Advanced Pass the Parcel, the prototype

What is this?

I’m developing a party game called Advanced Pass the Parcel. These are my thoughts on the first playtests, and where I’m thinking of taking it next.

It’s mostly for my own benefit, but other people might be interested. Maybe that’s you?

The parcel… and the cards


In Season 3 Episode 14 of excellent Australian animation Bluey (a fun show for kids but but even better for adults), the children play a game of Pass the Parcel according to the standard modern rules:

  • Every layer contains a small prize
  • The adult operating the music carefully ensures each child gets a prize (a secret rule that the players don’t know)

Lucky’s dad objects to this and runs the more traditional version:

  • There is only a single prize at the very end – but a much bigger one
  • The music stops at random

This has some dramatic consequences, but it also got me thinking about changing the rules for Pass the Parcel in other ways. I liked the idea of running a really weird version of the game with kids that had all sorts of chaotic shenanigans deployed layer by layer – swapping places, surprise rubber spiders, minigames, dance breaks, ability to control the music, general mayhem.

Unfortunately (or perhaps fortunately?) I don’t currently have access to a kids party where I would be able to do that. Fortunately, Clare pointed out that the general idea could work just as well with adults, and I realised she was completely correct!

Basic design

As a game, Pass the Parcel in both the above forms is notable for the complete absence of player agency! According to most definitions, this means it doesn’t even qualify as a game. That’s fine for the kids version, but I don’t think that’s something adults could as easily accept.

Immediately I thought of the most obvious mechanic to fix it: you can play a card after the parcel stops (an ‘Interrupt’) to move it further around the circle. If you got to open a layer it would naturally reward you with more cards. This certainly gives players agency and just might admit a bit of strategy.

There were a lot of other rules and changes I wanted to try, which gave me the second core idea: each layer also reveals a card with some sort of new rule! I love games that do this, but there aren’t many, and it brings some of that feeling of chaos / unpredictability I was originally imagining.

Interrupt design

As a quick way to test the idea I planned to use playing cards as the interrupts. Large numbers would be a bit boring, and I felt there had to be some restriction on what you could play to add some strategy and stop players from just playing all their cards in a single round. I settled on using only Aces (for 1) up to 5 as a range, but I couldn’t decide if an interrupt should be lower than the one played before it (so the parcel would sort-of decelerate), or be +1/-1 from the previous interrupt (for less predictability). I decided to try both by starting with one rule and changing it later on with one of the ‘new rule’ cards several layers deep into the parcel.

Imagined example play

To be very clear about it, this meant a game might play out as follows:

  • Music starts and the parcel is passed around
  • The music stops, so the parcel stops in one player’s hands
  • Anyone can play an interrupt (they have five seconds to decide to do so)
  • If they do (for example, by playing a 4), the parcel moves that many places around the ring – probably to end up at the person who played the card
  • Anyone can then play another interrupt if it is less than 4 (or under the +1/-1 rules, if it was a 3 or a 5)
  • This keeps going with more interrupts, until the parcel stops and after 5 seconds no more interrupts are played
  • The person who ends up with the parcel opens up a layer, gains more interrupt cards, reads out the new rule, and that round is now over
  • Music restarts to begin another round, until eventually someone opens the final layer and wins

Layer design

Gaining one interrupt card for opening a layer would create a net drain on cards in play; gaining three cards would create a very strong rich-get-richer feedback loop, since someone gaining so many cards would very likely get the parcel again. So it kind-of had to be two interrupt cards you gain in each layer.

Physically, the tradition of wrapping paper and tape seemed quite wasteful; fortunately we had a lot of furoshiki cloths of various sizes that were lightweight, quick to wrap and unwrap, and easy to re-use.

I realised there would be a bit of strategy around judging when the final layer was reached, so I planned to have some fun with that. First I had one layer reveal two parcels inside, the new rule being that the parcels moved simultaneously in opposite directions. One of those then ended before the other (with a ‘jackpot’ of 4 interrupt cards), and the remaining parcel was instead a large envelope – which would look quite final after all that furoshiki. However, in that envelope was a hankerchief wrapped around a smaller envelope, to create two further final layers!

Parcel Passing Protocol

A classic problem in the original game is that the music can stop mid-pass. Two children both have their hands on the parcel and will tend to fight over it.

With adults I assumed we could instead go with a much less ambiguous approach: the parcel could be lightly thrown from one person to the next!

This was a mistake.

New rules

Here are a few examples of the rules I tried adding to the different layers:

  • Parcel now reverses direction
  • All players draw 1 new interrupt card
  • Give these interrupt cards to the players either side of you
  • Read this out loud: “The next layer has a spider in it. Not including me, whoever is first to volunteer gets to open it.”

I also tried some rule cards that instead you got to keep and play at any time:

  • Shout ‘Objection!’ to cancel an interrupt, then discard this
  • Declare ‘Technically it’s my birthday’, the player with the most cards must give you two, then discard this card.

The 5 seconds to play an interrupt was quite long, but felt necessary when people were new to the game. So I also add a rule card that reduced this to 3 seconds a few layers deep.

What about the prize?

As you might have noticed, games played by adults routinely do not have any prizes at all. People play Monopoly, Catan, Hearts or Waldshcattenspiel purely for the honour of victory. But having the final layer of a parcel contain nothing other than perhaps a bit of paper saying ‘you win!’ felt far too anticlimactic.

Given the direction the game was headed, the answer seemed obvious: the final layer has a blank card, on which the winner gets to write a new rule for the next game.

Expected behaviour

Under these rules, I expected the following:

  • Players would generally play an interrupt only when it brought the parcel to them
  • Players would hoard cards a little bit to ensure they had interrupts to play when some people had run out, and when the crucial final layer approaches
  • Perhaps emergently some players would co-operate. Player A might play an interrupt so that player B got to open the parcel; player B might later return the favour (with the help of the additional cards) so that player A might get to open a layer in turn.

To make the prize a bit more exciting, and with the idea of co-operation in mind, I changed the final prize to be two blank cards – one for the winner to write, and a second they could choose to give to another player.

Playtest 1

I was very lucky to be able to test the game out three times in relatively quick succession with three different groups of people.

Here’s what I learned from the first group, 17 people who work in video game development (so were more game-literate than the average person):

  • Players were excited by the simple joy of passing the parcel and playing interrupts! It may have helped that some players were lightly inebriated…
  • Interrupts having to be lower than the last one played in a round immediately felt far too restrictive, and we quickly switched to a ‘same or lower’ rule. This led to a much more exciting round finish when an Ace could be trumped by another Ace (and possibly another, and another…)
  • We needed a way to adjudicate bad behaviour (holding on to the parcel instead of passing it, throwing it badly, deliberately not catching it and picking it up slowly and so on). I quickly opted for player’s voting which player misbehaved, and penalising that player by having them discard one interrupt card
  • Players were not very strategic with their interrupts, playing them early and often – even when it brought no immediate advantage to themselves. Probably inebriation played a part here too, but it did feel like doing a thing that triggers everyone else to do a thing (pass the parcel the number of times you indicated) is inherently quite fun?
  • Rather than co-operation, players instead formed enmities! They would play interrupts to deny certain players the parcel – players who previously denied them the parcel, or players who were doing too well, or just players where it seemed funny to deny them the parcel for some reason
  • The interrupt that reads “The next layer contains a spider. First player to volunteer gets to open it” is very silly and out of keeping with the rest, but still proved very fun, suggesting a whole other possible way the game could lean. (Also note, it is a trick – there is no spider of any kind)
  • In a large group and a loud party setting, most players can’t/won’t read the new rules loud enough, so a moderator (me) had to do that for then
  • Complicated new rules just do not work in a loud party setting
  • We later needed a way to punish bad behaviour even when a player had no cards left to discard. We concluded they would be excluded from the next round
  • The rich-get-richer problem was actually pretty bad, with one player getting so many cards they would just play them in bulk (e.g. four 3’s for a total interrupt of 12)
  • The ‘+1/-1’ interrupt rule led to extremely long combos that drained almost all cards from the group, suggesting a lack of strategy, or possibly a failure to guess how many more layers were left?
  • The drama of approaching what may or may not be the final layer was tremendous!
  • With 17 people starting with 3 cards each, and 10 layers of parcel with about 2 cards in each, the game went on for about 30 minutes – probably too long
  • The winners wrote their new rules, and while these were mostly good, there was one that was possibly a bit broken and needed a second draft. Not sure how to deal with that…

In general despite some poor throwing discipline and long running-time, it seemed to go very well. One player afterwards reported “I haven’t laughed that much in ages” – a pretty good sign!

Playtest 2

In the next playtest, I once again found about 17 people willing to play, and once again some of them were slightly tipsy.

As a warm-up I had the group simply pass the parcel in a single circuit, and the relative ease with which they managed (or rather did not manage) to do that convinced me the throwing method was not a good plan. Instead we went with regular passing of the parcel from one person to the next, with the rule being that if the parcel stopped when two players had contact with it, whoever was furthest around the circle (i.e. was just in the process of receiving the parcel) should get it, which proved suitably unambiguous.

I made sure the new rules in each layer were simpler. I also added more new-rule cards that would ensure cards were distributed or replenished to cut down on the rich-get-richer problem.

Finally, rather than using whatever party music was playing, I specifically chose Meute’s track “Rushing Back” because it had a nice beat for passing to (136 BPM, pass every other beat so 68 BPM), followed by Edvard Grieg’s “Hall of the Mountain King” (inspired by a video game it would kind of be a spoiler to name).

Here’s what I learned this time around, with a group that seemed to have a wider range of game literacy:

  • Once again the simply joy of passing and interrupting was much greater than I expected! Interrupts just seem incredibly dramatic
  • This group quickly cottoned on to co-operation as a strategy, enmity formation was much lower. Depends on the group I guess?
  • Given the absence of any agency during the passing phase I tend to assume the gaps between stops should be quite short. But based on how the mood was feeling and perhaps how much people enjoyed the music, it seemed clear that longer periods (with the parcel making 3 or more circuits) actually felt just fine?
  • Changing the core rule of which interrupt can follow which added more confusion than I think is fun for a few people. For a group that includes non-gamers and/or people who had a few drinks, it’s probably best to keep that rule consistent
  • The accelerando of Hall of the Mountain King worked just as well as I hoped, and I was able to time the whole game such that the final stop aligned exactly with the end of the song. So good!!
  • Once again the game lasted about half an hour, which actually felt fine in this case, perhaps because of the more suitable music
  • The nature of the occasion meant no further games would be played, and the appetite to write new rules was not very strong

This now gave me a lot of excitement about the general concept, since two quite different groups of people seemed to have a lot of fun with it.

Playtest 3

I updated the rules again and printed out large-font new rule cards for better legibility than the hand-written ones I started with. Then I tested it out with 14 friends who are much more experienced with my kind of games, and indeed games of all sorts. I repeated the music choices from Playtest 2 since they seemed to work so well. Here’s how that went:

  • The first few rounds were much calmer, with fewer interrupts played, and in particular much fewer ‘frivolous’ interrupts (those that don’t advantage the player)
  • The vibe was initially more serious, and the relatively low amount of player input into the whole process started to feel a bit odd…
  • … but as some new rules kicked in and a few interrupts surprised people (in particular a father playing an interrupt denying his own child the parcel!), I sensed the vibe picking up as people got into the rhythm of it all
  • The climactic ending of Hall of the Mountain King was once again great, although the winner then immediately began to unwrap the final layer, and I didn’t even realise that was wrong – such is the finality of the song. Players pointed out the error and we quickly played out the ending properly (allowing interrupts to take place).
  • Richard B felt that playing an Ace to take the parcel from the player next to you was such a distinct move it should have its own name; he called it ‘batsy’. I like it!
  • Once again there wasn’t a huge appetite to write new rules, although unsurprisingly a lot of this group were very interested in discussing what they could be

Tarim raised that the ‘same or lower’ rule meant that the higher numbers got very little play indeed. This was more obvious in this playtest with the more tactical play. Players had ‘solved’ this problem in the first two playtests by playing their highest cards as soon as possible and without regard for personal benefit!

My concern about the alternative ‘+1/-1’ rule is that it can encourage very long runs of interrupts draining ‘too many’ cards. That said, it does still have a nice slowing effect as players reach the lowest (or highest) numbers available, since only a single number can interrupt from there…

Where next?

In general the game seemed a lot more fun that I would have guessed, so seems worth developing further. I see two obvious but very different paths to try.

Path 1: Fix what seems broken: less luck, more agency, more strategy

Even with a more strategic group playing, it became clear that the game was barely more strategic, at least in outcome, than just stopping the parcel randomly and opening it. But in a game with many players, where only one will get to open a layer, how much strategy is even possible? Most of it will surely cancel out.

The ‘rich get richer’ problem is also quite annoyingly baked into the concept. Even though ensuring every other layer had a rule that to some extent redistributed cards or allowed new ones to be drawn, that dynamic does feel a bit antithetical to strategy.

My working theory for a solution to both problems is as follows:

  • Each card has an interrupt value AND a special effect, players choose which one to play
  • Players can only play one card per round. Having more cards gives you more choice, but not as much overwhelming power
  • When you open a layer, it contains three cards – choose one for yourself, give one each to the players either side of you
  • Special card effects could follow a ‘tableau’ style of play – you place them in front of you to grant some sort of mild buff that stays with you for the rest of the game. In this way, even though you don’t have much control over a single round, you get to be strategic over the course of the game. Even if you don’t win, it would probably feel better than just playing an interrupt to gain the parcel and then immediately have that play ‘wasted’ when someone plays another interrupt.

Possible example card effects:

  • Interrupt: Play this to reverse the direction of the last interrupt
  • Interrupt: If 3 interrupts have been played this round but you have not touched the parcel since the music stopped, draw 2 cards
  • Tableau: You can add 1 to any interrupt, but must discard a card whenever you unwrap a layer
  • Tableau: When a +2 interrupt is played, once per round you can draw 2 cards and discard 2
  • Tableau: Whenever a player’s hand reaches 6 cards, they must give you 1 and another player of their choice 1

The dual-function of cards (interrupt or effect) mitigates the weakness of high numbers under a ‘interrupts must be lower than the last’ paradigm. But I also like the idea of including a small number of ‘Joker’ interrupts which can be played after an Ace and pass the parcel directly to you – but can be followed by an interrupt of any value.

That all feels quite promising to me, but leaves a lot of unanswered implementation questions, and also just would not work with a large group.

Path 2: Build on what works! Frivolity, group activity, spontaneous alliances / enmities

Leaning in to the fact that apparently just playing cards and having the group do things in response seems fun, perhaps don’t worry about strategy and just have fun with it?! This leans in a direction I’m actually very interested in – pushing away from games and into something I’m currently calling ‘co-ordinated activity’, which I think is a vastly underexplored space.

Perhaps keep the ‘1 card per round’ restriction to stop the game running on too long and avoid the rich-get-richer problem a little bit?

In general have rules that are a lot more varied and surprising, or give the players a different kind of agency. Some example rule cards you might unwrap alongside your interrupts:

  • Ask a quiz question. First player to get it right (excluding you) draws 2 cards
  • Pass the parcel, each player adds one word to the story as they pass it. Parcel stops at the end of a sentence or when it completed a circuit. You choose who to award a bonus card based on their contribution.
  • Everyone High-5 one player next to you. Successful high-5s swap places.

Maybe each of these is great for very different groups – or maybe the right answer really is some kind of strange combination of the two? The experiments continue…

  • Transmission ends

Tim now posts on Bluesky as and has previously written about making games about sandwich making and blindfold roleplaying


Weird Family Fortunes


Family Fortunes (Family Feud in the US) is a brilliantly subversive quiz concept.

They set things up by conducting a large survey asking people all sorts of questions that have multiple answers. On the show, teams (families) then compete to guess the most common answers to those questions. So unlike a conventional quiz where you are rewarded for knowing obscure things, instead you are rewarded for being as similar as possible to the average person! Or at least being good at guessing what an average person would think.

I ran a ‘Weird Family Fortunes’ survey/quiz that compressed this idea into a single step. Players completed a survey of Family Fortunes style questions – but needed to anticipate what other people’s answers might be as they did so. Initially this was quite easy, and then it started to get a bit weird.

There were 3 sections, with 8 questions in each, and 38 people sent in responses – if you were one of them, thanks for that! Here are the results.

Section A: Majority Wins

In this section the rule was similar to Family Fortunes: ‘majority wins’. Players get a point if their answer matches the most commonly given response out of everyone playing.

For example, if 10 people answered ‘Pinball’ and 7 people answered ‘Snooker’, the 10 people who chose Pinball would each get a point, since theirs was the most commonly given response.

I grouped together responses that I consider equivalent, e.g. I would combine ‘Football’ with ‘Soccer’ when considering the totals. This sounds superficially simple, but gets into some quite difficult judgement calls later on, as you’ll see.

A1: What is your favourite colour?

Inspired a little bit by Monty Python and the Holy Grail, this seemed like an easy place to start. Perhaps informed by the first response in that film (but also just possibly what is most popular in general), the winner by quite a long way was ‘blue’.

Some of the responses suggested to me that people either had not realised the goal was to guess what the majority of all players would guess (‘Pastel Pink’ and ‘Aquamarine’ were two ambitious responses), or simply did not care.

I also think the kind of people I sent this to are generally very smart and original, and I often run games that allow them to flex those strengths. In this round at least, this is pretty much the opposite of what you should do, so perhaps it was just force of habit that led to the particularly unlikely response of ‘radio wave’.

A2: What is the best animated Disney film?

Given the demographic of respondents, perhaps ‘The Lion King’ was the inevitable winner. But I was very interested that ‘Toy Story’ came second, given that it was – at the time – a Pixar film, not a Disney one. Years after Toy Story came out, Disney bought Pixar, and the argument could be made that it has now become a Disney film. In a traditional quiz that would probably be debateable, but here it doesn’t matter! If everyone makes the same mistake, it becomes valid. Albeit second place in this case.

A3: Which is the least awful social network?

This one is a bit of a popularity contest, since I think on average most people only use a small number of social networks. Still I was surprised to see Instagram take a convincing lead. Good to know!

A4: How do you distinguish the file name of the final version of a document?

As a subject of personal interest, I find that the version of a document one thinks is final turns out not to be final in the vast majority of cases. This creates the problem of distinguishing the document you first thought was final from the one that is actually final (if there ever even is such a thing).

The results here suggest that other people don’t have that problem, or other people don’t think other other people have that problem, or they don’t care, or think other people don’t care. In other words, these results don’t actually tell us very much.

Still I was disappointed that only 3 of the responses reflected my experience – there is no final so name accordingly – and actually one of them was me, since I was the first respondent to the quiz!

A5: What is your favourite video game?

The big problem here was categorisation. ‘Zelda’ and ‘Mario’ do not uniquely refer to a specific game, and it seems too much of a generalisation to group all games in the Zelda  or Mario series as a single category. Then there is ‘Mario Kart’, which is also a series with many iterations – but ‘Mario Kart’ was what ‘Super Mario Kart’ was generally called (just as Star Wars: Episode 1 – The Phantom Menace was generally known as Episode 1), so probably stands. As noted, in general this quiz isn’t about me judging what is and is not a valid answer, but when it comes to joining up responses I did have to make that call.

There was a surprising breadth of answers here, but perhaps this is more a sign of how huge and varied the topic of games is… and is a sign that a lot of the respondents also work in a video game studio.

A6: Who is the best female character in the Marvel Cinematic Universe?

This question got at the problem of general knowledge (the average casual or non-MCU fan probably could only name Black Widow) vs. strong contenders from elsewhere if someone is more familiar with the material. Of course, this is compounded by people guessing what others will guess, making Black Widow a very likely winner.

A7: Who is the best droid in Star Wars?

Not only is R2-D2 clearly an excellent droid, they will be the obvious choice for the large number of people who haven’t seen the sequel trilogy or other Star Wars spin-off material. In a straightforward poll, I hope a few more might support the excellent L3-37 or BB-8, but when it comes to guessing which will win, R2-D2 is hard to resist.

‘Han Solo’ might look like a rookie error, but makes sense if you consider some additional context. To say more would be to spoil that other context.

A8: Which year was the best?

Another tacit test of the player’s demographics, 1999 and 2000 are the years that most of my peers were in their 1st or 2nd year of university – with maximum freedom and probably the lowest amount of responsibility. The internet was also gaining traction and was exciting and new. Well, I’m guessing that’s what people thought anyway.

Section B: Runner-up wins

For this section, the rules changed: runner-up wins. That means players got 1 point for each answer where their response was the second most commonly given. Of course, players needed to bear in mind that everyone else is trying to guess which will be the second most commonly given as well.

For example, if the question was “What is the best letter?” and 10 people answer ‘a’, 7 people answer ‘b’, and 2 people answer ‘c’, the people that answered ‘b’ will get a point since that was the second most commonly given response.

I originally considered giving people the chance to answer each question twice: once with their ‘honest’ answer, and then again where they are trying to guess what other people guessed. But I liked the elegance of compressing that all into one question, even though it makes it very difficult and probably quite luck-based!

This was made even more difficult since it was hard to tell how many respondents would truly understand what was being asked, as that could influence the responses. Let’s see what happened.

B1: What is your favourite mode of transport? (runner-up wins)

I made the call to separate ‘bike’ from ‘ebike’ and ‘quad bike’, and also ‘tube’ from ‘train’.

Now I think about it, I can actually see that helicopter and zeppelin are pretty excellent forms of transport, but it’s not too surprising that not many people thought of these. Or thought that other people might think of it not quite as much as the most popular thing.

B2: Which tree is the best? (runner-up wins)

At this point I slightly regret not asking the ‘majority wins’ version as well as the ‘runner-up wins’, since I’m curious how similar they would end up being. Oak seems an obvious front-runner, but well done to the 3 people who all agreed on Birch making it the runner-up.

B3: What is your favourite potato-based food? (runner-up wins)

Somewhat contentiously I separated ‘chips’ from ‘fries’, as in my experience they are very different forms of potato, and I have been to establishments offering both as separate options on the menu. In this case I’m very uncertain if the people who won by guessing ‘chips’ were under-thinking it or thinking-it just the right amount.

B4: Pick a word that begins with ‘w’ (runner-up wins)

This could have been very hard, but I deliberately included two possible words in the question itself, creating what I thought might have been a dilemma, as surely one of those would be the runner-up. Sure enough, these showed up close to the top – but out of nowhere, water was the most popular, and I’m very unsure why.

B5: You are invited to a thing you don’t want to do, but technically you are available. What do you do? (runner-up wins)

I’m very interested to know the ‘majority wins’ answer to this question, but as a person who frequently invites people to things, perhaps people would not have answered honestly in any case?

This created some tricky grouping problems, in particular I decided to separate the general (‘make an excuse’) from the specific (‘feign illness’), and nuances of timing (‘decline’ vs. ‘cancel later’; ‘go anyway’ vs ‘go for a bit and see’). I am mysteriously pleased that the successful runner-up was the perhaps surprising ‘go anyway’.

B6: What did Tim eat just before writing this question? (runner-up wins)

For those that want to know, the reality was that I ate a banana just before writing this question. But this is not about accuracy!

Given that you most likely needed to at least come up with something that a few other people would guess, I was surprised at the long tail of very specific responses – I had thought generic answers like ‘a snack’ or ‘breakfast’ would have been the way to go. But in the absence of examples, I can imagine it was not clear that general categories might be a good way to go. It was quite pleasing to end up with a 6-way tie between 12 people!

B7: Where is alien life most likely to be found in our solar system? (runner-up wins)

In this particular case it felt fairest to group up specific responses (e.g. locations on Earth) to a single major planet, but that did end up making Earth the most popular. Perhaps if I do something like this again it would be better to specify (where possible) what kind of grouping I will do?

B8: If you had to kill a vampire but you aren’t sure which vampire rules are in play, how would you first try to do it? (runner-up wins)

With an unlikely tie for first between sunlight and a stake, the successful runner-up was the garlic grouping. My personal favourite was the 3 people who chose not to kill at all, in a category joining ‘Don’t’, with ‘run away’ and ‘fall in love with them’.

Section C: Weird

This section grouped the weirdest majority and runner-up questions with a few that had rules all of their own…

C1: What is the lowest unique whole number that someone will guess in answer to this question? (For example, if two people guess ‘1’, one person guesses ‘2’, and one person guesses ‘3’ then the person who guessed ‘2’ will get the point, as that was the lowest unique response).

I thought ‘Whole number’ was a well-defined term, excluding negative numbers, but I failed to specify this. The fact that several very intelligent people opted to give negative responses made me realise I should have specified that restriction up front (and not just implied it by the example).

Under the intended rules – which is what I’m counting here – the winner was the person who selected ‘2’, funnily enough the winner in the example given! This meant 1 or even 0 could have won.

If we broadened the definition to negative numbers, I find it pleasing that two people both opted for the largest negative number possible in the restrictions of the format, meaning the winner (under those rules, which we are not actually using) was instead the person opted for a merely very large negative number.

C2: What came before the Big Bang? (Majority wins)

In grouping terms, I considered the winner, ‘Nothing’ to be distinct from ‘there is no “before” ’, since the latter implies time itself did not exist / was not meaningfully defined. I separated out ‘The Big Crunch’ from ‘A Big Crunch’, because while superficially similar I felt like the use of “A” much more heavily implied a cyclic behaviour.

Respect to the one person with the very reasonable response of ‘I don’t know’!

C3: What is the opposite of the thing most people will answer to this question? (Any answer that can be interpreted as opposite to the answer most commonly given to this question gets a point)

I asked this without any idea how people might answer, or how possible it would be for me to judge. As it turns out, variations on ‘nothing’ clearly took the majority, making both ‘everything’ and ‘something’ both winners as opposites.

You could argue that the response that simply repeated the question (and perhaps the one that said ‘asking the same question again’) should win, in the sense that they are the opposite of all of the others which are all ‘answers’. But I don’t think a response should dictate how I group things up, so that does not win, but is still a good effort.

I like that one person gave 42 and another -42. I also particularly like the surreal ‘you should put salt on it’, which is possibly trying to do something clever I haven’t been able to figure out.

C4: What is the average of the numbers submitted in answer to this question? Give your answer to the nearest whole number. (Closest to average wins)

With just one person opting for a very large negative number and 4 people going for very large positive numbers, the largest positive number ends up winning!

C5: Pick a whole number from 1 to 8. (Runner up wins)

I wondered if the earlier questions might somehow prime people’s responses here, but I can’t see a clear pattern. Perhaps the similarity of ‘runner-up’ with ‘2’ as a concept made it the most popular choice, allowing 5 to take the true runner-up prize.

C6: What would be a good question to ask in this survey? (Best 3 answers as judged by Tim will win, originality will be highly rated)

After struggling to come up with questions, I figured I should turn the problem back on respondents. I got a very widely varying set of responses, which were as follows:

  • How do you feel? (Majority wins)
  • What would be a good question to ask in this survey? (Best 3 answers as judged by Tim will win, originality will be highly rated)
  • Did anyone else’s brain start hurting around C1?
  • Pick any whole number between 1 and 50 to stand on and a second such number to place a trap on. You score a point if nobody has placed a trap on your number.
  • If you had to be on an island with only one type of bird, which would it be and why?
  • What do you consider to be the greatest virtue?
  • Number of peanuts in this pack of peanuts. Points at peanut pack.
  • Thumbs being incredibly useful, why do we only have 2?
  • How many bees would we need to attach to Tim to be able to fly him to the moon?
  • Why doesn’t glue stick to the inside of the bottle?
  • Why did the chicken cross the road? (runner up wins)
  • If Tim was a phone, which phone would he be? Bonus points, what colour/case would he be?
  • If you wanted to get a Guinness world record, what would be your best shot?
  • If you could choose someone to narrate your life, who would you want to serve as the voiceover?
  • What do you get when you cross a rhetorical question and a paltry attempt at being funny in a quiz answer?
  • What is Tim’s favourite colour?
  • If asked to define yellow, how many people would not use the opposite of a not a banana to explain it? (Wooden spoon wins)
  • If everyone were to answer the best year of their life minus the worst year of their life, what do you think the average of all answers would be?
  • How much wood would a woodchuck chuck if a woodchuck could chuck wood?
  • What colour are the clouds in the sky right now?
  • How many bison dollars is to a single US dollar?
  • Best burger chain
  • What’s your favourite question? (Majority wins)
  • Favorite question in this survey, runner up wins
  • Why does Tim write question that we don’t understand and have no relevant or real answer
  • What is the best number?
  • favourite integer (runner up)
  • what’s your favourite thing about Tim?
  • You have a pet dinosaur and you don’t want to give it neither a human, a dog/cat nor a scientific name. What would you name it? (Majority wins)
  • Name something you find in a bakery that can also be used as a term of endearment
  • Pick a number. If the total of everyone’s numbers is odd, evens win. If the total of everyone’s numbers is even, odds win.
  • Do you think that this format of quiz will be popular enough to repeat (wholly or in part) for next time?7
  • Assuming the location of each person responding to this survey forms the vertex of a polygon, how many KitKats would it take to trace the perimeter? (One point for answers within one standard deviation of the mean)
  • What would be a good question to ask in this survey?
  • Who, What, Why?
  • How many distinct named colours do you think there are?
  • Probably something like what you just asked, so that you could effectively outsource creating the next survey for your next event.

I said originality would be highly rated, but another (but less important) criteria was how well a question might actually work in this context – something I now have a much better feel for. My top 3 were:

  • Pick any whole number between 1 and 50 to stand on and a second such number to place a trap on. You score a point if nobody has placed a trap on your number.
  • If everyone were to answer the best year of their life minus the worst year of their life, what do you think the average of all answers would be?
  • Assuming the location of each person responding to this survey forms the vertex of a polygon, how many KitKats would it take to trace the perimeter? (One point for answers within one standard deviation of the mean)

Also my respect and bafflement to the complicated question about yellow that then reveals that ‘wooden spoon wins’, I am very curious what the responses would have been to that!

C7: Rock, paper, or scissors? (Those who select the option that beats the majority win)

My favourite question! This ended up being a little tricky to resolve, with a two-way tie between Scissors and Paper.

Ultimately though the answer is quite clear: the majority is a tie between Scissors and Paper, so we simply consider the following:

  • Scissors ties with Scissors and beats Paper
  • Paper ties with Paper but loses to Scissors
  • Rock beats Scissors but loses to Paper

A tie and a win is clearly the best result, to ‘Scissors’ is the overall winner.

C8: When considering all responses to these questions, including this one, which whole number is most commonly answered? (Closest to the actual most commonly answered whole number across all questions wins)

Perhaps the ultimate test of figuring out what other people will choose, while ‘1’ feels like a pretty safe choice the winner was ‘5’ – and it would not have been if 3 fewer people chose it in response to this question!

For reference, the actual distribution of numerical responses to all questions is as follows (excluding the very long tail of numbers with only one occurrence):


Did this work? Did it make any sense?

In retrospect the ‘Runner up’ round felt a bit too confusing and random. It would have been neater and also more enlightening to have people give a straightforward response and then go on to guess about the runner up. One reason I chose not to do that was that it would make the survey a lot longer – or the same length with much fewer questions in play. The second reason though was that I liked how mind-boggling it might be!

The grouping problem got pretty difficult. If I did this sort of thing again, I would definitely try to anticipate how grouping might work and specify guidelines for it within the question itself.

The responses to this quiz went on to inform an actual ‘Family Fortunes’ style quiz game I ran in person, specifically because I wanted to do something that was like a quiz but a lot weirder, and I liked how that worked a lot – that will be written up elsewhere.

Thanks again to everyone who participated!


As promised, I said I would publish the top 10 scorers. There were quite a lot of ties though, so some of these tables get quite long.

In section A, a nice warm-up with 8 ‘majority wins’ question, 5 people managed to get 6 of them right – well done!

In section B, there were 8 ‘runner-up’ questions. Figuring out what will come 2nd out of the options submitted by people who are all trying to guess what will come 2nd is very difficult, but somehow Jordan got an incredible 6 of the 8 correct, a strong lead against the 6-way tie for 2nd with 3 points. Amazing!

Section C had a lot of weird questions, but also a lot where just by design very few players were likely to win the point (with the exception of the rock/paper/scissors one), so it’s especially impressive that Phil got 5 out of a possible 8.

I said I wouldn’t reveal who gave which answer, but I will say Phil did answer 3 Section C questions tactically and this did help him win at least one additional point… quite how he might have done that is left as an exercise for the reader.

Adding those all together, Jordan takes the lead with 13 out of 24!

  • Transmission Ends

You’re In A Room

What’s this all about then?

Fruitbat (that’s us – Tim Mannveille and Clare Huxley) ran a game called You’re In A Room at the recent(ish) first of Hide & Seek’s “Games with Audiences” Sandpit series, on Friday 25th May.

We’re particularly pleased with how it turned out, since it involved players creating challenges for each other (uh-oh…) but the nature of the format meant that players pretty much always completed those challenges by the skin of their teeth! (Hooray!)

This is the story of how that came about!

And why do you need me?

The game involves a small team guiding one blind-folded player through a virtual dungeon, with that player generally asking a bunch of questions. We’re having form reflect content by you playing that role.

Where am I?

You’re In A Room! This was the repeated-and-therefore-magical phrase that was invoked whenever a player advanced in the late 80s/early 90s kid’s TV show Knightmare, in which a player in a bucket-like sight-blocking helmet was blue-screen-comped into a matt-painted / CG dungeon populated with actors, props, and virtual peril, there to be guided by their team-mates watching over them through a television, thereby creating a game perfectly designed to make you want to shout at the screen. You can get an idea of the show very quickly by watching this clip:

So where does your version fit in?

One day at work, Tim discovered a desk-side bin hidden inside another desk-side bin, with a piece of paper taped to it featuring an upside-down face and the word ‘Friday’:

Clearly, the previous occupants of the office would wear this bucket on their heads on Fridays in some kind of ritual, the purpose of which we can only guess at.

On hearing this, Clare was reminded of an impulse from long ago to make some kind of “bucket head” game, using the sighted-players-guide-player-with-bucket-on-head device so familiar from Knightmare.

A few days later, Hide & Seek announced the dates of their summer 2012 sandpits, starting with one themed around “Sports and Game Shows,” and even though this was only four weeks away and all we really had as an idea was “Bucket Head”, we figured this was the perfect opportunity to do it.

“To achieve great things, two things are needed; a plan, and not quite enough time.”
Leonard Bernstein

Clearly, we didn’t quite have enough time. So all we needed now was a plan.

What’s in this room I’m in anyway?

What? Er, a computer. Anyway, the plan was to very quickly come up with a workable idea, playtest it once, tweak it, and run it.

After quickly veering away from an ambitious idea involving multiple simultaneous players guided by walkie-talkies (which we may may well come back to), we were ultimately drawn to a more literal idea of creating a kind of mini, Sweded version of Knightmare, the hope being that some of the back-seat-player audience involvement engendered by the original show would carry across.

"You're in a room, it looks like there's a locked door at the far end..."

Clare suggested one key twist: that the players should have the opportunity to come up with dungeons themselves, thereby solving the problem that the original Knightmare had more people running the game than actually playing it. But could it work?

We distilled what we thought were the key elements from Knightmare (a character, a puzzle, a reward, a physical challenge, an obstacle that could only be passed with the right item or spell), created three-room templates, rustled up some friends, and just gave it a go.

"Where am I?" "You're in a room!" "Is this fun?"

So did it work?

The optimistic vision was that a game might take 5 minutes, and Tim in particular was drawn to the idea of a constantly rolling drop-in format in which one could rapidly move up a sort of engagement ladder:

From passive audience member, to…
> Back-seat-player
> Player
> Dungeon designer
> Dungeon operator.

The test very quickly revealed a problem with this idea: people naturally tended to design complex dungeons that took more like 15-30 minutes to play, and some of their rooms would take longer to set up than to play – not ideal.

On the positive side, it had one particularly outstanding feature: dungeon-designers wanted the players to experience everything they had made, so when playing the part of monsters/hazards/obstacles, they would tend to tweak things on the fly such that a player would succeed – but only just. This is great, because we’ve seen before that the moments that players enjoy the most are those in which they feel they succeeded by the narrowest of margins. We had stumbled upon a game format that pretty much generates these moments by default. Wow!

So how did you tweak it? And does this computer have internet access?

Er, yes it does. First we revised the instructions (and briefing) so that, with some guidance from us, a team could hopefully invent a dungeon in 15 minutes that would itself take 15 minutes to play.

Second, we had another look at the logistics, and were drawn to the idea of a structured rolling format with two tracks of players – dungeon-designers, and dungeon-players – hoping that this would create exactly the opportunity for progressive involvement (from audience member on upwards) mentioned at the start.

Finally, we realised that riddles were a fun feature of the genre, but it was tough to come up with them on demand, so we prepared a random selection that players could draw from if they got stuck.

On liaising with Hide & Seek, we were advised that this 15-minute version probably wouldn’t schedule well with other games (which tend to have defined 30-to-45-minute slots); a better format would be to have defined 1-hour slots in which 3 teams could first design their dungeons, then rotate through the roles of audience/dungeon-operators/players over three 15-minute slots. Although this removed the possibility of allowing spectators to become players and then performers, it seemed the most practical way to try the game out in this context – and in particular, to see if we really could have people stick to that 15-minute time limit.

What happened?

What happened was this: for the first group of players, we actively pushed to have dungeons designed and run as quickly as possible – and ended up finishing the hour slot with just under 10 minutes to spare. We let things follow a more natural pace for the remaining two hour-long slots, and amazingly they worked just about perfectly.

The other nice feature was that in the absence of specific theming, people came up with a huge range of dungeon types. Here’s a quick overview of what was produced:

7pm-8pm, Team A: North Korea Dungeon. Great use of one of our hat props for that authentic single-party-state feel.

7pm-8pm, Team B: Fun House Dungeon (a kids gameshow from the same era and channel as Knightmare). We particularly liked the use of the whiteboards as story-board-like captions, for example, the guy playing Pat Sharp holding a sign that says “PAT SHARP APPEARS” as he appears:

7pm-8pm, Team C: The Mirror Dungeon. The players find a map with mirror writing showing that the left exit from the room is deadly. Which way do they go?

8pm-9pm, Team A: A somewhat adult-themed dungeon we later realised was probably called the Knightclub dungeon. Notable for use of the spell ‘B-E-E-R-G-O-G-G-L-E-S’.

8pm-9pm, Team B: The Jelly Dungeon, featuring an extra challenging variant on the form: with the entire dungeon made of jelly, the blindfolded player could only ever stand on one leg. Tricky, but as it turns out, doable – hooray!

8pm-9pm, Team C: The Classical Dungeon, in which a player had to play ‘Stairway to Heaven’ on a paper flute to put Cerberus to sleep. Most notable for this exchange:

One of Cerebus’s heads: That’s not how it goes at all!
Player: I’m doing the intro!

9pm-10pm, Team A: Robot Science dungeon. An amazing use of the form in which the player first discovers that they are actually a robot, before solving the final room by discovering the human within themselves. Not bad for 15 minutes.

9pm-10pm, Team B: The Not-At-All-Based-On-Indiana-Jones Dungeon, most memorable for the entire dungeon-operating team stamping their feet just behind the player to simulate the approach of a giant boulder.

9pm-10pm, Team C: The Pirate Dungeon, featuring some great pirate acting and buried treasure recovered seconds before the Sandpit event came to an end. Hooray!

I write a blog post about You’re In A Room. So what would you change for next time?

A lot depends on the setting. If the venue was suitable, we’d love to try some version in which the game is ‘rolling’ – probably with some volunteers on hand to do the dungeon-operating part by default, but who could step aside if people want to come up with a dungeon themselves. This means players get to see how the game works before designing a dungeon (a major challenge in this first run of the game), and also brings back the possibility of bystanders graduating from back-seat players to actual players.

Another setting-dependent issue is the noise level. With two other games making use of amplified sound nearby, and the general hubbub, it was tough for the dungeoneers to hear their directions. As a quick fix, teams simply stepped into the room themselves, which made things a bit surreal, but did work. Unfortunately, this also meant that any bystanders / audience members would struggle to follow what was happening unless they too got right up close to the action. In a similar situation, some kind of amplification may be called for.

The most important change would be to further distill the instructions. The set we used worked – players were indeed able to come up with suitable dungeons within the time limit – but they did look quite intimidating. If players also get to see a dungeon in action first, this also becomes much easier.

Finally, we’re also thinking about adding more hats, because hats are where it all began, and because everyone loves hats.

Okay, I publish the post.

-Tim & Clare, now known as Octopus Fruitbat

analysis game

Learning to Cheat Without Breaking the Rules, Part 2: Frameworks for Cheating

For the same reason as in Part 1, these posts are illustrated using the Hand of Fate: Comic Strip Playing Cards by Karen Rubins. This is Part 2. Part 3 is here.

In the first part I explained how a sequence of games taught me to become comfortable with bluffing and even lying outright, from a position of not being able to do either. That’s simple. The harder question is this: what kind of games can teach you to bend or break rules?

To tackle that, we should probably step back and ask how games teach us anything. And to tackle that, we should probably make sure we have a working definition for a game.

Here’s one provided by Greg Costikyan (emphasis added):

A game is a form of art in which participants, termed players, make decisions in order to manage resources through game tokens in the pursuit of a goal.

Yes, something like that. From there it seems clear that playing with decisions in a game is how we learn. So the question becomes instead: what kind of game could let us play with the decision of whether or not to break rules?

Form or Content?

I don’t know of one, but I imagine you could make a game about cheating (in the rule-breaking sense, not just bluffing or lying). Perhaps players take the role of political factions within a Kafkaesque bureaucracy, with in-game costs and benefits for following or breaking in-game rules. Or it’s Moon colonists with competing sets of societal rules, struggling for domination while following or breaking the rule sets of the competing factions. Or, I wonder, do football simulations these days give you the option of going for a risky tackle, knowing that you risk a red card if it goes badly? I don’t know, but they certainly could.

But in the process of abstracting away the costs and benefits of cheating, it feels as if something gets lost. Just as making decisions in poker has a very different feel when real money is on the table, a game with a cheating theme seems far removed from the potential power of a game that is about cheating the very rules of the game itself.

So we want a game whose rules somehow give players interesting decisions to make regarding whether or not to follow those rules, which sounds like an almost Gödelian paradox. So let’s take a step to one side and look at exactly what it is that drives whether or not rules are followed in a game.

Game frameworks for cheating

There are two key dimensions to consider here, I think:

1) Do people play the game to win, or do they play it for fun? Perhaps more accurately, does the pleasure come from winning – or striving to do so – or just from taking part? There’s a spectrum here, of course, and players of the same game could be at different points along it. For the sake of simplicity we’ll assume it’s binary.

2) How are the rules enforced in the game?

  • By design. In video games this is simple: the game should only let you do what you are allowed to do. Other kinds of games can also facilitate enforcement by their design – for example, playing a piece in Connect 4 makes a distinctive noise, so it’s hard to secretly take an extra turn if your opponent is briefly distracted.
  • By the other players. In most well-defined tabletop or party games, it’s understood that the players are watching each other to ensure they all play by the rules.
  • By a moderator. Generally speaking, if a game can’t work with either of the above approaches, it resorts to some kind of moderator who enforces the rules.

Let’s take a look at how cheating works across the 6 kinds of game this framework implies, and in particular whether choosing to bend or break rules is an interesting* choice for the players to make.

I should probably define ‘interesting’ here. Let’s say that an interesting choice is somewhat balanced (one choice is not obviously better than another), not damaging to longer-term goals, and is fun. And then let’s not try to define ‘fun’ or we’ll be here all day.

Games Played to Win

Play to Win, Design enforced (e.g. PvP video games)
The rules are enforced by the design of the game. As a corollary, anything you can do ought to be allowed, since cheating is theoretically impossible. In reality, it’s not that simple.

A reasonably notorious example is ‘snaking’ in Mario Kart DS. By performing a particular manoeuvre when cornering you get a speed boost. For certain karts, you could perform this manoeuvre almost continuously, alternating left and right – ‘snaking’ – and you could gain more from those speed boosts than you lost by taking a wiggling route. This takes some skill and practice, which makes the benefit seem fair. And since it’s possible in the game, it should be permissible, even if it’s not clear it was the designers’ original intent.

In practice, however, snaking gives such a big advantage that a player that does it will almost always beat a player that doesn’t, no matter how skilled. People are playing to win – that’s what usually makes a race fun – but if even one player is snaking, anyone that can’t snake can’t win, and so, implicitly, can’t have fun. For that reason, the London DS meetup group I played with had a no-snaking rule, and I see on the internet that other groups did as well. Snaking was considered a kind of cheating, and nobody would want to play with you. So no-snaking becomes an informal rule, and this becomes effectively the same as Player Enforced Play to Win.

(You can dive much further into this topic with Sirlin’s “Playing to win” archives)

Play to Win, Player enforced (e.g. Chess or Go)
As noted by Kirk Battle in this Kill Screen article, players can apply a certain level of flexibility to how rules are enforced in a game like Magic: The Gathering, and get a better game as a result. But this is not a situation in which such ‘cheating’ represents an interesting choice.

In it’s more blatant form, cheating in these games will always be considered bad form, and if discovered can have bad social and future gaming consequences. Chess or Go (when played without an audience) fall into this category. A player would only choose to cheat if they thought the short-term advantage would outweigh the risk of long-term negative consequences if they were found out. That’s not a very interesting choice either, according to the criteria above.

Play to Win, Moderator enforced (e.g. Football, Tennis)
Cheating outside of the moderator’s sight gives an advantage, and can’t be stopped in this system. You’re playing to win, and if you don’t exploit this fact, maybe the other guys will.

We see this in competitive sports like football or tennis, and it seems to be (I say this as an outsider) a key part of the entertainment: arguing about whether something counts as a foul, which side of the line the ball bounced on, and in general whether or not the moderator’s decisions are accurate.

Does this mean there’s some scope here for a game that could teach us about cheating? It seems close, but if people are playing to win, it gets dangerous – some degree of violence attempted outside of the moderator’s views seems likely, so I’m going to disqualify this on the grounds of encouraging non-fun behaviour.

Games Played for Fun

Play for Fun, Design enforced (e.g. Endless MMOs)
If you (and everyone else) is playing for fun, things change. If something is possible, and makes for more fun, few can blame you. If you were (somehow) playing Mario Kart just for fun, and through a streak of bad luck found yourself in dead last by some margin, perhaps it would be okay to snake your way to 7th place – that would be more fun for all concerned.

In Kingdom of Loathing, a kind-of MMO in which fun is primarily derived from exploration, an ‘exploit’ was discovered to generate more meat (the game’s currency). Who could resist such an exploit? Why resist it? Could you really begrudge those that used it?

In the case of online games like this, the rules are enforced by design, but there’s also some moderation in the form of game updates and code changes in response to things like this. In this particular case, the exploit effectively crashed the game’s economy, which impacts everyone’s ability to have fun. (Brilliantly, this was fixed with the addition of some entertaining currency sinks, rather than some kind of hard rollback).

So cheating is kind of interesting here, but again becomes more about player- and moderator-enforcement, so strictly speaking this category is ruled out.

Play for Fun, Player enforced (e.g. Mornington Crescent, DDR)
At this point, what constitutes ‘cheating’ is massively dependent on the players: having fun is more important than the rules. We’re right on the edge of what constitutes a game here and it’s an area I think most adults struggle to give themselves permission to enter. The most well-known example I can think of in this category is Mornington Crescent.

Less directly, this arises in Dance Dance Revolution (aka Dancing Stage), in which (for the benefit of the one person reading this that doesn’t know) players must step on specific directional arrows in time to the music. On the arcade machines, a raised bar is supplied behind the player, ostensibly to prevent anyone from falling off the dancing platform backwards. At a high level of gameplay, working out how to shift your bodyweight between feet while meeting a high-speed series of instructions is part of the challenge. However, some players realised they could lean their weight back on the bar and tap away with their feet without worrying about this issue; this also uses less energy. Is this cheating? As could be expected, that depends on who is playing.

In both of these cases, the decision of whether or not to cheat tends to have very little riding on it, so is unlikely to be interesting.

Play for Fun, Moderator enforced (e.g. Schooner or Later)
The Moderator’s role is to ensure people have fun, arguably as a higher priority than ensuring that people follow the rules. In this context, you might ‘cheat’ but do so with the moderator’s implicit or explicit blessing; or you might try to cheat by hiding your action from or misleading the moderator, and since people are only playing for fun this shouldn’t lead to anything particularly harmful. With the right set of incentives and approach by the moderator(s), this could well provide a framework in which a decision to cheat is actually interesting.

In conclusion, returning to our grid of possible games, here’s what the options for learning to cheat look like in each:

This final category is where The Haberdashery’s Schooner or Later comes in, the game that led me to cheat in a manner that could only be described as brazen, and shocked me into this whole line of thinking. I’ll describe how exactly that came about in the third and final part of this series.

Tim Mannveille tweets as @metatim, and has previously not cheated in order to earn stickers he made up at the Edinburgh International Film Festival.