Suddenly, two buses – another new arrival in the Go family General forum

10 replies. Last post: 2018-12-21

Reply to this topic Return to forum

Suddenly, two buses – another new arrival in the Go family
  • Richard Moxham at 2018-12-17

    Back in the spring, Nick Bentley unveiled his latest brainchild: Blooms. As many were quick to recognise, it’s a real triumph – and, for other inventors, a dauntingly difficult act to follow.

     

    Nevertheless, characteristically undaunted, I’d like to present a new game of my own: NECNON. It, too, is a surround game (actually two games at once) for the hexhex board – less intricate than Blooms, but purer. Purer than Go itself, actually.

     

    You can study the full rules (which will take you several seconds) here, on my website, together with the outcomes of 100,000+ playtesting runs. Those tests were conducted on Stephen Taverner’s wonderful site mrraow.com, where you can try NECNON for yourself against the bot on a wide variety of boards. Just download as invited and then click on ai ai.jar (you’ll need up-to-date java, and may have to modify your permissions).

     

    Please note that NECNON also has its own page now on boardgamegeek.  Any comments and ratings there would be most welcome.

  • ypercube at 2018-12-17

    Thnx for the nice game.

    Nit picking 1: is there a site where we can play it, without the need of downloading executables and Java?

    Nit picking 2: why is the game tainted with the 1/2 point? Do game designers hate draws so much? ;)

  • Richard Moxham at 2018-12-17

    Hi ypercube. And thanks for responding. If you don’t mind, I’ll answer your two points in separate posts.

    As I’m sure you know, a kind of Catch-22 operates with regard to new games. On the one hand, it’s pretty difficult to get them implemented on gaming sites until they’ve proved their quality and popularity across a wide human-v-human player base; but then on the other, it’s hard (as you yourself have just demonstrated) to achieve that proof until they’re available on ‘proper’ sites.

    I’ve already contacted Richard Malaschitz here and Dave Dyer at boardspace about the possibility of an implementation, but am currently still awaiting replies from both of those gentlemen. A third possibility might be iggamecenter, which, though still owned by Arty Sandler, is currently curated by Luis Bolaños Mures. Luis himself is not keen on coding, but has intimated to me that, subject of course to arty’s permission, he would be very happy to see NECNON at iggc if there were someone else out there (Carroll? mmKALLL?) willing to undertake the enterprise. FWIW, I’m told that NECNON is not too challenging to code. If anyone at all interested would PM me to say so, I’d be very grateful.

    In the meantime, I’m hoping people might be prepared to override their perfectly understandable reservations and give the version at mrraow.com a try. Stephen really does have an astonishing array of abstracts available there.



     

  • Richard Moxham at 2018-12-17

    Hello again, ypercube.

     

    And yes, my time on the Abstract forum on BGG has taught me that a fair number of designers (and indeed players) really do hate draws that much.

     

    I myself don’t share their view. On the contrary, my oft-asserted position is that a small degree of drawability can actually enhance a game, and I have consistently rejected enquiries about ‘tie-breaking’ Morelli, whose incidence of draws (<5% at 9x9; <1% at 11x11; essentially 0% at 13x13) seems to me not merely acceptable but actively desirable.

     

    But it’s all about proportion. The first thing I playtested for in NECNON was drawishness, because intuition said it might be a bit high. And lo and behold, it turned out to be around 23% on the hexhex4 board, around 11% on hh5 and 8% on hh6, gradually diminishing from there on until it finally ducked below the 5% line on hh10. I considered these figures excessive, especially as I wanted the game to be genuinely playable (as otherwise it is) right down to hh4.

     

    So I introduced the half-point bonus for final stone, and from then on I playtested with the bonus in place (producing in that way all the stats displayed on my site). I found no clear pattern of the half-point being earned more regularly by one player (first or second, winner or loser)  than the other, but – interestingly – I did find that the gap between Black and Green win percentages appeared to be narrowed by its inclusion. And I think you will agree that the outcomes reported on are rather encouraging.

     

    So although I don’t have a definite view yet about whether the placing of the last stone is something that players can realistically incorporate into their controllable objectives, I do think that the mechanism does its job pretty well. A “taint” upon the game, as you call it…? Well, surely a very, very slight one, if so. 

  • ypercube at 2018-12-17

    Ah, many thanks for the replies. Yes, “tainted” was rather harsh.

    If the probability of draws is rather high (as much as 23% in size-4 and 5% on size-10, so much more than the numbers in Havannah on the same boards), yes, it makes sense to eliminate draws.

    I was thinking of Go when I said “tainted”, where the (ancient) designers and players felt the need to add half point to the komi, even though the occurrence of draws is tiny, even in small boards, like 9x9.

  • The_Burglar at 2018-12-17

    if a drawing game, can the slimmest lines be remembered by a human (usually a big avantage to stay in book longer)

    at reversi the white slim book requires 56 positions and the black slim book 475 positions

    to remember the whole of the draw book it’s maybe 80000 positions (gave up after 32k because Excel slows down)

  • Richard Moxham at 2018-12-17

    Hey Burglar.

    By “a drawing game”, do you mean one that’s a forced draw with perfect play on both sides?

    If so, I’ve no idea whether NECNON is or isn’t, nor who would be capable of memorising what, but 23% draws on hh4 (actually not an astronomic figure,  considering that you’re only talking 37 cells), still struck me as worth removing.  The trick for doing so, however, was to find a mechanism which felt a little more ‘part of the game’ than perhaps last-stone bonus did.

    So the discussion got me thinking, and I ended up with a bonus criterion I liked better, and which I announced this morning on BGG (here – mocko post timed 10.13 am ). See what you think  - and ypercube, you too. I’d be very interested to know.


  • Christian K at 2018-12-20

    This game sounds pretty fun actually :) I will try the ai ai implementation.

  • Carroll ★ at 2018-12-20

    This is a great idea !

    But it is another impossible game where you should be an oracle to settle an advantage early, and then in the endgame when you have played well, this is a chaotic beast that can tip over suddenly !

    Even the ai is sometimes surprised and changes his evaluation drastically before being able to compute the result.

    I was thinking that keeping a high mobility should give you an advantage, but I had a game where for long the ai had few moves but was happy with that and then I was forced to play opening moves and it came back to win... a very depressing experiment to avoid at this time of the year !

    Did anyone manage to win to the ai on a decent size, I certainly can not.

    Richard can you please post an update to the rules here when you have found the right way to mitigate draws ?

    I myself would prefer the bonus for the first to play an air-hole that the last move bonus. 

  • Richard Moxham at 2018-12-21

    Carroll, thanks very much for the comments and the (I think!) compliments.

    It certainly doesn’t  appear to be an easy game – I’ve never beaten the bot on any size at all (but then, I’m such a rubbish player that I rarely beat AIs at anything). The nice thing about being a game’s inventor, though, is the built-in consolation in defeat :)

    Anyway, I’ll certainly update the rules (and post a link to the update) once I find a corrective mechanism worthy of the game. At the moment, I’m nowhere near any decisive conclusions.  In the meantime, if anyone’s interested in a summary of my recent thoughts on the matter, you’ll find one on this page.

    Happy Christmas all round.

    R


Return to forum

Reply to this topic




Include game board: [game;id:123456] or [game;id:123456;move:20] or [game;id:123456;move:20;title:some text]