I’m posting to show off my game Glorieta with an example match (click through slide show above) and commentary (below), because I’ve gotten fond of it and because the rules alone don’t fully convey how the game unfolds.
In the game, 2 players compete to be the first to form a loop of stones on a hexagonally tiled board. This would be an impossible task except in each match, stones on board are slowly converted to neutral stones, owned by both players. As a result, the barriers to forming a loop gradually erode, and the player who most cleverly navigates the rubble completes a loop.
Completing a loop, especially a big one, feels good to me. Like I’ve successfully threaded a bunch of needles in succession. Like I’m the most precise badass on the planet. Like I’m Inigo Montoya.
In a normal connection game, you aim for a big target: a wall. Here you aim for the tail ends of chains of stones strewn about. I think the sense of “hitting a small target”, combined with having to overcome the minefield of your opponent’s short-loop threats, is where the feeling of precision comes from. There’s also something wholistically nice-feeling about completing a loop on a hex board, but I don’t know why. In addition, the many shapes and sizes of the winning loops lend a sense of variety.
The example above was a match between two beginners with experience in connection games. We allowed take-backs (not shown), which means the match didn’t end in a short loop like it often does for beginners. It ended with a big cross-board loop.
If you don’t want to go through the whole game, just look at the last 5 or 6 slides to see how the game ended. Note how the Blue player won by simultaneously making a short loop threat and a long loop threat, such that his opponent couldn’t defend both.
It’s a good example because it illustrates how the neutral stones and perimeter spaces come into play in loop formation: the winning loop here utilized 5 neutral stones and 6 perimeter spaces, as you can see by looking at the last two slides. It’s also more advanced than a normal beginner match thanks to the take-backs.
More on the Origin of Glorieta
When I play a game, generally speaking, I like it when there’s a threat of a game-ending move throughout. I like the “sweating bullets” feeling that I get under such circumstances, and the possibility of both long and short matches adds to my sense of variety.
It’s a little strange then, that I love connection games, because they generally don’t have that quality. They tend to be heavily strategic, and the threat of game-ending moves doesn’t arrive until many turns in. I like connection games for their wonderful strategy considerations, but I think I would like them even more if there were some short threats to contend with as well.
A few years back, it occurred to me that a connection game where the goal is to form a loop might be different. The reason is that loops can come in a wide range of sizes. For example, on a hexagonally tiled board, a loop may be comprised of as few as 6 stones, or it can be a many-lobed monstrosity draped across the board, or anything in between.
The existence of small loops allows for short threats, in principle. In fact, one game, Christian Freeling’s Havannah, employs them to this end, and it works beautifully.
In Havannah, loops are a side dish, not the main course. There are two other goals in that game, which tend to be the focus, and a loop is what happens when you concentrate too much on them.
So for a long time now, I’ve been trying to invent a game where loops are not only the main course, but the only course.
There’s a problem though: depending on the board you use, loops are either too easy or too hard to form, so if you try to design a straightforward loop goal connection game, you either end up with a tactics-fest or a stalemate-fest.
I’ve come up with several solutions to this problem, and Glorieta embodies one of my favorites so far.
It’s a bit more “grand” than the games I typically design. These days I usually strive to make complexity hidden so that casual players won’t realize it’s there. Glorieta isn’t like that. You’re aware you’re playing something challenging. The game tree is both large and bushy, and it takes experience to reliably see the basic tactical threats.
I’m interested in knowing what kind of first impression this game makes on you, if you look through the example game. I’m so deeply into these kinds of designs that I have no sense of what kind of impression they tend to make.
A favorite pastime among abstract game designers is to invent games entirely in our heads, without playtesting. This practice doesn’t always result in good games (to say the least), but it’s a great way to shape and test one’s understanding of fundamentals.
I do it all the time, and I thought I’d share one with my free time today. The idea comes from my desire to make abstract games with large game trees, but which feel coherent, simple, exciting, not arbitrary, and not discombobulating. I’ve no idea if the following game achieves any of those goals (except it does have a large game tree), but here goes:
The Rules
The game is for two players with white and black stones on a hex grid or a square grid (playtesting would determine which is best). One player owns the black stones, the other owns the white.
Each player has a number of small bowls in front of him, and each bowl contains stones. All the stones in a given bowl have the same integer printed on them. All the stones in the first bowl have “1″ printed on them, all the stones in the second bowl have “2″, and so on.
I don’t know how many bowls there should be, but this number is critical to the quality of the game and should be found through playtesting. More on that below.
Players take turns. Each turn consists of two actions, which must be performed in the following order:
- You MAY move a stone already on the board by a Chess Queen’s move. If your stone ends its move next to an enemy stone whose value is exactly one less than your stone, and your stone wasn’t next to that enemy stone to start its move, remove the enemy stone, and any other enemy stones in the group of which it’s a part, from the board. If a move results in multiple possible captures, the moving player must pick one to make. Note: your “1″ stones can capture the highest-numbered stones of your enemy (i.e. capture is circular). Captured stones are returned to the opponent after counting.
- You MUST place a stone from any bowl onto any empty space.
The game ends either when a player has captured X enemy stones (X is a number to be determined by playtesting, along with a consideration below), in which case he wins, or when when the board is full, in which case whoever has the largest group on the board wins (or second largest group if largest groups are same size, and so on…)
Note the two win conditions create conflicting incentives: you want to keep your groups small to avoid losing them to big captures, but you want to keep your groups big in case the board fills up and you need the largest group to win.
Let’s consider how game-play changes depending on the number of bowls a player has:
If each player has two bowls, then captures will be very common, and the game will always end when one player has captured X stones of the other.
If each player has some very large number of bowls, say 100 (unrealistic, but this is just for illustration), then neither player will ever capture enemy stones, and the board will always end full.
Somewhere between those two extremes, the game will end about half the time on one win condition and half the time on the other, which balances the conflicting incentives to make big groups and make small groups.
The number of bowls need for that depends on how you set X above (the number of stones needed to capture to win), so that gives you some freedom to keep the number of bowls reasonable if needed. I think a good target would be around 6 bowls. You can also reduce the number of bowls needed by restricting movement power to something with less range than a Queen’s move.
I’m not sure whether the win condition is the best, but a range of different win conditions are possible, so there’s some flexibility to play with there. I’ll discuss this later if it turns out, with playtesting, that an alternative win condition is better.
There is also an open question regarding what happens when you move one of your stones adjacent to an enemy stone which could capture it. Is your stone captured in that case or not? As the rules are now, the implication is that your stone is safe, because a stone can only capture an enemy stone when it moves. And for now I think it should be that way because it limits the number of captures, and therefore limits the number of bowls needed to otherwise limit capture. But not certain without playtesting.
On the heels of yesterday’s announcement of a dedicated play site for Ketchup, today we have a second announcement from a completely separate developer: an iOS version of the game is coming.
Whereas I had no involvement in the playsite, I’ve been involved on the app since they started working on it some time ago. I’d like to comment on an excerpt from the announcement:
The AI is already TOO good, (a problem we also had on For the Win) and one of the challenges we have yet to tackle is how best to make it interesting to play against at all levels of difficulty. Ideally, I want to have some kind of automatic scaling of difficulty so that it attempts to play at or just above your level, always giving you a challenge, but not making it impossible for you to win. I’m not yet sure the best way to do this, so there is probably a lot of work left in that department.
He’s not exaggerating when he says the AI is really good. I’ve played the game about 500 times now, put effort into analyzing strategy, and I basically never lose to other humans anymore, but I’ve been playing the AI in a beta version of the app and I’m lucky if I can beat it 10% of the time, even when I’m straining every last neuron. So yeah,it’s good. I’m impressed because it’s running on weak hardware and Ketchup isn’t a straightforward AI challenge.
The AI has also taught me that my understanding of the game isn’t what I thought it was, which is both humbling and thrilling.
[EDIT] Because interest in this game keeps growing, I’m close to pulling the trigger on a Kickstarter. Whether I do will be contingent on whether I can find a way to create an unusually beautiful physical set at a reasonable price point. I have some ideas about how to do it, which I may post about later, but tell me what features/design elements you’d like a physical set to have, if you have any such desires.
I love it when my games get attention, and no game of mine gets more attention than Ketchup. I woke up this morning to a pleasant surprise. Someone has developed a real-time play site for the game:
http://joshnelson.info/ketchup/
Screenshot above. Note that it fills up the whole browser window. You can view the game in in the screenshot here.
I found out about the site through this reddit thread.
It’s lovely: clear, big, beautiful interface, and quick responses. It could still use a couple of features (it’s missing the ability to resign), but a) it looks like the developer will add that stuff; and b) even without it it’s a really good implementation.
It seems like lately there’s increased interest from people/companies who wish to publish my games (and by “increased” I mean up from zero), something I never expected, because I’m an abstract game designer and abstract games aren’t exactly Call of Duty: Black Sniper Terrorist Nexus Ops War the Return, if you know what I mean.
I don’t know how far this will go, but anyway, a Portuguese company has published a nice wood version of a game I designed 7 years ago. The published version is called Produto. Actually, the published version is a collaboration, because two game-designer friends, Bill Taylor and Joao Neto, modified one of the rules and the published version includes the modification.
I invented the original version of the game way back in 2006, as a sub-game of a meta-game I invented called Mind Ninja, and it’s been available to play online as a Mind Ninja pattern at igGameCenter since around 2007, if memory serves (in that context its name is “Product War”)
It turned out to be one of the most exciting sub-games from that system, and so it turned into its own game. I don’t think I’ve ever posted about here on this blog (I’ve invented hundreds of games and I’m way too lazy to write about them all, even the good ones), although I posted about it over at rec.games.abstract in 2008 (which site, by the way, don’t even bother perusing because it has since become a wasteland ravaged by exquisitely boring flame wars)
The basic idea is simple: players take turns placing stones on a board, forming groups. When the board is full, each players calculates the product of the sizes of his two largest groups. Whoever has the larger number wins.
In the original version, you could play one stone of either color (black or white) on your turn. In the published version, (and here is Bill and Joao’s modification) you place two stones per turn, except for the first player’s first turn, whereupon she places a single stone.
The game is WAAAAY more intuitive to play than it sounds like it should be. I don’t have any precise explanation as to why – the human brain is just good at processing the patterns/variables needed to play this game well I guess. There’s almost never a need to actually calculate the score during the game, weirdly, and even when the game is done you often don’t have to actually do the score the calculation to know who won. You can just “see” it.
This version is being produced specifically to promote the Portuguese Tournament of Mathematical Games, which is a large event for young students, designed to help them cultivate thinking skills and which is the best possible use I can think of for the games I design (certainly better than their other function of giving neckbeards like me an excuse for endless mental wheel-spinning). Look at all those beautiful kids learning!
My game is going to be a part of that!
I’ve fallen asleep thinking about abstract games every night for more than a decade, and I’ve designed several hundred such games, the rules for which fill out bundles of notebooks I keep in my garage.
Having spent such a colossal portion of my waking life designing abstract games, I’ve long wished there were a way to monetize some of this work. I revisit the challenge periodically, and often end up frustrated. Most commercially successful games aren’t abstract, most abstract games remain unpublished, and most which are published struggle commercially.
Nonetheless there are exceptions. In this post I focus on two commercially successful abstract games, Blokus and Pentago, and the lessons they may offer
Blokus, in mid-play
Pentago, in mid-play
I’ve been studying these two because I see both everywhere: game stores, book stores, toy stores, big box stores, novelty shops and even drug stores carry them, across the US. Both are traditional, luckless, themeless games with simple rules, and each has sold more than a million copies, making them among the best-selling board games of any kind.
Why do these games succeed so spectacularly where others fail? Research and contemplation has led me to believe the following factors are critical:
Hustle - First, as with any product, the most critical among the many critical factors is probably old-fashioned salesmanship and hustle. Generally speaking, products don’t cut through the noise of the marketplace unless they have someone behind them willing to press and press and press their appeal to all prospective buyers. Obvious though this may be, many games don’t have that kind of support. Blokus and Pentago do, or at least did.
Brand Focus – Both Blokus and Pentago, in their rise to commercial success, were published by companies that treated each game as its own brand, and published only that game. Many publishers think of themselves as the brand, publish many games, and treat those games as extensions of the brand. I don’t think this works well to maximize the success of any particular game. Focus is critical. The company should be built around the game, not the other way around. This is how it was for Blokus and Pentago. A company called Sekkoia was formed for the sole purpose of selling Blokus, and a company called Mindtwister was formed for the sole purpose of selling Pentago. Note that things later changed for both games, but only after commercial liftoff: Blokus was sold to Mattel, and Mindtwister started publishing other games/toys, though they still seem to put far more resources into Pentago than any other game.
Short Play Time - The game should be short. Pentago takes 5 minutes to play, and Blokus takes 20. Consumers apply a different standard to abstract games than they do to other kinds of games, in this respect. I’m not sure why, but here’s my best guess: abstract games are mentally taxing, and most people only enjoy mental taxation in short bursts.
Quality Threshold – A game must satisfy some minimum level of quality: it must be enjoyable to play for some sufficiently broad group of people. I emphatically don’t believe a game must be among the best of its kind to succeed commercially. I’ve no doubt you can find many other abstract games that would beat Blokus or Pentago in head-to-head “taste tests” (assuming equally appealing sets of components; more on that below). My claim is corroborated by the abstract game ratings at boardgamegeek.com. There are many games with higher average ratings than both Blokus and especially Pentago there. This isn’t to say abstract games don’t have to be good to succeed. Both Blokus and Pentago are, in their own ways, excellent. But they aren’t the very best. The idea that quality doesn’t matter beyond a certain point is an important one for game designers like me to bear in mind. I spend most of my time trying to create the Best Game Ever Designed. But for commercial purposes, some of this focus wasted.
Form Factor! – The most overlooked item on this list. I hereby coin Bentley’s law: the more minimal an abstract game is, the more care must be put into making its physical aesthetic absolutely drool- and coffeetable-worthy. Minimalism is hard to do well, but it can be amazing (ask any Apple product designer). Both Blokus and Pentago have excellent product design. Both have eye-catching color schemes, for example, and both have pieces which snap into place on board, which makes the games look neat and ordered in play (a feature many of the most commercially successful abstract games seem to have – see Abalone or Othello for example). Of course, one constraint here is that the amazing form factor has to be achieved at a reasonable price point.
The practical corollary to Bentley’s law is simple: don’t publish an abstract game without hiring a top-notch product designer. Few who publish abstract games do so, because it’s expensive and the cost seems too risky given the commercial record of abstract games. But if Bentley’s law is true, failure is partly the result of poor product design, so the choice to skimp on design could be self-defeating. Note Pentago has been through several design revisions (I count three wood versions and two plastic versions, not including the multiplayer versions), and widespread commercial success didn’t come until after revision. How do you know if a game has the right form factor? Answer: the Coffeeshop Test. Set the game up in a coffee shop and if people play it, unprompted, you’re good to go. Otherwise go back to the drawing board.
Novel Components – a game must feature some physical components which feel novel to the average consumer. Consumers must feel like they’re getting something new, and that they’re getting some kind of toy in addition to a game. Novelty is key for getting attention (says this Neurobiologist). Blokus has clear acrylic polyominoes which snap into place, and Pentago has that neat twisting board. The average consumer has seen neither of these things in any other game, and both have a pinch of “wow” factor when you first behold them.
I believe commercial success is only possible when an abstract game has every one of the above factors working in its favor. If any one is missing, the game will never be among the best-selling board games. There may be one exception: it may be that an extraordinary form factor can overcome the need for novel components, because a beautiful form can itself act as a kind of novelty. But we should take care not to fool ourselves when our games aren’t physically novel enough.
There’s one other factor which, while not as critical as the above, probably also helps:
Familiar References – a game can be described as related to something else with which buyers are already familiar. For example, you can tell a person that Blokus is “like Tetris”, and she’ll instantly know it’s about fitting polyominoes together. Or you can tell her that Pentago is “like tic-tac-toe, except the board twists”, and she’ll know she’s in for an n-in-a-row game.
This kind of reference-to-the familiar is probably important in successfully pitching product pickers at retail chains. Most retail gatekeepers don’t know or care about games per se; they care about whether they can sell widgets. For that reason, familiar references can help them feel comfortable with a product. This is my speculation anyway.
Where does this leave me?
Only one of my games has proven the right kind of appeal: Ketchup. It’s the only one non-gamers eagerly request. It also passes the shortness test, clocking in at 15-20 minutes.
However, it’s weak on two points: the physical components aren’t novel enough, and it’s hard to describe it as related to something else with which the consumer is already familiar. I think the only circumstance under which I would consider publication is if I could create a physical form that passes the Coffeeshop Test with flying colors. I have some ideas about how to do it, but it’s a pretty stiff challenge.
Are there published games which could do better if they were promoted differently?
If my analysis above is correct, then yes. I think there are a bunch of games with commercial potential but in the interest of brevity I’ll focus on just a couple: the games of Kris Burm, from the GIPF project. His games are already commercially successful relative to most abstracts, but they haven’t reached the rarefied air of Blokus or Pentago, and I think at least one of them could. His games are short, one or two of them do very well on the Coffeeshop Test, and I they handily beat Blokus or Pentago in head-to-head “taste tests”.
I think this games have fallen short in the Hustle and Brand Focus categories. Kris Burm is a better game designer than a salesman, and no one has yet built a company focused solely on one game of his.
The game I would choose to build a company around is Yinsh. Even in its current incarnation it does well on the Coffeeshop Test (though I think it could do even better with the help of more product design – I would keep the pieces as they are but redesign the board), its rings have a novel, toy-like feel, and it can be described in terms of familiar references: “Othello crossed with Tic-Tac-Toe”.
Questions for readers
What have I gotten wrong in my analysis? What have I missed?
What abstract games, published or unpublished, have the potential to be (more) commercially successful, and why?
This isn’t necessarily the last version of Glorieta (but it could be). I’ve made great progress on the game, using (for a second time) a mechanism that I’ve not seen anywhere else, so I’m reporting on it. But the more I design games, the more I’m convinced they’re never actually done. Every ruleset is just a launchpad for another, better ruleset, always in pursuit of a platonic ideal.
Glorieta is an attempt to design a game on a hex board where the goal is simply to form a connected loop of stones. I’m convinced that the goal has great potential, and it sounds simple enough, but realizing it has proven tricky, especially because I want it to satisfy a bunch of other constraints as well (example: draws should be impossible, and the game should be highly balanced, along with other constraints I won’t bother to record here). I’ve been working on this project on and off for five years, and I have designed many, many (unpublished) games in an attempt to achieve it. Once in a while, I post an example: see here for the preceding version of Glorieta, and here for the one before that, or here for a completely different attempt to solve the problem. Or don’t bother, because at the moment I like this new one more.
Equipment: Glorieta is played on a hexhex7 board with black and yellow stones that are brown on their undersides. The board is also surrounded with a ring of black and yellow spaces, as this picture of an empty board illustrates:
Definition - Loop: a connected group of like-colored stones, and (optionally) brown-side-up stones, which completely surrounds one or more spaces, regardless of what’s in those spaces. The picture below shows a board that contains two yellow loops and two black loops. Note that loops can also include like-colored spaces that surround the edge of the board (as illustrated by the small yellow loop on the right). The smallest possible loop is six stones/colored spaces surrounding a single space.
Rules
1. The board begins empty. One player owns the yellow stones and the other owns the black. To start, Yellow places a stone on any empty, uncolored space.
2. Then each player takes six of her stones and holds them in her hand.
3. From then on, starting with Black, the players take turns. On your turn, you must either take 1 or 2 stones from your hand and place them on any empty uncolored spaces on the board, or you must flip any one of your stones on the board so that it’s brown side up. If you run out of stones in your hand, your turn is over. If the board fills completely, you must keep playing by flipping a stone on each turn.
4. You must choose to flip a stone at least once for each handful of stones. You can do so after you’ve used all the stones from your hand, but before you pick up your next hand of stones, or on any earlier turn. After you’ve used up all the stones in your hand, and flipped at least one of your own stones, pick up another hand of six stones and continue.
5. The game ends when a loop is formed and the player who owns that loop wins.
Notes
-The game will always end with a loop and there will never be a draw.
-If I’ve designed the game right, the board will rarely fill completely before a loop forms. In any case, if you’d prefer to play a shorter, more tactical game, just reduce the number of stones in each handful.
-The picture at the top of this page shows a finished game, won by black, who has a loop near the bottom of the board.
-This mechanism can be applied to any pattern-completion game (as long as empty spaces aren’t part of the pattern), and it will make that pattern inevitable. I love this. Since pattern-completion games are a huge category, it’s cool to know that if nothing else, this game shows how to make a much wider range of patterns possible as game goals. I mean, a lot of those goals probably won’t make for good games, but maybe some of them will.
-What are the novel mechanisms here? There are two: one is the use of neutral stones to make loops (and many other patterns) inevitable. The second (which is actually more just rare than truly novel) is the employment of a “hand” of stones. While “hands” of items like cards are among the oldest game mechanics, they’re rarely applied in no-luck, perfect-info, abstract games. It strikes me that there’s great scope for innovation here. “Hands” allow you to enforce variety of turn/move types, of many different kinds, in an intuitive way. For example, you can force players to make sequences of hot and cold moves, as is the case in Glorieta: flipping a stone is a cold move, but placing stones are hot moves, and in Glorieta you have a hand of stones to ensure you make a cold move on a certain percentage of your turns. The concept of “hands” is such a general and generally unused idea in abstract games that I plan explore it heavily in future games.
[edit] Fellow game designer Rey Alicia, who is a graphic designer, took pity on me and created, without solicitation or compensation, this much more beautiful layout for Glorieta:
The neutrals seem to really work for me, though I can’t say why exactly. Some graphic designer perceptual magic that’s beyond me. Endless thankyous, Rey!
