From small beginnings
Here is a new idea -- possibly it's a completely original idea, and how often do those come along? So bear with us.
The game is called Gamete. In its basic version, Gamete 1, the rules are these:
- There are two players.
- Each player has 5 points to buy characters with.
- There are three types of character: Wizard, Knight and Dragon. Each costs 1 point.
- The players buy five characters of their choice. These are not revealed to the other player yet.
- Each round, players choose one of their characters to fight. The choices are then revealed simultaneously.
- Wizard beats Knight beats Dragon beats Wizard.
- A player scores 1 for a win, 0 for a lose or draw.
- Repeat until all characters have been used up.
- The winner is the player with the higher score.
"So what?" you're thinking. "That's just rock paper scissors." But hold your horses. That's not the original idea. That's just the taking-off point. The point of Gamete is that it evolves. And here's where you come in.
You can make any variant you like on the Gamete rules. For example: "Players start with 6 points instead of 5." (I know, boring variant, you can think of much better rules. Good, because we'll be asking you to do exactly that.)
Variants will be maintained on the Gamete page for now. (We'll set up a GitLab platform later.) Variants based on the original Gamete 1 rules will be labelled Gamete 1-1, 1-2, 1-3 and so on. Variants that spin off those variants will develop in a branch structure. So the first variant deriving from 1-11, say, will be 1-11-1 and so on.
You're seeing it now, right? This might start off looking like rock paper scissors, but it could develop into anything. Gamete 1-5-15-6 could be a fifty-page roleplaying game or a strategy boardgame for unlimited players. It can and will go anywhere -- guided by you.
Over to you to get creative. We don't know where this will go, or what variants the gestalt will come up with. The core rules are a game universe at the moment of the Big Bang. Now you get to play God... Fiat ludus!
Here's a more interesting root-level variant: you could have four character types instead of three. Wizards beat Knights and Dragons, Knights beat Dragons, Dragons beat Elves, Elves beat Wizards and Knights.
ReplyDeleteOK, so Elves and Wizards are more useful. They beat two opponent types each; it's no longer symmetrical. How would we deal with that? One variant (which might be Gamete 1-1) is to still charge 1 buying point per character and let the players gamble. If you pick only Elves, figuring to play the odds, you're sunk if your opponent chose Dragons.
But there's another subvariant (call it Gamete 1-1-1) in which you have to pay 2 points to buy an Elf or a Wizard. But won't that mean you run out of characters? Well, one option is to let you go into negative points. So if I start with an Elf, a Wizard, and three Knights that means I go into the game with -2 points. Those are deducted from my win scores, so I have to win by a margin of more than 2 to come out ahead.
I like 1-1-1, I would propose 1-1-1-1 you actually get ten points at the start to buy characters, and points left over count toward your score, but you get two points for a victory (otherwise a player could buy no characters and always guarantee a draw)
DeleteYou may buy as many or few characters as you like, and can choose to enter any round or not
Would the other player know how many characters you bought? Actually, there are two possible variants there rightaway.
DeleteWell, here's a rules question (of course :-)) How much of the game are we allowed to change between variants?
ReplyDeleteWe had nine clauses in the root version, your variant 1-1 can be incorporated by changing clauses 1 and 6, which seems reasonable. Are we to propose that you may amend, add or remove up to two clauses from the parent variant to form a child variant? Part of the challenge will be the ingenuity required to squeeze any changes into the two clause rule
Are these meta-rules themselves subject to change? and how would we track that?
On a separate subject, does anyone know a handy way we can keep track of these variants? Dave and I actually created a GitLab account already, but it's a bit of a fight to autogenerate all of these rapidly diverging versions (Git in general allows this, but really aims to merge back to form a single version most of the time)
By the way, since Dave has already taken v 1-1, I propose v 1-2 (that is a new variant from the root)
Clause 4 is amended to say:
"The players buy five characters of their choice. These are revealed to the other player who may make a note of them"
As the players use up characters the remaining combinations become more known, making (hopefully) for an interesting ending (whereas in v 1 the result of the entire game is effectively random, isn't it?)
Of course this could also be applied to Dave's v 1-1 to make v 1-1-2
Also perhaps someone can figure out how to make the game multiplayer, whilst sticking to the (admittedly so far unilaterally declared) two clause meta-rule?
I was wondering about this. I suppose it should relate to the number of clauses changed, but that would make your variant closer to the root game than mine (because I changed two clauses) which we don't currently have a way to reflect in the nomenclature...
DeleteThis reminded me of a similar problem, namely classifying the many variants of SARS-CoV-2. Nomenclature there uses systems like Pango (hierarchical, like our proposed numbering system) or Nextclade (measures relatedness with a clade). We'll also need to rationalize the original expression of the rules, which is different on the blog post above from the version on the Gamete page. (https://www.cdc.gov/coronavirus/2019-ncov/variants/variant-classifications.html )
DeleteSo, we already have have the makings of a mini-campaign game, each player buys an army from a pool of points, and moves secret portions of that army to occupy locations
ReplyDeleteWhen opposing armies occupy the same location they fight using the rules above, instead of victory points, defeated units are eliminated. Each round a player may play any of his remaining cards, perhaps we say he can't play the same card twice in succession unless it is his only card. An army may cut and run at any stage by discarding one card and retreating. Locations have a recruitment value, you may buy cards each turn to the value of your total recruitment value
Question.
ReplyDeleteYou are so centered on natural numbers (small ones to boot),
but why you omit possibilities of floating point numbers? ;-)
As to idea itself -- to make upgrade of rules in small steps -- it's obviously great, but...
it's better to be automated -- let computer add some rules and test how they'd work.
We might take longer training the universal gaming model than we would actually designing any games! Though it would be fun to do. DeepMind could add generative game design to MuZero, which already knows about a whole bunch of strategy games.
DeleteWell... my musing here was -- about trying to use more sophisticated model from the very beginning.
DeleteAnd. In my opinion. To make it with floating-point numbers -- could be greatly easier.
As it would allow small tweaks, naturally.