The Paradox of Catch Up Mechanics

Playing Power Grid the other day I thought “I could build an extra house, but I don’t want to, because then I’ll buy resources later and that’ll cost more money than the extra house would earn”. There’s a mechanic that penalises the player who’s ahead so good play becomes – in part – about fooling the mechanic so that the game thinks someone else is in the lead.

Fooling the Game is Good Play

There are a lot of games with catch up mechanics where a bonus is offered to a player who’s losing or a penalty is applied to one who is ahead. Playing these games well often involves manipulating the metric that the game uses to determine who’s winning so that it does a poor job at detecting your progress.

This is true even of “soft” catch up mechanics that have no formal implementation. When Munchkin was popular with groups I played with people very quickly realised that the first player to level nine almost never won. There’s no formal catch up mechanic but players are obliged to draw cards of the form “Play when another player tries to do something, it goes horribly, sucks to be them.” so inevitably the first person to nearly win would face a slew of these. The single skill most consistently correlated with winning was an ability to fool other players into believing that you expended all of your ‘screw you’ cards on the last player’s near victory so that they squander theirs on the current player, you can stop them and there’s nobody left with resources to oppose your final fight.

Whether you’re fooling people or systems, convincing them you’re behind improves your chances.

The Paradox of Catch Up Mechanics

A catch up mechanic needs to determine which players are winning and losing to be effective.

Who will benefit from the catch up mechanic is a factor in determining which players are winning and losing.

I’m convinced that a truly accurate catch up mechanic would find itself in a paradox really quickly. I can see it being of the form “Jeff has the least stuff so we’ll let him buy the cheap resources but if Jeff gets the cheap resources their value exceeds that of the difference between how much stuff he and Sally have. So really Sally is losing so we’ll let her buy the cheap resources, but if Jeff doesn’t get the cheap resources…”

So What’s the Point?

Honestly, I don’t know. We’re all going to die in the end anyway.

But in terms of catch up mechanics they might not achieve “Find weakest player, help them catch up” but they’re still doing something valuable.

If a winning player determines that their best move is to use their advantageous position to break some of their things or trade away useful resources inefficiently in order to get the catch up mechanic bonus – they are still breaking some of their things or trading away useful resources inefficiently.

Consider the following:

We’re playing a game and I’m winning.

I have 8 points and the option to add or subtract 2 from my score

You have 7 points and have used up all of your options for this turn

At the end of the turn whoever has the least points gets a 3 point bonus

If I add to my score I go up to 10 and you get the bonus, also hitting 10 – we’re tied!

If I subtract from my score then I go to 6 but then claim the bonus shooting up to 9, while you’re still on 7 – I’m two points ahead!

So in the surface because I’m winning and have more effective options I can manipulate the catch up mechanic so it benefits me and not you. Leaving me 2 points ahead by abusing a mechanic that was supposed to help you.

Digging a little deeper however, consider what happens if we delete the catch up mechanic: In that instance I add 2 to my score and am leading 10-7 – I’m three points ahead!

So despite the fact that I got the bonus from the catch up mechanic and it didn’t change your score one iota, the mere fact that it existed meant that I’m leading by 2 points rather than 3. Just be encouraging me to make the otherwise suboptimal move of subtracting two from my score.

Catch Up Mechanics as Incentives

The problem that I’d identified was not in catch up mechanics themselves but in how I’ve been conceptualising them.

“A catch up mechanic is a mechanic that identifies the losing player and gives them a boost allowing them to catch up” was the wrong way to think about them.

“A catch up mechanic is an incentive for a leader to make moves that would otherwise be suboptimal to allow others to catch up” seems a better way to conceptualise it.

This way of looking at them leads to some differences in how they’re designed. It seems naïve to go “We’ll assume our players play the same way as if the mechanic didn’t exist and offer people who are behind a leg up” when players are generally pretty good at games and taking advantage of the rules in order to win.

Instead respecting the players capabilities and going “Everyone will try to use the mechanic, including the current leader, who may even have a greater ability to access it if their position has given them more options. The goal is to make sure that when the leader is encouraged to use it the price of doing so is smaller than the reward, but the net effect is that they can’t be a runaway leader because they’re still paying those prices.”

A Good Catch Up Mechanic

Viewed through this lens a good catch up mechanic has these traits:

Stratified: If there are multiple players having the mechanic be effective at each step means careful thought needs to go into when to use it and when to pull ahead. If there’s no advantage in dropping from 1st to 2nd place it won’t work because the boost won’t be big enough to justify dropping 1st to 5th

Positive: If a catch up mechanic is adding to a players progress then each round everyone will progress, even if some players are deliberately losing progress to access the mechanic they wouldn’t bother if the boost didn’t outweigh the loss. A negative catch up mechanic risks creating a game that will never end if the penalty for being ahead is high enough then someone who deliberately backs off and someone who gets hit by it are both behind their starting point and the game may never end.

Manipulable: Players should be able to manipulate the thing that the catch up mechanic is assessing in a timely fashion. If the leader can’t deliberately lose some resources to access it then either it doesn’t propel people past the leader or it was inevitable that someone will be propelled past the leader. In which case they were only ahead on some metric that failed to capture winning since their being surpassed was inevitable.

Costly to Manipulate: There’s no sense in offering a catch up mechanic to whoever has the least cash if players can buy gold, get the mechanic and sell their gold to have the same amount of cash in the very next turn. To work there needs to be some friction in how the mechanic is accessed.

Incidentally you might notice these are all attributes shared by the mechanics of Power Grid and are weak to completely absent for Mario Karts Blue Shells. So I feel like they work for at least those two examples 😉


This week I’ve been testing Genesis and there’s a broad agreement between playtest groups that the most troublesome thing in the game at the moment is if someone decides to be a god of madness. This domain grants the power “This power targets the world, for the rest of the turn any champion with a power that targets one or more champions instead targets only itself.”


It’s a problem because it’s a power that modifies other powers, a metapower, which have two important features:

1) Players get excited about them, they offer a lot of flexibility adding a relatively large number of things someone can do in the game for a relatively small complexity cost. Generally they’re some of the powers that get the most positive feedback.
2) They’re a massive pain in the arse to design because they impact the design of every other power.

The fact that madness exists obliges me to look at every other power and go “What does this power do under the influence of madness? Is it obvious? Is it balanced? Is it interesting?” which is a lot of work for one domain.

It goes beyond individual cards too and into combinations of cards. For instance if there’s nothing in play that can remove a madness wielding champion without directly targeting it then everyone will target themselves constantly all game. This leads to degenerate games where the madness domain might as well read “Stop playing now, the game is won by whoever has the most targetting powers that give bonuses and the fewest that give penalties.”

Metapower Combos

Astute readers may have noticed the construction of the madness power is something that’s necessitated by the existence of powers like it. Having some of the odder powers read “This power targets the world, this rule applies for a turn” means that metapowers that mess with things that target champions exclude them – meaning you don’t get combos of metapowers on metapowers.

Consider a situation like this:

Power1 targets everyone and makes them target themselves.
Power2 targets each players highest level character and makes them target everyone on their own side.
Power3 targets everyone and cancels their powers, making them have no effect.

What the hell happens? Does power 3 make everything have no effect so the other two don’t matter? Or does power 1 mean that power 3 is only making itself have no effect (which means it doesn’t which means it does)? Or does power 2 mean power 1 only applies to its controllers stuff? If that’s true does that mean it then targets only itself so it retroactively doesn’t apply to the rest of its controllers stuff?

You can do metapowers that affect each other, but you either need a rigid framework that makes the ways they influence each other relatively uninteresting or to manage them carefully which creates exponentially more problems the more of them you have.

Implementing Metapowers

Despite the caveats they are neat and offer advantages. Lets agree that we’re going to do them and get into how to implement them.

The main thing to shoot for is consistency of language. This is generally desirable anyway since it makes games easier to learn and rulebooks read more smoothly. For this sort of thing it really matters though. Consider the original wording for the power a player gained if they elected to play as a god of vengeance:

“When this power is in play and a champion dies then the champion with a power that killed them also dies.”

What happens if that power is madness’d? The wording makes it unclear. Consider some alternative wordings:

“This power targets the world. For the rest of this turn if a champion dies the champion with the power that killed them also dies.”
“This power targets all champions. If a target dies, the champion with the power that killed it also dies.”
“This power targets any champion that’s power has killed at least one other champion. The target dies.”

Standardising powers to always start “This power targets X” obliges the power to be worded in such a way that it’s clear. You can see what madness would do for either of these wordings. It opens the way for any number of metapowers of the form “modify target” (So long as there’s some system for how they interact with each other to avert paradoxes)

Astute readers, who get time in the spotlight for the second time this post, will have spotted the problem with all of the suggested vengeance wordings above. Namely that the champion of vengeance always kills itself – since if (say) Water kills Fire then Vengeance kills Water. In that case Vengeance has killed Water so Vengeance kills Vengeance. It needs an exception for itself, but I left that out of the example to keep us on topic 😉

Aside from standardising power wording to make sure that metapowers work consistently there are two other approaches to consider:

The first is to look at every 2 card combination of powers and do a mental “Is this okay? Is it clear what these do together? Does that combination break the game in some critical way?” check.

The second and most important is to do lots of playtesting. You will miss things or a thing will seem clear to you while being obtuse to your players. Always do everything you can that isn’t playtesting before squandering tester time on a thing you could’ve fixed yourself – but never skip playtesting. It’ll catch problems you wouldn’t have dreamed of.

What are We to Do About Madness?

I’ve said what I wanted to about game design in general, but it feels like a tease to discuss a topic on a game I’m working on without describing the resolution.

In reality there isn’t a resolution yet because I’ve written some new versions of the affected cards for the next playtest and that playtest will undoubtedly change things again.

The change I’m looking at is “This power targets each players champion(s) with the lowest level. If they have abilities that target champions they will target only themselves instead of their usual targets. If a champion with this power targets itself its ability is unaffected.”

The reasons for this are:
1) There are some standard forms for the game in there. Opening with explicit target information is standard. “Champion(s) with the lowest/highest level” is standard (and applies a general game rule of “In a tie the thing happens to everyone whose tied”)
2) It moves away from the “targets the world” construction which while hardened against some interactions also makes things less interesting. I’ll revert to targets the world if testing shows this generating more problems.
3) Modifying it only to hit low level champions deals with a lot of the complex interactions. You can’t have degenerate games as easily because someone can drop a high level champion and then a low level champion that targets and kills the madness champion. The most broken combos were madness + domains with high level champions balanced by giving opponents bonuses in some circumstances – these are harder to pull off if the madness card can’t target those high level champions. Low level champions also tend to have more interesting and powerful abilities so hitting everyone’s lowest level champion is likely to preserve the meddling troublemaking options that made people like the madness domain in the first place.

I’ve no idea if it’ll work, but we’ll see after the next playtest.

Incidentally I’m recruiting more testers for this game soon. I worry that my existing testers are too familiar with the game which distorts feedback. If you’re already on the 3DTotal mailing list you’ll get a message about it in a week or two. If you’d be interested in PnPing a version or trying it as a Tabletop Simulator mod drop me a comment or email to [email protected]

Mathematically Equivalent Rules

Why Change A Rule?

Genesis is at a point that it plays great with people who know how to play it and people have started actively requesting it at games nights (Always a good sign that a prototype is approaching being ready for publishing!) However I’m finding a noticeable proportion of new players are having a hard time with the game in places and sometimes it’s a downright unenjoyable experience.

Why Not Change A Rule?

The game is working great. People are really enjoying it and it’s working smoothly and well. The balance is approaching where I want it. Games have emotional high points that I’m pleased to deliver. Feedback is generally good.

How Do We Change Something without Changing It?

We use a mathematically equivalent rule!

Let me give an example:

For most abilities timing doesn’t matter, they can go off more or less simultaneously. However for a few the order makes a difference so these each have a speed from 1-10. You resolve all speed 10 abilities, then speed 9, then speed 8 and so on.

A sticking point for the game was that some players would read the rules, but then treat speed 1 as the fastest and speed 10 as the slowest. When quizzed on how this happens a reason emerges: Players intuited that speed 1 goes 1st and 2 goes 2nd and so on.

So the mathematically equivalent rule in this situation is to simply multiply all speeds by 10. Now cards are speed 10, 20, 30 etc. The problem just disappears, while there’s a reason to intuit 1 is faster than 10, there’s no equivalent to believe 10 might be faster than 100.

Critically this doesn’t change how the game is resolved at all – the same abilities go in the same order under both sets of rules. There is no situation in which using one rule rather than the other changes the outcome (in terms of game state) – but using one rule rather than another changes the outcome (in terms of ability to grasp the rules first time) so it’s clearly a better choice.

I’m pretty sure I’m not the first designer to go through this specific process. In fact as a teenage player some years ago I always thought it was a bit silly that Robo Rally used speeds that were multiples of a hundred. It seemed pointless to busy up the cards with extra digits that were always predictably meaningless. But seeing the difference in how players react: Now I get it.

I wonder just how often designers worry at some problem, hit upon a solution and then realise they’ve seen that solution before but didn’t recognise it as a solution because the problem doesn’t exist in that game?

Another Problem

Presently Genesis abilities often choose targets based on their power and can sometimes modify their power. For instance an ability might be something like “All champions with a power of 5 or less gain +3 power”

Abilities always target based on a cards natural power – under no circumstances will they select a target based on its total modified value. So in the above example if a power 6 card was reduced to 5 by some other ability, it wouldn’t get a +3. The +3 is only for champions with 5 or less printed on the card.

This seemed like a fairly solid rule to me, certainly compared to previous iterations where some things cared about unmodified values and others cared about final values. “Always unmodified” didn’t strike me as difficult.

I dramatically underestimated how complex that’d be for players with different backgrounds. There are a lot of different ways in which people think and some portion of players just can’t grasp “always unmodified” as a rule.

But if we flip it over to abilities like “All champions with a level of 5 or less gain +3 in combat” it becomes easy. It’s exactly the same rule, but moving away from “modified power” and “unmodified power” to two values called different things makes it much easier to intuit.

Alright, but how do we apply this to design in general?

The commonality between situations helped by this sort of intervention is that they have the following traits:

1) The rule works really well for people who understand it.
2) A portion of players play the rule incorrectly.
3) These incorrect plays are due to intuiting a competing rule and the incorrect intuition is known.

From there the rule can be modified to remove the intuited rule as a valid interpretation without changing the thing that’s making it work well for a lot of players.

In the first example the intuition was “1 means 1st” so the modification is to get rid of 1.

In the second example the intuition was “If this alters power I can use it to get the right power to use this other thing that cares about power” so the modification is to stop using the word power to refer to what are essentially two different things.

Sometimes it’s better to embrace the intuition wholesale and actually change the rule to match whatever people intuit that it is. It was an option to change powers to say 1st, 2nd etc. if all powers were in play every turn I’d probably have done that – but in this case avoided it because the notion that 2nd goes 1st (because 1st isn’t in play) might prove equally unintuitive with players trying to execute one of the unmarked powers before it.

So if you’re working on a game, have a quick scan of your playtester feedback and see if you’ve got situations like that. It’s easy to dismiss “We got this rule wrong but then read it again and then it played fine” as something that worked itself out, but it’s perfectly possible to use it to improve the odds that more people will get it right first time.

After all, some players won’t give your game more than one chance 😉


I’m presently working on a game in which players control runners who are navigating a maze with the assistance of various teleportation devices. Each player has a hand of devices and uses 1-3 each move to help their runner get to where they want to be. I’d like the teleporters to all feel different to use and the ideal move not to be obvious so there are some meaningful choices in what’s used when and what’s held back. I figured it might be interesting to use todays post to brainstorm ideas and talk about their pros and cons.

The Lovely No Limits Teleporter
Effect: Place yourself on any space within N spaces of your current location.
Pros: Simple. Useful.
Cons: Simply better than almost anything else. Would need a small N to balance with other cards.

The Leeroy Jenkins
Effect: Move forwards N spaces ignoring all intervening stuff.
Pros: Simple.
Cons: Very similar to normal movement. Board design or other player obstructions need to be significant to justify its inclusion.

Cardinal Teleportio
Effect: Move in a particular compass direction N spaces ignoring all intervening stuff.
Pros: Simple.
Cons: Could be completely useless to someone moving towards an objective in a different direction. Potential to be a “lucky” or “unlucky” draw irrespective of players skill at setting up opportunities.

Place Thief
Effect: Choose another runner within N steps, swap position.
Pros: Could have take that style effects.
Cons: Could have take that style effects.

Murphy’s Teleporter
Effect: Your opponent chooses a space you go there.
Pros: Puts an interesting choice for one player in another players turn, potential for dramatic desperation moves.
Cons: Would need some significant other advantage for anyone to ever be willing to play it.

Effect: Choose another runner who is N spaces away. Place yourself on any space within N spaces of them.
Pros: Potential to act as a catch up option, potential for teamwork in setting up several runners to use each others positions.
Cons: Might be too many turns on which a useful anchor doesn’t exist and its just a dead card. Board size and population density strongly impact usefulness.

The Escher Step
Effect: Choose another space with N spaces that is the same as your current space in all respects but location (If you are in a 1 wide corridor next to an exist, it must be a 1 wide corridor next to an exit)
Pros: Leads to interesting possibilities with a high N.
Cons: Awkward to codify in rules, possible analysis paralysis.

These also work in most combinations. For instance you could have Murphy’s Slingshot which anchors to a player of your choice but your opponent chooses in which direction it fires you.

What will work in practice will depend a lot on the interaction between the teleport steps and the base moves – but having a small list to populate a simple prototype allows for some basic testing and refinement (and all game design is testing).

Even the obviously bad ideas belong in this sort of early test. Murphy’s isn’t going to make it through in its unmodified format – but throwing in the card to see what the implications are when it’s played give an indication of the best thing to combine it with or what sort of magnitude of payoff it requires to justify its inclusion.

Early testing is as much about finding exactly why and how things don’t work as finding what does work. The lessons taken from the more obvious missteps can prevent more subtle ones from creeping in later in the process.

Hope you’re all having fun, I’m off to test a bad game.

Consequences of Precise Probabilities


I’ve been playing the digital adaptation of the Pathfinder Adventure Card Game lately. Besides the bugs (and dear me there are a lot) it’s a pretty faithful recreation of the physical game – yet I’m finding one aspect of it is making me play very differently: Every time you do something to modify a die roll the game shows you the probability of the roll succeeding and permits you to undo that modification.

Probabilities in Rise of the Runelords

Many of the bonuses in the game are represented by extra dice. At any given point you usually have access to a whole load of these bonuses, but use of them has a very noticeable opportunity cost leading to you seeking to use as few as possible. On a typical turn you might hit into a situation like this:

You’re righting something that needs a total of 22 to beat.
You subtract one from each die rolled.

You can punch at D6 or attack with a sword at D6+D8 or cast a spell at D10+D6+D6+2
You have four blessings which will give you a D6 if you’re attacking physically or a D10 if it’s a spell.
One of those blessings gives two dice instead of one if it’s used for a D6.
You could discard the sword, losing the ability to use it again and get an extra D6
Your mate you shuffle part of his hand into the deck to provide an extra D4+2
You have an item you can discard for a bonus D4 but that applies to all fights this round (Which is at its best if you save your blessings since each blessing could also be used to flip another card and maybe get into another fight if it’s not used here)
One of the other players has a spell that gives +3 to strength rolls so works with the sword but not if you’ve cast a spell.

Computing the probability of success using each combination of bonuses would be a massive headache. Heck even the base combinations would take some doing. But the game does it instantly and displays the number for you.

How does it change the game?

The obvious first degree effects are that it means I make optimal decisions. If I can get a bonus one of two ways and the cost to both is the same then I’ll always pick the bonus that’s best in the situation. That’s not such a big deal, I’d probably have done it anyway most of the time.

The major impact it’s had has been in how I’ve learned the game over repeat plays.

In tabletop I doubt I’d notice the difference between a 91% and 97% chance of success. It’d always be at the point of “I’ve got a big pile of dice and the average expected result is at least twice what I need to roll – it’s gonna be fine.” Yet providing the probabilities makes me sensitive to it. It’s a game that can be won or lost on a single roll – tripling the chance of failure for that roll (from 3% to 9%) is actually a really important difference in those critical situations.

It’s harder to pin down how these things are altering the emotional experience of the game. I’m making better moves, but am I enjoying the game more or less for that? The reaction to a roll seems different – on the one hand I get “Well I decided to stick at 95% and not throw another card in, 1 in 20 chances happen all the time” where I’d have got “That roll is so absurdly below average, I hate you dice.” but on the other hand failing a >99% roll is a worse kick in the teeth than it’d have been if you weren’t aware of just how good the odds were.

There’s also a conflating factor in the game being single player. “Shall I throw in my other characters extra card to squeak an extra 2% chance to win out of this roll” is a fundamentally different question to “Shall I ask Jane to give up her extra go so I can have a 2% extra chance to succeed?”

Overall I think being aware of the exact probabilities sharpens the game. Dice are fickle, but over time, over many rolls, fair. It makes me more aware of smaller changes which in turn means I think about choices that I might otherwise have discarded out of hand. Ultimately it frees me from arithmetic to enjoy what the designers intended the game to be.

What’s the take away for designers?

Games can be more or less explicit about the probabilities of success involved in doing things. A magic computer box that does the numbers is not a necessity: Settlers of Catan dots its pieces to show dice probabilities. It would not be possible to do something for more complex mechanics like the examples shown here – but the mechanics themselves can be streamlined.

It’s also not an all or nothing approach. For instance Race for the Galaxy includes a card that allows you to guess the cost of the top card of the deck and draw it if you were correct. That card contains a little table showing how many of each cost of card are included in the deck. This doesn’t mean you can calculate the probability (I mean you could if you wanted to sit and count the number of cards of each cost currently in play and that you’ve personally discarded) but it makes it less obscure – you have a better idea of what it might be than if you didn’t have the contextual information.

Most games are likely to benefit from giving the players more information and more powerful tools to make decision – if the rest of the game supports that. There are some games that being able to work out the odds of each move *is* the gameplay and these calculations are intentionally on the cusp of human ability – but for more games working out the best move is the gameplay and understanding the odds of various outcomes is a tool in reaching the more interesting factors that define the best move in that particular game.


One Trillion Dollars

I’m presently working on a game in which a team of players take the role of evil masterminds trying to get one trillion dollars out of the UN. Each turn they get cards and can either play half of them or threaten to play all of them and offer not to for some fixed amount of money. Their goal is to get enough money doing this. The UN player can catch and arrest them all, but the cards they play make this harder, so they need to strike a balance between giving up too much money and losing and letting too many cards be played and be unable to win (Which as the evil players can get money in some other ways will also result in a loss).

Tone of Threats

The initial threats deck I’ve made by marking cities on a world map in a “This city is in a convenient place for my map” kind of a way and then picking something a mastermind might do to each city to threaten it.

This has lead to threats which have a tone all over the place. Kidnapping the queen of England is maybe a bit silly, but something someone could plausibly do that would definitely cause an uproar. Awakening the Egyptian mummies would definitely cause problems in Cairo but is clearly stepping into fantasy. You could definitely do something to Taipei and Beijing by annexing Taiwan but… actually that whole situation is pretty real and real people’s lives are going to be quite affected by how it plays out.

Just going through cities and writing the first thing that popped into my head has lead this version to be all over the place. This is a problem.

Who cares about tone?

I was going to start this off printing a user review of Escape the Nightmare but the user has deleted it :O

It boiled down to “I really like this game mechanically but can’t play it with my preferred groups because the tone of the art means it can’t be played in those circles.”

Which is a fair cop. I *did* make a light party game which had art including dismemberment and disfigurement. I probably shouldn’t have done, it made a great game less accessible to some audiences and less enjoyable to others.

How do you get tone right?

I don’t know. I really don’t. “Inconsistent tone” is an accusation that could be reasonably levelled at all of the games I’ve made.

I’m hoping that being more conscious of it is a good starting point. Previously I’ve tried to generate as many good ideas as possible that fit the theme – which has lead to a lot of good ideas. I’ve then winnowed them away through playtesting, removing things which don’t pull their weight gameplay wise – which is also very important.

Designers talk a lot of about theme vs mechanics and the compromises made in order to carry both. I think that it might be wise for me to look at this set of cards in terms of theme-tone-mechanics and start removing cards from the set that significantly fail in any one of the three areas.

I think it’ll be hard to say goodbye to something that fits the theme and presents great gameplay because of tone inconsistencies – but it should lead to a stronger game. I’m a very mechanics oriented player so I’m constantly fighting the desire to go “Screw everything else” but for most players enjoyment is a combination of factors. There will be opportunities to make relatively small sacrifices in other areas for relatively large gains here, so I just need to be alert for them.

What do you think?

This post has been less useful to designers reading it than others because rather than it being “Here’s a thing I’ve figured out, let me share” it’s “Here’s something I’m figuring out and where I’m going with it at the moment.” So some of the readers almost certainly have a better idea about it than I do – more than usual I’d encourage folks to comment and for readers to read the comments. I know I will be 😉


Dollar Auction

Dollar Auctions

A dollar auction is a thought experiment in economics in which someone auctions a dollar under the following terms: You can bid any amount you want, you can win a dollar for a cent, but if you make the second highest bid you still pay up.


Using the economic “perfectly rational” actors things get quite pricey.

Initially it makes sense to bid $0.01 to win $1
When someone else bids $0.02 it’s still a good proposition to bid $0.03 to win $1
This continues until someone bids $1. At this point the other bidder is looking at bidding $1.01 for $1. That’s a clear loss. But it is only a loss of $0.01, compared to a loss of $0.99 for stopping now. So a higher bid is still rational.
This continues until…actually it just continues. There’s no until. The dollar will sell for infinity dollars.

In economics this is a criticism of a particular means of conceptualising “perfectly rational” actors and the suggestion that the idea is flawed or at least needs some refinement. But we’re not economists.

Auctions in Games

Board games have been using auctions for a while. They’re a really neat mechanic since they sidestep a lot of balancing issues.

If something is auctioned the players are setting the price. You don’t need to say “Card A costs 4 and card B costs 5” and risk having valued those cards incorrectly. You just auction them off and the players will set the values. This is particularly powerful because it is sensative to the meta-context of the game – some things can be impossible to set costs for becuase their value changes dramatically depending on what else is in play, the state of the game and (sometimes) the sort of people playing.

There are two popular models for auctions in games:

Traditional auctions, in which players take it in turns to bid higher and higher until everyone but one player has backed down. They pay and get the thing, while everyone else takes their money back. Like you’d do in Monopoly or Power Grid. (I probably don’t need examples to illustrate this point but something tickled me about being able to use those two together)

Blind auctions, in which players simultaneously choose how much they’re going to bid and reveal all at once. All of the bids are lost, but only the highest bidder gets the thing. Or in some cases players are bidding for something akin to turn order and everyone is placed relative to their bid.

I can’t think of an example of a game that uses a dollar auction model in which the second place (or more) players lose their bids.

Could it work in a game?

Almost certainly someone has made it work and I’ve simply not played the game or games where it does, but let’s speculate anyway 😉

The good news is that it creates a complex situation in which there’s not an obvious dominant strategy. If all players but one refuse to take part in this sort of auction the player who does is at a significant advantage. If everyone but one takes part the one who stays out has a definite advantage. Optimal play with this sort of auction will be based on meta-knowledge of what the other players are likely to do. It’s also somewhat fertile ground for table talk and informal agreements beyond the structure of the game. Some players hate that, but a game that sets an expectation that it is to be played that way can attract players who don’t.

The bad news is mostly loaded into the difficulties of making it fun to lose these sorts of auction. Inherently the player who comes second will lose resources, potentially a lot of resources, and have nothing to show for it. It’s important this not remove them from being in a position to make meaningful decisions in the game. There’s also the danger of a run away winner where someone says “I have the most shinies, I am going to outbid anyone who tries to get this thing, so if you bid you will lose money for nothing, therefore you will not bid and I will get it for one shiny. With everything costing me one shiny it is easy for me to have the most shinies forever.” In theory a strategy like that could be broken where other players cooperate to break their strangehold on the “most shinies” position, but such cooperation is difficult where’s an intrinsic motivation to be the contributing member who contributes the least.

What would the working game look like?

I think there’d be two main ways to mitigate the problems auctions like this could cause:

The first would be to make a game in which the thing players are bidding with refreshes frequently. If you had 10 coins a turn to bid and couldn’t carry them over from turn to turn then nobody could maintain a “most shinies” position and someone who lost everything to get nothing is back in the action fairly quickly.

The second would be to make a game in which the social dynamics took centre stage and other game mechanics were supporting the making and breaking of alliances in a way that mean someone who bid second and lost *was* getting something meaningful. They were knocking down the money available to the winner which is somehow good for their alliance and good for them.

Is it a good idea?

Maybe! I don’t know. The point of the blog is to discuss design ideas, see what it inspires and get used to thinking of mechanics in different ways. In that it’s been an interesting topic to divert onto for a few minutes 🙂

Space Food Truck

Recently I’ve come across a computer game that I wanted to talk about, becuase it’s basically a deckbuilding board game with sprinkles and has some interesting ideas. It’s called Space Food Truck.

The Game

The gameplay will be very familiar to anyone who’s ever played a deckbuilder:

The game is cooperative and each turn starts with a random event (usually but not always bad, escalating in severity as the game goes on). Once that’s done you play all of the cards in your hand. These let you do things related to your job, which in aggregate will let you fly your truck around, pick up ingredients, cook them and deliver them to the target planets winning the game. Then you discard all of the cards you have left over and buy one or more cards to add to your discard pile. Finally you draw a new hand, shuffling your discard pile if your deck has expired.

So far, so standard. So why bring it up?

The devil is in the details. This game uses a host of different mechanics, some of which I’m wouldn’t usually to think of as a deckbuilding mechanics, in interesting ways. There’s nothing you won’t have seen before in this game, but there are times that the mix of mechanics generates something more than the sum of it’s parts. This is a good goal for any designer, so I’d like to explore them by way of looking at how using a mechanic in a new context can generate new gameplay in the hopes of inspiring ideas for other genres.


All cards have a power value. The most important cards in your deck will scale based on how much power you choose to supply them with. For example if you play a repair card you might need to discard one power for each thing you want to repair. But of course a card that’s discarded for power can’t be played.

I’ve always been a big fan of “choose cards to discard to be allowed to play other cards” as a mechanic, meaningful choices are central to good gameplay and this is the sort of mechanic that helps to develop it. It also provides extra delineation between cards, allowing the presentation of choices between buying cards worth lots of coins, cards which have good abilities or cards which provide a lot of power.

It’s not used to its fullest possible effect here though. It could have been a means to flatten out some of the luck of the draw and make player skill more central. In a lot of games a hand of all of one type of card then a hand of all of another is often less desirable than getting a good mix. If important action cards had high power values then drawing them all together would lead to thoughts like “Huh, all three repair cards, well we’re not repairing for a few turns after this while I draw through the rest of the deck – but at least this turn I can fix everything!” Instead important actions have been assigned low, or even 0 power, which means the mechanic instead exacerbates the luck of the draw.

That seems like a missed opportunity.


The variation between playable roles is quite intensive and well implemented.

The captain is the only person who can move the ship and restock the supply. If the ship doesn’t move there’ll be no new cards to buy at the end of the turn – if you don’t buy a card you get a useless leftovers card that bloats your deck to no advantage.

The scientist is the only person who can improve people’s ability to do their core job. Everyone can buy cards each turn, but these come from a common pool of generic cards. The captain can never buy an extra “Fly the ship this turn” card, which is bad news since decks will inevitably get larger as the game goes on so the main jobs get done less and less consistently. The scientist’s research power adds new core cards to everyone’s decks, letting them do their main jobs more effectively.

The chef is the only person who can cook meals, necessary to win the game. However that’s a skill that’s only needed on a few turns, so the designers needed to give him a second strength. What they’ve settled on is an exceptional ability to destroy cards in both their own and other people’s decks. Even the most expensive “destroy card” options in the generic deck are typically less good than the chef’s starting options.

The engineer is there to make sure you don’t lose. Events will damage the ship and crew, lowering its capabilities in various ways. If it’s hull gets damaged enough it explodes and everyone loses. On the other hand different sorts of damage might cause people to draw fewer cards each turn or be unable to play some actions. The engineer decides what to fix in what order.

Given that the distinction between the roles is generated by approximately 4 cards in a 10 card initial deck (plus whatever the scientist researches over the course of the game) the roles feel very different. They all involve some sort of meaningful decision – whether it’s research priorities, which deck to burn cards from or where to fly – and they all wind up feeling very different to play.

Asymmetrical roles can add a lot to a game and give it some decent replayability. I suspect some of that is wasted here as I reckon the average person playing the game is probably on single player as all four roles – but as a primer on how to make asymmetric roles interesting and distinct it’s got some good ideas.


One of the defining features of a deckbuilder is how it limits which cards a player can buy to add to their deck. Dominon established the genre by selecting a limited number at the start of the game, but allow players access to all of them every turn. Others, like Legendary, have a limited supply drawn that’s refreshed after players take cards from it.

Space Food Truck takes an interesting step in treating “available options” as a resource. If the captain is willing to play a jump card and expend fuel and avoids crossing the path of places she’s visited before then she gets to draw a random selection of cards which she may add to those that are available to buy. There are a limited number of slots for these cards and it’s up to her which ones are discarded and which ones are available for the rest of the team to buy.

Having stocking the cards that people can buy from as an explicit game event opens up more meaningful choices. The actions that will lead to the most frequent restocks and doing several restocks to get the ideal mix of choices rather than “whatever was lying around” will often conflict with other desirable outcomes like “take the most efficient route” or “move primarily through systems where we know that we want what will happen to us on that planet to happen”.

This is a mechanic I’d like to see someone do more with. The decisions in Space Food Truck are often relatively trivial “everyone getting a leftovers is sufficiently bad that the jump to a new planet is worth it regardless of other consequences” or “we have fewer new cards from this planet than we have slots so we’re taking everything” aren’t tense, exciting, decisions. I’m sure a game built around playing with the stock situation some more could do a lot with it, but this shows how the potential is there to be realised.


Space Food Truck is not the best game I’ve ever played (well worth the price though) but it demonstrates the power of taking mechanics into unfamiliar territory and messing with assumptions. As much as there are ideas from this particular game I’d like to see taken forward, I think more generally it’s good to bear in mind how we don’t need to always be trying to come up with something new as much as innovative mixes for ingredients we’ve already got on the shelf.

An Unfair Game

Today we’re going to turn an assumption about what makes a good design on its head and see what shakes out. The assumption is this: An asymmetric game should be balanced such that players of equal skill have a reasonable chance to win no matter which side (or position or whatever) they start with.

Players Have Unequal Skill

This seems obvious, but it’s worth saying out loud, since it’s the reason to do the thing. In most games if one player is better they will tend to win. Sometimes this is desirable, especially in something like a competition scenario, but I’m not sure it’s always desirable.

If a parent plays with their kid should they always beat them, or else decide not to and lose by deliberate poor play?

Should the outcome of a game between an experienced player and someone they’re introducing to games for the first time be a foregone conclusion?

If people like playing together socially but are at different levels of skill is the best possible design one where one constantly beats the other?

I don’t think the answer to any of these questions is “yes”. I feel that in a lot of situations players are best served by a design if they have an exciting close game that comes down to the wire and whichever one of them most surpasses their personal best comes out on top.

I think there’s a place for an unfair game, in which positions are intentionally unbalanced. With the goal of creating a tense close game between players of different levels of experience.

What Features Does It Need?

The gameplay on the “hard” side needs to have a high skill ceiling. In principle the game is allowing one player to make up for a gulf in raw power through good play, so there needs to be a lot of delineation in how well it can be played. There’s no sense playing snakes and ladders with one player starting 10 spaces ahead, the lack of options in the game gives the player who’s behind no means to catch up.

The gameplay on the “easy” side needs to have accessible play, FOO strategies and at least a moderate skill ceiling. Let’s break those down one by one:

Accessible play is necessary because the side is intended as a means to introduce new players to games, so it shouldn’t assume too much prior experience of familiarity with different mechanics. You’d want it to be as playable as any gateway game.

FOO strategies are a part of accessible play and it’d be helpful for developing the game. Some relatively obvious strong plays offer a new player somewhere to get started and provide a baseline for what the game might expect a player in that position to do.

However it’s important that there be better approaches available. If one side is more powerful but its gameplay is just “Flip a card, do what the card says” then you’ve essentially relegated one player to running part of a solo game for the experienced player. There needs to be the genuine opportunities for subtle and clever plays from both sides for everyone to be truely involved in the game.


In discussing mechanical issues it’s easy to nudge theme over to one side and forget about it, but the funny thing is that there are loads of games that use themes more suited to this sort of situation than the games they’re actually in.

I think it’s to do with the nature of stories that people like:

A plucky group of heroes entering a dungeon filled with monsters and traps.

A single ninja infiltrating a castle packed with guards.

A small rebel group taking on the might of an Empire!

We like underdog stories. So why not make a game in which the underdogs really are underdogs? Making major mechanical decisions just to deliver on a theme can often generate a fairly flawed game, but if we’re looking at an unbalanced game anyway why not really nail a theme that’s been attempted a bunch of times but never in any fidelity?

Pulling it Together

I’ve started trying a few prototypes for this sort of game this week. My first attempt was one called HappyLand in which a director of an amusement park was trying to get his guests to have enough fun while his guests were desperately trying to leave and get on with their lives.

The visitors started with a deck allowing some short moves and obtained resources to add cards to their deck by moving through certain spaces, permitting them to work up to longer moves or to overcome obstacles that blocked movement such as walls and mascots.

The director draw two cards and played one each turn. This let her move visitors or mascots around or set them objectives (Like “ride the waterslide”) such that failing to do so added 0-value (“smile”) cards to their deck.

The visitors won by escaping the park, the director won by emptying the smile pile.

Having worked on it for a week and taking the time to write a short article about it today I’m not sure that HappyLand is the game I’m looking for, so I’m likely to dissect it for parts. Some thoughts on this prototype:

Deckbuilding is a strong mechanic for the “underdog” side. It requires a degree of planning ahead and creates a lot of tension between “needed now” and “needed ever”. It’s also somewhat subject to disruption and can help create the skill ceiling for the other side, especially if they’re adding cards do the deck that are anything other than unalloyed evil.

I’m not sure HappyLand was the right theme, but something that’s at least child friendly seems important, given age is going to be one of the main reasons for disparity of experience. It’s important the game not come across as “for kids” though, that’s going to be a fine line to walk.

The many against one aspect needs some consideration. The one vs one situation is easy enough and many vs many is going to be much harder to work with – but which way around should it be? Is a hypothetical group more likely to be a bunch of experienced players and one new player or the other way around? Is it better for several players to gang up on the strongest player or the weakest player to be thrown into a position of power over everyone else? Those seem appealing in different ways.

There’s still plenty to do with this, but it’s an idea I’d like to explore some more.


Social Deduction for Introverts


I introduced One Night Ultimate Vampire to a new group this week and we got talking about it after the fact. Some folks liked it, others didn’t, there’s one criticism I wanted to highlight and talk about:

“in large numbers the shy folks really have to fight to be heard”

I thought this was really interesting because it describes something that I think applies to the genre rather than the game.

The Problem

A social deduction game has two elements – broadly “social” and “deduction”

You have to work things out. Partly from hard facts and partly from softer information about what people say and do.

Having worked things out you need to persuade people of some truth (either the actual truth or some alternative you’ve made up based on what you’ve figured out).

If the conversation is dominated by a few voices then the quieter people don’t get to play the second half of the game. Which is obviously a bad thing.

But the incentive structure for the game is that dominating the conversation is a good thing. People will try to build several narratives and the successful player will be the one who can make their narrative the strongest (if they’ve got their deductions right).

So the challenge for the designer is structuring the game in a way that rewards players for making more space for other players to speak up. Even if they wouldn’t usually, perhaps especially if they didn’t usually.

Not a Solution: Play with nicer people who make sure everyone gets a go

I bring this up briefly because someone always suggests it for problems like these 😉

This is a solution for *players* but it’s not a solution for *designers*. You can’t print on the side of the box “Only play this if you’re nice” and expect that to work.

A player should seek a group that they enjoy.

Where it doesn’t compromise other design goals, a designer should aim to make their game enjoyable by many types of different group.

Existing Solutions: Information Dispersal

The #1 way these sorts of games handle this issue is to make sure that all of the players have different useful bits of information. The theory being that if everyone knows something critical to their teams success then everyone will have something to say in the discussion and it will be in their teams interests to make sure it’s heard. Since the average player is on the biggest team the average player should have the support of most of the table.

It’s a neat theory, but there are a few ways it can fall down.

The first is if someone’s information isn’t useful to any other person. If all you have is “I am not a baddie” that’s not worth much. You can figure things based on that, but when you say it to the table it’s meaningless because everyone claims to be not a baddie.

The second is if the group prioritises the social over the deduction. Players new to the genre very often see leading the discussion and influencing who dies as the winning strategy. Information dispersal can only work if two conditions are met:

1. The players need the information to correctly know what they need to do to win
2. The players recognise that they need that information.

In new groups there can be a tendency to try to control the conversation without finding much out, which makes the conversation controller look powerful. If they lose anyway it can be easy for a player to attribute that to “bad luck” rather than “poor deduction” and not realise that giving players with extra hidden info more of a voice would’ve lead to a win and it was the conversation control that cost them the game.

The third is where info comes out relatively trivially and discussion on how to interpret it dominates the game. If you have 20 seconds of “Everyone says their thing” and then half an hour of arguing about it – the half hour argument is most of the game. The 20 seconds of being involved hasn’t really solved the problem, because the player still sat out of a lot of the game.

Existing Solutions: Regulated Speech

Some games will make explicit rulings about when and how a player is able to speak. In the online mafia game Town of Salem if there are enough votes to hang someone everyone else’s chat is literally disabled for 20 seconds and only that person gets to speak. Then everyone makes a final yes/no vote on whether to kill them.

Giving a player a moment to speak mandated by the rules ensures that they get to make a point at the most critical juncture means a person can get a word in edgeways at the most critical juncture.

Existing Solutions: Power at a Point

In Mafia de Cuba one player is the godfather. They will ultimately decide the outcome of the game, there’s no vote, just their word for who they think is a thief.

This has an interesting effect on group dynamics, in that one person is imbued with power in a way that distorts the conversation and that person is also the person with the most to gain from getting all of the information. If the godfather says “I want to hear from Eric now” then Eric is going to get to speak. Whether he wants to or not 😉

Other games harness this to a lesser extent, giving a player the right to choose what thing is being voted on or some special power they can use during or just after the discussion – but the purpose is the same: Get someone to chair the meeting.

Existing Solutions: Parallel Discussions

Two Rooms and a Boom takes a different approach. If a big discussion with a bajillion players is going to lead to some folks not getting to sidle into the game. So what if it was broken down into a whole bunch of smaller discussions? What if there was an active advantage to having quiet 1:1 discussions that other players didn’t notice or hear?

The answer is that play becomes very different, but everyone will be getting to do something most of the time. It’s not a perfect solution in that you can still wind up with sidelined players in 2RaaB due to its other mechanics, but it does very neatly deal with this particular issue.

Potential Solutions

Because “Information dispersal” is baked in, I’m not sure how many designers have been actively considering this as an issue rather than it being a thing that gets worked out naturally by tinkering with rules in playtesting.

Heck I’ve published a social deduction game and never really consciously thought about it!

So perhaps there are untapped solutions that haven’t yet been tried. The main one that occurs to me at the moment is likely due to ongoing discussion about asymmetric resources. What would a social deduction game look like if the capacity to talk was a limited resource?

Could you have a game in which people said things like “Rose just spoke for the third time which means she can’t be a villager, I am which means this is the last thing I’m going to be able to say.” I think there’s some unexplored potential there.

I’m sure there are other solutions out there waiting to be discovered. I look forward to playing them 😀