, and Nestor has said that a stand alone version is coming (which won’t be the wooden board).

Even more recently, Dieter posted an image of what appears to be a Manalath app “coming soon”.

So if Yavalath was No. 2, what about No. 1 and 3-19?  That is tomorrow’s topic.

Ratings from the Opinionated Gamers


I love it! James Nathan, Eric Martin
I like it.
Not for me…


I love it! James Nathan, Eric Martin
I like it.
Not for me…

Related Reviews:

Monday – Cameron Browne’s “Automatic generation and evaluation of recombination games”
Wednesday – Ndengrod/Pentalath, Valion, and Elrostir
Thursday – Volo, Feed the Ducks, and assorted puzzles
Friday – Brief Interview with Dr. Cameron Browne

This entry was posted in Reviews. Bookmark the permalink.

11 Responses to James Nathan: Yavalath and Manalath

  1. Pingback: James Nathan: Cameron Browne’s “Automatic generation and evaluation of recombination games” | The Opinionated Gamers

  2. Love this series of articles, James! Thanks much for writing them.

  3. xitoliv says:

    Thanks, Eric. And thanks for reminding me that Yavalath existed. It’s interesting delving into abstracts a bit, as almost everything seems innovative.

  4. mlvanbie says:

    The previous article discussed metrics for evaluating games. How does Yavalath compare with Manalath under these metrics?

    • xitoliv says:

      It’s hard to say. Manalath wasn’t part of the LUDI program/process, so it wasn’t ranked by the same survey participants.

      In his thesis, Cameron stated that the six stand out criteria that appear to be the most important are:
      Uncertainty (late) (“the tendency for the outcome of the game to remain uncertain for as long as possible”)
      Killer moves (“an estimate of the tendency of killer moves to occur in a game”, defined as “one that significantly improves the player’s position and typically swings the outcome of the game”.)
      Permanence (“the tendency for players to immediately recover from the opponent’s last move in terms of board position”)
      Lead change (negative) (“the tendency for the lead to change over the course of a game”)
      Completion (“the tendency for games to reach a win or loss within a reasonable number of moves”)
      Duration (negative) (“a measure of the average number of moves required to complete a game”)

      Both do fine with coming to completion in a reasonable duration. I do like the longer duration of Manalath, as it allows more time for the drama to unfold.

      For the others, I’d personally give the edge to Manalath. The differences between several of these criteria are very subtle, so the broad reasoning here applies to several, but in two-player Yavalath, once you’re on the ropes, it feels hard to recover. In Manalath, my experience has been that you can be on the ropes, but make the right “killer move”, and then possibly gain the upper hand.

      The criteria were used to triage the 1300+ results, but it was up to the human survey participants to rank the final 19. For my rankings, Manalath would be No. 1.

  5. Pingback: James Nathan: Ndengrod/Pentalath, Valion, and Elrostir | The Opinionated Gamers

  6. Pingback: James Nathan: Volo, Feed the Ducks, and assorted puzzles | The Opinionated Gamers

  7. Pingback: James Nathan: Brief Interview with Dr. Cameron Browne | The Opinionated Gamers

  8. Pingback: BGGCON Day Three (2018) | The Opinionated Gamers

  9. Mikk says:

    Interesting to see that few people mention Renju (proffesional go-moku) when talking about Yavalath. It’s a classic game in Japan, and while the n-1 loses rule only applies to one player, it is a notable precedent for Yavalath

    • xitoliv says:

      Thanks, Mikk. I wasn’t aware of that one before. At first glance, the distinction is that Renju disallows creating multiple patterns of n-1 or n-2 rather than doing so triggering a losing condition (though perhaps that’s not an important technicality.) From what I can see, n-1 is allowed, it is just more than one group of n-1 that isn’t permitted (and even then there are exceptions for when the opponent’s stones block the group.) It is also quite interesting that all of this only applies to one of the players.

      I’m glad you brought that up. I like to be overly thorough with citing the heritage of games, and while it’s hard for me to call it a precedent for Yavalath (as it wasn’t included in the gene pool -though Dr. Browne did include 17 variations of Gomoku, none appear to be versions that used any of the Renju restrictions), I think it’s certainly worthy of a mention.

Leave a Reply