Addressing the Kingmaker Problem in Other Games

After yesterday’s article about the kingmaker problem in Invasion, I thought it might be interesting to examine some other games and see how they avoid a similar kingmaker problem. I’ll focus on area majority games, since that’s Invasion’s style.

El Grande: This game was the inspiration for Invasion. The thing that prevents a kingmaker situation here is the simultaneous selection provided by the dial choice in the Castillo; no one player gets to affect the board last.

Evo: This game provided the inspiration for Invasion’s Will Smith rule. The rounds are tracked with a marker that also shifts the weather. After a certain point, a die is rolled after every turn. If it rolls high enough–against a threshold that gets lower and lower–the meteor hits, ending the game.

Chaos in the Old World: For a game with so much fighting, Chaos actually has a simple underlying area-majority mechanic for scoring Domination. But because of the uncertainty provided by battle dice, where the destructive power of even the smallest fighter is potentially unlimited, each player just has to set up their forces to increase their chances and pray for the best.

Carcassone: One of the important features of the other games I listed is that players have fairly free rein to attack and interfere with any other player at any time. By contrast, in Carcassone, your ability to interfere is heavily limited based on what tile you draw. You can usually improve your own position, and targetting a particular enemy is usually pretty difficult. These factors add up to no particular kingmaker problem.

Louis XIV: With its seemingly open, deterministic scoring and one last player, Louis XIV seems ripe for a kingmaker problem. However, the secret scoring bonus for crest majorities provides that tiny bit of uncertainty to make it usually more advantageous to increase your own score rather than tearing down a particular opponent.

Montana: I have to add a plug for my own design in here! Secret goals were added early to Montana and they manage to fill two important design goals at once. First, they increase the degree to which some regions are more valuable to some players than others. Second, they provide such a large chunk of the winner’s points (possibly a third or so) that attacking a “leader” before you know whether they managed to even managed to hit their secret missions is risky indeed.

I considered some kind of secret goals for Invasion, but I don’t think it’s the right choice. Invasion is supposed to be about making the best of a rapidly-fluctuating situation, not pushing towards one particular outcome.

Invasion: Rules PDF

As I promised on Twitter, here’s the latest rules for Invasion. Along with the latest rule and terminology changes, I also moved from Word to LaTeX, which changes the content in no way but makes it look super classy. (The last time I used LaTeX, I just used a text editor. I tried LyX but never got into it. This time, I used Texmaker, which is a really nice environment.)

Invasion Rules

Addressing the Kingmaker Problem in Invasion

When I wrote about my recent playtest of Invasion, the last point I mentioned was that I found myself in a kingmaker situation: I could not win myself, and I could stop either of two other players from winning, but not both. I ended up making the play that maximized my own score, but it made an unsatisfying end to what had otherwise been a tense and active game. I left this point until last not because it was least important, but because unlike the other points, I was completely unsure what I was going to do about it.

I want for players in a position like mine to be able to credibly play to maximize their score–or to narrow the gap between their score and the leader’s–without the futility of “the game is ending this turn so why bother?” The first way I attempted to solve this in a previous Invasion iteration was the “Will Smith Rule”: when the game would otherwise be over, roll a d6. On a 6, Will Smith’s heroics buy the humans a respite and the game lasts another turn. (Yes, the name was inspired by his role in Independence day.)

This mechanic works, but it has some flaws that kept me from ever becoming fond of it. It feels tacked-on and arbitrary, because it is. It comes as a surprise for first-time players, because it comes up only once, at the climax, and doesn’t fit the pattern of the rest of the game. It’s out of the players’ control. And every so often a game is going to drag on senselessly, boring everyone involved even when the winner is clear. Invasion is already tuned to be the “right length”; occasionally lasting for another turn is fine, but it shouldn’t happen frequently or for many turns.

To try to put the decision in player hands in a later iteration, I scrapped the Will Smith rule and created a Technology card: “Futile Heroics”. The text of this card essentially read “If this card is in your hand when the game is over, reveal it; you get 2 VP and there’s a 1/3 chance that the game lasts another round.” On one hand, this was an improvement, because it gave control to the players and didn’t require any extra rules explanation. On the other hand, it was a disappointing and low-impact card to draw, and when players avoided the Technology deck, it almost never came out at all. Also, in most situations players wouldn’t want to keep a Technology card unused in between turns. When I transformed the random-draw Technology deck into the common-pool Gadget deck, Futile Heroics didn’t make the cut, but that meant nothing was left.

My change for the next iteration is: a few of the Gadget cards now have an hourglass in the corner. (Currently, 4 out of 30.) When the game would otherwise end, you reveal one; on an hourglass, you set that card aside and play another turn. This seems a lot like the Will Smith rule, but I’m optimistic about it for a couple reasons:

  • There are enough Hourglass cards that one of them will almost certainly show up in the first two or three rounds. At that point, if there are first-time players, the rules teacher can reiterate its effect on the game end. The icons on cards act as a quiet reminder that the rule exists.
  • Because the icon is printed on existing cards, it seems like an organic part of the game. It also doesn’t require another component.
  • Because the cards are removed from the deck as they trigger, this provides a natural way to reduce the chances that the game keeps going on, without having to add any new rules or complexity.

Towards a Holy Grail (Part 3)

Here, as promised, is a short rundown of the changes I am planning for the next playtest of Walsingham. Everything is subject to change, of course, even mid-game if it ends up not working out.

  • Theme: As I’ve been mentioning, the current title is now Walsingham, after Francis Walsingham, Queen Elizabeth I’s spymaster. Conceptually, the players are merchants buying and selling secrets to win favors from the Crown–but one is secretly a spy for the Spanish! I am relatively pleased with this theme and will probably stick with it if the game ends up being fun.
  • Graphics and Terminology: The resource that the Spy wants is now known as Intrigue. Intrigue is now no longer harmful to non-Spy players; in fact, players get a small reward for stopping the Spy from winning by collecting more Intrigue. Victory Points are still Victory Points. Both of them have their own icons, which I got from game-icons.net:
    VP

    VP

    Intrigue

    Intrigue

  • Auction: The complicated simultaneous auction is replaced with a dollar auction in which non-winners pay half. The highest bidders get first choice of the rewards on offer each round. (This is the same mechanic as For Sale and similar to Steam.) This should speed the game significantly, since I think one of its largest weaknesses is likely to be an anticlimactic period in between when the Spy becomes obvious and the end of the game.
  • Secrets: Several rewards allow the players to draw 3 / keep 1 card from a Secret deck. This deck contains both VPs and Intrigue. This lets players accumulate Intrigue without it being obvious; of course, since there are plenty of VPs available, a player going for VPs could attract false suspicion. A few cards let you peek at the Secrets an opponent has collected.
  • Narrower Rewards: Previous iterations of the game featured a wide array of rewards. The only real rewards now available for bidding are VPs, Intrigue, and information. A few cards are worth extra VP or information in if you collect both cards of a set. Some of the reward chips do come in two different colors, and Loyalty cards reward players for collecting more of one than the other. This is to give the bidding some “texture” and cover for Spies to bid erratically.
  • Stinginess of Information: At the end of the game, starting with the player with the least (public) Intrigue, everyone gets to make an Accusation of the spy. An incorrect accusation carries no penalty. A correct accusation gives two rewards: first, a sizable VP award, and second, the Spy loses both VP and Intrigue. On one hand, sharing one’s information about the Spy can help everyone correctly accuse the Spy and stop the automatic win. On the other hand, being close-lipped–or lying outright–can help a player reap the successful-accusation reward alone. Since there is no penalty for being wrongly accused, players may even choose to deliberately act suspiciously to throw off rivals.

Invasion Playtest Report

Tonight was the first playtest of my post-Protospiel changes to Invasion. I was pleased that all the changes seemed to be more or less successful, although a few changes are in order.

  • The previous iteration had players choose any free region to place their Mothership in on the first round of the game. This iteration has each player choose a Portal card, which indicates a location for a mothership and two bonus ships. This narrows the choices and jump-starts the game, so I approve. This playtest version also granted some Energy based on the relative desirability of the Mothership spot. I will remove that as too complicated in the next iteration; any players sophisticated enough to care about the small difference will settle it through bidding.
  • The previous iteration used turn order cards 1-6, refreshed every turn, and used dice to distribute bonus Action Point chips between them. This iteration uses a small deck of turn order cards 1-16, some of which have bonus Action Points built in. This works well; it not only cuts down on the components but speeds the turn up slightly. Some of the cards gave bonus Energy in this iteration. That will change to a discount in Gadget cost to encourage using those and also to reduce the amount of component-shuffling.
  • I replaced the large and small washers that served as money with red and yellow plastic winks (transparent bingo chips). This worked great; players had no trouble remembering which was which, they look classier, and the game term “Energy” works much better than “Iron Points”. I replaced the pennies that indicated increased region value with small laminated paper chips with a “2″ on them (the amount they increase the value by). This worked great except that they were too small, so I’ll print some slightly larger ones for the next iteration.
  • The big new change in this iteration is that the one-shot Technology cards, purchased in one turn from the deck and used later, were replaced by Gadget cards, which are refreshed at the start of the turn and are purchased and used immediately. They also have variable costs. This ended up working very well, infusing some more “action” into the game. However, some specific changes are in order:
    • Remove all the cards that don’t directly affect the board. One gained energy, one gained VP, neither led to an interesting situation or decision.
    • Rebalance card costs or effects to account for the fact that destroying or moving enemy ships is generally more powerful than moving or deploying friendly ships. (It lets you focus more strongly; it doesn’t lead to increased casualties; it makes it easier to lean on a leader.)
  • A secondary effect of the Gadget cards is that going earlier in turn order is now better than it used to be, because it gives a better choice of Gadget cards. I already accounted for this somewhat by reducing the spread of Action Points available between high- and low-numbered cards; I’ll keep an eye because if it makes the cards too equal, I’ll need to de-equalize them again to make bidding more interesting.
  • As the last player on the last turn, I found myself in a kingmaker situation: I could stop one of two opponents from winning, but not both. This is not ideal, and the possibility of such kingmaker situations is the worst problem that has plagued Invasion since its inception. I think the game is good enough to survive with this flaw, but I’d still like to see it resolved… somehow.

Thanks to Paul Jacobs, Eric Steiger, and Carla Schober for playing!

Towards a Holy Grail (Part 2)

I put together the first prototype for the game that would become Walsingham back in 2010. The idea was less focused then. I have always loved auction games, and I married the idea of hidden roles with another idea I had been kicking around: a simultaneous auction (as in, say, Homesteaders or Cyclades) that rewards players for being outbid. (Thematically, the auction house is paying out kickbacks, I suppose.) The auction drove a set-collection game: players were rewarded for having sets of colors as well as the majority of each color.

There is plenty of discussion as to how soon to give a game its theme. My opinion is that the game gets a theme as soon as it needs it. Montana was always a simple game, and stayed Space Poker until it needed art and a name to put on the box. By contrast, Invasion was Invasion out of the gate, and the theme helped drive both the names and abilities on the Gadget cards. (Likewise, it seems pretty likely that Ra was given its theme when it needed art, while it’s impossible to imagine the mechanics of Battlestar Galactica with any other theme.) This game didn’t need a theme yet, but it did need a working title. I was using colored cubes for the set collection: gems. Some of the gems were cursed, and only one role wanted to pick those up. Who likes curses, instead of being afraid of them? Witches, I suppose. OK, the working title became Gems and Witches.

So, Gems and Witches, 2010. Curses were negative points for every player except for the player who was secretly a Witch, who got positive points. One of the late-game resources available provided an Accusation, which gave a player big points for identifying the Witch, and nullified the Witch’s huge bonus points for spending the whole game blatantly picking up curses. The auction/set collection game was pretty solid if derivative–I think 1997′s Reiner Knizia could have developed it into a 1997 hit without too much trouble–but the aspect of guessing and finding Witches didn’t work. I had hoped that the players could use the bonus-for-getting-outbid auction mechanic to leverage their knowledge, but the math never worked out, and the Accusation always ended up an anticlimax.

Boldly charging down the wrong road in 2011, I added more roles: unaligned Merchants that got better bonuses for certain colors, a Hunter that got a bigger bonus for correctly accusing the Witch, a Decoy that got bonus points for being falsely accused. I added special gems that players got big points for having exactly one of at certain intervals. The reasoning for those was to give players a cover for otherwise-erratic bidding. Again, it worked out interestingly from an auction perspective, but fell flat from the hidden-role perspective. There was just no happy medium between the Witch playing coyly (and not getting enough curse points to win) and playing openly (and drawing an obvious accusation from whoever pounced on it first).

2012 came and I tried again. First, a retitle. Montana had been published with its working title of “Space Poker”, so, ok, let’s use “Space” to mean “working-title name for a prototype game”. Honest traders don’t want contraband, because they don’t have the contacts to sell it; smugglers do, but they have to stay hidden. Hence, Space Smugglers. I ripped the set-collection and points-for-being-outbid mechanics out entirely; this isn’t the right game for them, and they took too long to explain and didn’t fit with the game. Most of the cards gave VP and Trouble, as the the curses became known, directly, although there were still some “flavored” VP to give Merchants the incentive to bid on one rather than another.

The Space Smugglers playtest was possibly the most disastrous playtest I’ve ever conducted, and I think I redeemed myself only by cutting it short after about 15 minutes when it was obvious that everything had gone to hell. Sorry, Matt, Eric, and David. What didn’t work about the game that time? Well, everything. There were a lot of hidden roles, and they all had to be explained up-front so that players’ action in the game could make sense. And then they didn’t anyway, because the auctions didn’t work. The auctions had too many moving pieces–bids going back and forth, chips gained, player pieces moving on tracks. Most of the scoring happened only at the end, and it wasn’t at all clear how any game turn related to the next. The set collection, which was the strongest part of the previous iteration, was gone with nothing to replace it. I wrote up some notes, shelved the game, and procrastinated working on it until now.

Tomorrow, I should be finished prototyping the latest changes to the game, now working-titled Walsingham. I’ll discuss what I believe I learned from the previous iteration and the changes I made (many!) as a result.

Towards a Holy Grail (Part 1)

The new prototype of Invasion is ready to try out, modulo some color printing that I need to get out to OfficeMax for. So today I’ll pick up work on what I consider a Holy Grail design, Walsingham.

The Holy Grail I am questing for here is for a hidden-role deduction game where the hidden roles are the foundation of the game, not an add-on or just another piece of the puzzle. In other words, players are playing a basic game, one that probably seems very simple–but one or more of the players has a secret goal. The business of covering up and ferreting out secret goals needs to be the point of the game; otherwise, you’re just playing an ordinary game with different VP conditions.

The reason I call this a Holy Grail is that I believe it’s a very difficult goal to achieve, and one that has never really been done to my satisfaction yet:

  • If the only consequence for having your secret goal found out is that it’s somewhat easier to work against your goals, that’s not terribly interesting. I’m used to having my goals worked against. The good news is that you can still have a fine game; it’s just that the deduction isn’t an interesting part of it. Consider, for example, Lords of Waterdeep. Usually most players’ Lord cards are pretty obvious by the end; keeping them secret just helps to throw a little bit of uncertainty into the scoring.
  • By contrast, if the penalty for having your secret goal found out is devastating, it means that having goals revealed needs to be the climax of the game. Because information should ideally drip out slowly during the course of the game, I take this to mean that the Holy Grail game needs to be fairly short–30-40 minutes or so–so it doesn’t drag on past the climax.
  • Because of the physical nature of board games, it’s very difficult to achieve any kind of fog-of-war effect: to have information available to some players but not others. About all you can do is have cards or chits that are concealed and allow some players, but not others, to look at them. In all but the simplest situations this also introduces an element of memory, which I don’t particularly care for.
  • The game should not allow enough direct spying to deduce what player is what; that’s blunt and boring. (I think this is one of the main reasons Shadow Hunters falls a bit flat.) Rather, most of the clues should come from players’ actions in the base game: what resources or opportunities you value or ignore.
  • All players should have a reason to lie or otherwise not share information freely. This adds the element of mistrust and intrigue that makes the game compelling. Consider One Night Werewolf, where the secret swapping of roles makes originally non-Werewolf players need to lie to figure out what others know; or Two Rooms and a Boom, where a huge number of the advanced roles either reward, punish, require, or forbid sharing information in different ways.

I believe, by the way, that The Resistance, Two Rooms and a Boom, and One Night Werewolf are all fine games, but these games are driven by the team and social aspect. By contrast, my Grail is free-for-all where players might well choose either to lie (to throw others off the track and reap the rewards of successful deduction alone) or tell the truth (to catch a runaway leader that can’t be brought down any other way).

Next time I’ll talk about the history of Walsingham and some of my lessons and failures along the way.

Favorite Tools and Resources

I spent a good chunk of this morning fighting with Paint Shop Pro, where I had created an old version of the Invasion map. In honor of that frustrating experience, here is a list of some of my favorite online tools and resources.

nanDECK: This an amazing tool, specialized for laying out cards. It uses its own scripting language, but the scripting is easy and the manual and tutorials are helpful. It can import data from a CSV file, create your cards according to your script, and output a PDF at the touch of a button. A mini-IDE and CSV editor are built right in. As the author’s hobby project, there are some quirks in the behavior and writing, but don’t let that stop you: this is an easy-to-use and powerful tool.

Inkscape: This is a free vector graphics application. It’s not a perfect Illustrator replacement, but it’s powerful, novice-friendly, and I can usually make it do what I need. You often hear this mentioned in the same breath as GIMP, which sells itself as a Photoshop workalike. That’s unfair: Inkscape is a fine program, while GIMP is a morass of bugs, terrible usability, bizarre nonstandard UI choices, and missing features.

Pixlr Editor: I was only recently introduced to this raster graphics tool. It’s Web- and Flash-based, but it’s amazingly full-featured, free, and quickly my new favorite free raster graphics tool. Paint.NET is fine but not quite there (plus you have to install it, so early-21st-century) and you already know my feelings on GIMP.

Game-Icons.net provides hundreds of icons useful in games, with a Creative Commons license that is attractive for prototyping. The icons available lean a bit towards fantasy and fighting. NounProject is a similar resource that contains a much broader array of icons, although they don’t share the unified graphic style of Game-Icons.net.

Incompetech Graph Paper: this bizarrely named site provides very good graph paper generators including hex grids, triangle grids, and dot grids. Gridzzly is a quicker, less full-featured tool that provides the most common papers you might want to print.

Invasion: Playtesting Anecdote

A story about the Gadget/Technology cards in Invasion from Protospiel. A playtester had just purchased a card with the following 12 words of text:

Move up to 3 of your own or enemy ships to Europe.

Playtester: “I hate this card, this is stupid. Europe isn’t scoring this turn so this is useless.”

Me: “Since Europe isn’t scoring, why don’t you move another player’s ships there?”

Playtester: “You can’t do that with this card.”

Me: “I’m pretty sure it says you can.”

Playtester: “No, it doesn’t.”

Me, now concerned that there is an error on the card: “Can I see?”

As it turns out, reading the card aloud was enough to convince the player that it could be used to get enemy ships out of the way.

This is a difficult situation. On one hand, I don’t want to confuse players, and I want to take playtester feedback seriously. On the other hand, this is not a complicated effect, and I don’t want to coddle recklessness or incuriousity. This player should be beaten by a more thoughtful opponent.

I hope that my change to Gadget cards that makes them public mitigates the effect of a player drawing a card and being frustrated because the right way to use it isn’t obvious; the other players can use the card instead and forcefully demonstrate.

Invasion: Overview of post-Protospiel Update

Project #1 for the sabbatical is updating my second-oldest prototype, Invasion, based on the feedback that I got at Protospiel. Here are the changes:

  • Rename “Technology” action cards to “Gadget” cards to reflect their single-use nature. Rename “Iron Points” to “Energy”; the old name “Iron Points” is a holdover from when VP (“Gold Points”) could also be spent. This is not a gameplay change, but it will require recreating most of the components.
  • The most common criticisms of Invasion are: some players want more direct aggression against the other players; and players complain about buyers’ remorse after purchasing a Gadget card and not seeing an immediate use for it. To address these issues together, I will rework the way Gadget cards work: Rather than being purchased and spent later, a set of Gadget cards will be available to all players, refreshing each turn. On your turn, you may pay for and use one of the Gadget cards, which is then unavailable to the other players; so there will be an advantage to going first. I may even add in free, weakened effects so that being broke does not shut players out of the cards.
  • Currently, the turn order bid works like this: The turn order cards are laid out 1-6; dice are rolled to see what cards get bonus action points. Mechanically, this works great, but it requires three sets of components: turn order cards, dice, and bonus action point chips. A Protospiel playtester suggested that I could use a slightly larger deck of turn order cards with bonus action points “built in” to the appropriate fraction of the cards, and thereby make the dice and chips unnecessary. I think this is a great idea, and the cards will need to be redone anyway (to handle the change in Gadget cards) so I’ll give it a try while I’m at it.
  • These changes will require a new board and player aids. I hope I still have enough of my old files that regenerating the board is not too hard.
  • All of these changes will require an overhaul of the rules and explainer’s script. I may go to the trouble to typeset the rules in LaTeX; I saw some of Gil Hova’s prototype rules done up in LaTeX a few months ago and they looked very sharp.