Games

[woodpijn]

Last night at GamesEvening we played A Fake Artist Goes to New York, which is like Spyfall but with drawings instead of conversation.

(Quick description of Spyfall for those who don't know: All players but one know a location where we supposedly are (e.g. supermarket). The remaining player is the spy. The players all ask each other questions. The

Comments

[simont]

One really major flaw is that the QM is on the FA's team

I always think in cases like this that you need a scoring system that acknowledges that there are three different parties involved, all with different goals - some players are trying to guess, some are trying to prevent guessing, and the QM is trying to set a problem of a difficulty grade that balances those sides well against each other, and each should be rewarded for how well they do the appropriate one of those things.

I don't recall which, but at least one of the games of inductive reasoning (Eleusis or Patterns II or some such), in which more than one player is trying to guess the same secret and scores more highly the faster they do so, rewards the QM-equivalent player by means of measuring the spread of the guessers' scores, so that the QM is incentivised to pick a secret that some people will get faster than others, and avoid either one that's so easy everyone gets it or so hard that nobody does

[cartesiandaemon]

I think the game where you model things out of plasticine, and also dixit have the "more points for setting a middling-difficulty question" mechanic. And I was always impressed at that as a solution[1].

Although, now I'm starting to think it usually actually works better in practice if the setter isn't really incentivised either way, just socially encouraged to think of something which will likely to be fun to guess. (I don't know if that would work in woodpign's case.)

[1] Although I did have to decide to avoid setting clues based on one person's particular knowledge: the system was easy to rig in dixit if you took it competitively, by finding a clue SOMEONE would get easily and most other people wouldn't know. And I decided it was more fun if I just assumed I was playing to an audience of roughly average equal knowledge.

[alextfish]

Yeah, my problem with Dixit is that it rewards cliquiness and excluding people. It's the opposite of a welcoming game: playing it in a group where some people know each other well and some don't is a recipe for the clique scoring high and the newbies / outsides scoring low.

And you're right, it does work better if you deliberately avoid doing things where you know person X will/won't get it, and try for something "moderately obscure".

Barbarossa (the plasticine-modelling game) works somewhat better in that it's purely on a relative scale, and also it's harder to make a plasticine model that one person will know but others won't!

[cartesiandaemon]

Yeah. I think I just happened to play Dixit (and codenames) with a group it worked fairly well for. I had a stranger time playing it with people who didn't know each other so well; no-one was trying to be too clever, but there were a lot of clues that just fell flat.

I was indeed just thinking that it's harder to model in plasticine an obscure reference :)

[gerald_duck]

While there are games which try to set up the scoring so that the setter is rewarded for things some, but not all, people guess, there are also games where the setter is trusted to act in a way that maximises fun. Penultima and Zendo work fine like that, for example.

When we tried Fake Artist last night, I set one of the phrases the fake artist didn't guess. I chose zebra crossing deliberately as something a bit more challenging, without regard to the fact that I was notionally trying to help the fake artist. I think that works slightly better

[woodpijn]

Yeah, maybe the QM should sit out of the scoring as well, and just serve the function of choosing a fun object for the sake of the other players.

Then again, that's what you pay game makers to provide. Otherwise you've just spent £20 on a small box of felt pens.

[gerald_duck]

But if the person who writes the cards doesn't choose the thing to be depicted, their role becomes 100% administrative.

I'm not sure how you fix that. Spyfall does it by printing N cards for each location, but it feels like Fake Artist needs many more things than Spyfall needs locations.

Probably, the solution is to use an app instead of physical cards: create a new game identifier, everyone logs into that game, people suggest some things then, once there everyone's submitted enough, it chooses someone to be the fake artist then picks a thing to tell everyone else, which is not something the fake artist submitted.

[ghoti]

I've always found that the difficult part of early games is counting on so I tend to avoid early games that involve that element. Snakes & Ladders type boards are the worst, because children want to go to the written number that they rolled, not count on that number, but the 'start at zero' just takes the fun out ime. THat's not just my children, I've spend a sizable amount of my life working with childrne in the 2-6 age range, and it's been pretty universal.
I haven't observed the path finding thing, but it doesn't greatly surprise me either.

That said, the game sounds cool, and Andreas is coping well with the 'follow the track' games, so we'll give it a go.

[gerald_duck]

Tangentially to games: when I was young, they left teaching zero surprisingly late. Nowadays, I think I'd give serious consideration to teaching it along with other natural numbers very early. Then you could count zero as where you are, one as the next space, etc.

[woodpijn]

That is indeed what Zoe is now doing to help her figure it out :) She learned the digits mostly from Numberjacks, which has characters shaped like all the digits including zero.