Thursday, October 31, 2013


I've been wanting to write something about Tzolk'in for quite some time. It's a game I had to be convinced to even try – my experience with CGE has been mixed, and the gears struck me as a gimmick. When I finally got around to trying it though, I liked it a lot and it's almost cracked 10 plays.

I've still been a bit conflicted, because the game has always seemed a bit vacuous.  These days I tend to prefer games with thematic focus. Tzolk'in's physical presentation is very cool, and both the art and the gears use authentic Mayan imagery. Somebody obviously did some research. It's just not clear whether this is all just flash, or if it informs the actual game systems in any way.

The subtitle of the game is "The Mayan Calendar", and Tzolk'in appropriately does seem to revolve around time and tempo. The goal is to score lots of points, which you do by gaining the favor of the gods, installing crystal skulls in Chichen Itza, and building buildings. The resources you'll use to do these things are wood, stone, gold, and food. Acquiring those resources and then turning them into points involves making efficient use of your workers, who you will commit to various tasks for various lengths of time.

Tzolk'in looks a bit like a worker-placement game, because there are workers and you place them. But it's not, at least not in the traditional sense. Here, you have 5 different cities you can assign workers to, each of which specializes in a different task (Palenque does farming, Yaxchiclan resource gathering, Tikal building & technology, Uxmal trading, and Chichen Itza crystal skull mounting). Each turn you must either place one or more workers onto the cities' starting spots on the moving gears, or remove one or more workers and do the action for the space they have advanced to. After everyone has taken a turn the big central gear rotates, driving all the other city gears and advancing all the workers on them by one spot. Crucially, you can't pass. If all your workers are on the board, you have to activate one or more of them. If none of your workers are on the board, you have to place one or more. Placing more than one worker has an increasing cost in food, the currency of the game. You can never be blocked from placing a worker in a city, but if someone else is already there you have to pay more food to get in (although you'll start on an already-advanced spot).

So, it's a game about timing. At some of the cities (resources, food, and skull-mounting), staying on the track longer is usually better, so you'd prefer to keep your worker riding for as long as possible. There are enough exceptions to the general rule that it's not really a rule – for example, some spaces are forests that need to be cleared (for wood) before they can be farmed (for food), and some technology makes early spaces very efficient – but it holds generally. Other tracks are not as straightforward. On the building & technology track, you might want either building, or technology, and those spaces are disjoint and spread out across the dial. The trade track is a true grab brag, with a focus on the third spot which gives you another worker. It all rewards planning. When you jump on a track, you'll want to know what you plan to get out of it (say, you need at least 4 corn, or the space with two technology advances). But you also want to be able to be flexible, letting a worker ride if an opportunity arises. Fortunately, you can always take an action you've already passed by (for a cost in food), so it doesn't always require extreme precision.

Less obviously from this description, but more crucial in practice, is that the requirement to always either place or activate a worker makes it important to find a rhythm. Most of the time you want as many of your workers as possible riding the gears, moving towards better action spots. Maybe you'll want one or two workers on long-term tasks, while the rest cycle on and off of shorter-duration ones. As long as you can keep productively placing and removing workers a few at a time, your costs stay low and you're getting a good number of advances every time the gears rotate. As soon as you have to activate or place larger numbers of workers, you take a hit: 4 workers are extremely expensive to place all in one go, and even 3 aren't cheap, but workers off the board aren't working. You want to maximize the amount of time your workers are being productive. My first few games I won comfortably by simply focusing on efficiently keeping my workers working cheaply and not worrying too much about where the points were ultimately going to come from.

Tzolk'in is a no-randomness, 100% open-information game and this juggling act is obviously a bit complicated.  Also, it's a CGE game, so the rules are somewhat involved – lots of different action spaces, different buildings, technology advantages, temple tracks with different payoffs, and so on. You could easily imagine this bogging down into total analysis paralysis as Dungeon Lords or Shipyard so easily can. But while it's unquestionably a detailed and thinky game, the gears don't seem to lock up, or at least not as much as you might guess. I think it's because you operate with a lot of constraints – you'll never have enough workers or enough food to do what you really want to do. If you're short food, which you often will be, there is nothing for it but to go to Palenque. With only 5 spots to put workers, the game focusses on giving you a limited number of highly distinct, high-leverage actions and there is very little of the micromanagement that so often weighs games down.

The big scoring opportunities in the game come from 3 major sources: Monuments you can build, which provide familiar endgame bonuses for having done a range of things during the game (advanced tech tracks, built specific kinds of buildings, farmed, and so on); getting Crystal Skulls and placing them in Chichen Itza; advancing in the temples; and to lesser extent, building buildings. Temple advancement is fairly diffuse and there are opportunities to do this, usually at the cost of resources, on all three of the non-resource-gathering gears. Advances on the temple tracks also feed back and give you resource bonuses twice a game. Crystal Skulls are the most focussed: there is exactly one spot on the resource gear that gives you a Skull and only a Skull, and these are useful only for mounting in Chichen Itza, which gives you points and a bump in one of the temples (and Chichen Itza does nothing else). The rest of the resources are quite flexible, being used for buildings, monuments, buying technology (which makes action spaces more productive), sacrificing to the gods, and selling for food. The net effect feels fairly well-balanced. There aren't any strategic cul-de-sacs, it's easy to feel like you're making progress, and scores will tend to be close. Unlike a lot of 100% open-information, no-luck games, this is not a game designed to make you feel like an idiot.

Even though Tzolk'in plays smoothly and feels pretty elegant once you get into the groove, there are still a fair number of details here and it's moderately tedious to have to explain the game to new players (I haven't even mentioned a fair amount of detail). So the question inevitably arises: what is it all in service of? That's the question that nags at me.

I think the answer is, it's in service of itself. Tzolk'in is thematic in the same way that Ra is thematic. Ra is an auction game with nice art and nods to the familiar Egyptian tropes: floods, pyramids, gods with funny heads, and so on. Tzolk'in is a game about scheduling and timing with nice art and nods to less-familiar Mayan tropes: calendars, corn, step pyramids, even their (as postulated by Jared Diamond) deforestation problem. Does Tzolk'in grab on to anything fundamental about Mayans, their calendar, or their civilization? Not really, and I guess my instinct is that for a game of its complexity, it should.

Ultimately I think my instincts are wrong in this case, though. Tzolk'in is a very playable, engaging, meaty euro that gives you lots of planning and resource optimization opportunities without many of the hazards that usually come with this class of games: systemic imbalance, reliance on brute-force analysis, downtime, effective player elimination. Worker placement games can be especially problematic, since they rely on figuring out how "hot" various action spaces are which requires you figure out not only what each space is worth to you, but what it is worth to every other player. This contention for spaces is a very limited element in Tzolk'in, and you virtually never have to do this sort of analytical heavy lifting. Another nice feature is that Tzolk'in is only minimally about infrastructure development, so early choices don't weigh particularly heavily and it's about making good choices for the whole game. Tzolk'in feels tight, moves at a good pace, and while it presents you with interesting tactical and strategic choices, it doesn't require or reward excessive forward-looking analysis at the expense of strategic judgement.

Tzok'in is definitely quite different in texture from the first generation of euros – it's not as mechanically tight, has a much broader canvas of options, and is quite complicated by 1995-2005 standards. However, the core elements do feel a lot like these classics: focussed, with tight pacing, good balance, and nicely presented. Because I found the gear gimmick distracting, it took me a little bit to understand that was what Tzolk'in was. Once I figured it out, I quite enjoyed it.

Saturday, October 26, 2013


On first inspection, Trains looks like a bigger Dominion rip-off than usual. It's got the basic structure we have come to know and love (draw and cycle 5 cards each turn; choose 8 random stacks of cards available to buy at the beginning of each game; copper, silver, and gold have been renamed Limited, Express, and Limited Express trains). A significant number of the action cards even are just Dominion cards rephrased and with new titles. As always, though, raw mechanics just aren't that revealing on their own. It's how mechanics are put together and calibrated to produce an effect that is interesting, and here Trains is very different from Dominion.

The idea in Trains is that you are a railway company building to connect cities and suburbs. For now they're in Japan, because that's where the designer is from; two hex grid maps for the environs of Tokyo and Osaka are on a double-sided board. You build track to connect those urban areas. Stations the players construct in those urban areas are worth points when you connect them. In order to do these things, you use two core types of cards, around which all else revolves: Rail Building and Station Expansion. There are a few other ways to get points – you can buy the equivalent of Provinces (Skyscrapers), Duchies (Towers), and Estates (Apartments), which go into your deck as deadweight – but these points are marginal in most cases. You can win with a very limited rail-building strategy with extreme card mixes, but for the most part it's about building the stations and rails.

The kicker, and the thing that makes the game hang together in interesting ways, is Waste. Anytime you do any of these things that help you win (build rails, build stations, or buy VP cards) you add one or more Waste to your deck. Waste is just a dead card – it's not worth negative points like a Curse, it doesn't have deleterious effects like a Disease, but it does cramp your style. There aren't many reasonable ways to play the game that won't involve getting rather a lot of Waste into your deck over time. How to manage your Waste is a key strategic choice, ranging from very aggressively trying to get rid of it via the action cards that allow you to recycle or landfill it to ignoring it, trying to counterbalance by purchasing lots of positive cards, and hoping it doesn't hurt you too much.

Waste is what gives Trains its unique feel. With its many expansions Dominion is pretty wide-ranging (I admit I checked out back around Prosperity), but basically it's a game about maximizing your chances of a monster hand you could use to buy expensive but high value-density Provinces. Once your deck started clicking, it accelerated rapidly to the end since the Province cards themselves weren't a significant drag. Because of the stream of Waste in Trains, instead of optimizing for the monster hand you're generally trying to make sure you can do at least something useful as often as possible. Usually that means Rail Laying or Station Expansion plus some cash. You almost always need card combinations to move forward (Rail Laying cards don't come with cash), which makes keeping your deck balanced crucial. When it comes to building rails and stations, cash is key but it's rare you'll need really large ($5+) amounts; but you will need to build a few expensive connections, so you do need to make sure your have that capability. You can do all this the with some amount of Waste in your deck, but beyond a certain point it starts to throttle you, and you need to make sure you don't drown in it. This balancing act is then further complicated and made interesting by the many action cards that allow you to do Dominion-esqe card and deck management.

The other major positive effect of Waste is that it gives the game a pulse, a push-and-pull that is absent from almost all deckbuilders. Because of the nature of card randomness, you'll usually go through bursts of activity which add a lot of Waste to your deck, and need decide whether to ignore it and press ahead, add action cards to your deck to deal with it, or spend time to remove it (you can pass your turn to remove all the Waste in your hand). As Waste is added to and removed from your deck in chunks, and as the randomness of the card flow takes its toll, each deck cycle has a different feel. Sometimes it's about improving efficiency, sometimes it's about building things, sometimes it's a balance, and sometimes it's about powering ahead in the face of declining efficiency. It's not impossible for Dominion to have this feel also with the right card mix; but in Trains, creating this ebb and flow – key to making a narrative game interesting – is an essential element of the design.

There are a number of key things that Trains wisely did not import from Dominion. There are no limits on buys and actions; buy and play as many cards as you want. Getting rid of these inorganic limits makes for more intuitive play, makes lower-cost cards more worthwhile, and gets rid of boring "+ action" and "+ buy" cards. Other than for Waste  there is no card removal, so there are no boring deck-pruning strategies and you're simply going to have to deal with eventual deck bloat.

Also gone are attack cards, replaced by the more interesting on-board competition for routes and cities. Track is owned by players but stations are neutral, so there is an incentive to leech off of other players' networks. You can never be blocked from a space, but it costs extra (both in money and Waste  to build where someone else already is. There are definitely cooperative-competitive tensions, although I think in general the high costs of building into other players' networks rewards getting there first – it's definitely not impossible to win by building up your own isolated track and station network. As always, though, the mix of cards available for purchase can change the balance between building and leeching somewhat.

Like Dominion, Trains offers you 8 random sets of cards to build your deck from as the game goes on. There are a number of game-changer cards: The Tourist Train can make winning with a small deck, minimal on-board network, and lots of VP cards viable; The Freight Train offers dramatically easier Waste disposal; Collaboration makes building into other players' networks and cities much easier. Because your deck generally has to do more fundamental things (build rails, stations, remove waste, generate cash) than other deck-builders, there seems to be a wider variety of interesting cards for any given situation, and games rarely revolve entirely around one or two big power-cards. Like all of these games, it seems likely that Trains will need expansions to maintain long-term interest once players mine out the balance implications of many cards. That having been said, I think Trains gets great mileage out of it's 30 or so different cards, and is much less susceptible to boring or degenerate mixes. I'm a hardened veteran of deck-building games, and Trains has yet to feel repetitive after 15 plays (Trains is very easy to teach and relatively short at about 45 minutes).

Trains looks like a Dominion ripoff – especially if you just read some card texts – but it most certainly is not. Clearly a lot of thought has gone into how to create a game with a very different texture. Instead of a pure race based on one powerful combo or a couple key cards at the outset, you have to make turn-to-turn and deck cycle-to-cycle decisions based on evolving game state. Instead of a relatively themeless game of interacting card powers, you have a clearly realized theme of building up infrastructure and capabilities in order to expand a physical network. I tend to like deck-building games generally, but Trains really hits a sweet spot for me, and it shows how powerfully a coherent design focus can work to breathe life into a game.

Thursday, October 24, 2013

Guns of Gettysburg

Gettysburg is a battle which has spawned a ton of wargames. Which, just from a gaming standpoint, seems odd. It was a battle where the Confederates had very little chance of achieving any meaningful military victory. It mostly featured them impaling themselves on extremely strong Union defensive positions. The Union on the other hand is usually completely reactive, as they entrench on the historical lines and throw the dice to resist Confederate frontal assaults. Most games which study the historical battle as it developed degenerate into marginally interesting slugfests.

Guns of Gettysburg brings a refreshing new perspective. Instead of trying to simulate the historical events, the design instead asks: what is the one big thing about Gettysburg – not individual elements like the fighting at Little Round Top, but an overarching thing – that was interesting? And let's do that instead. The answer Bowen Simmons has is that Gettysburg was an accidental battle. Nobody really wanted to fight there, but that's where the armies ran into each other and as reinforcements came pouring in the battle escalated. The importance of Cemetery Ridge was not inherent, it was just that's where the Union line ended up and Confederates couldn't walk away. Although the game deploys an impressive density of interesting design elements, this is the core idea that makes Guns of Gettysburg not just work, but be one of the most fascinating game designs to come along in years.

Starting from the premise that Gettysburg was a confused meeting engagement, not a set-piece, Guns of Gettysburg does three key things.

The first is to make unit arrival random, and combine this with victory conditions that move as the battle progresses. Divisions and corps from both sides show up at random times and in random order, although within a broadly historical structure. As a player, you'll know which units are arriving next on each incoming road, but not when they're going to show up. Obviously, this means that it's possible the Confederates (or Union) could have almost their entire army show up before noon on the first day, while their opponent get only token reinforcements until late in the evening. Obviously, this changes the entire face of the battle and would be a recipe for dire play balance if we were fighting for a fixed set of objectives. So, Guns of Gettysburg deploys a simple but very clever objective system. The Union line is initially defined by three objective areas which starts out near McPherson's ridge in the center, Warfield's Ridge on the left, and Culp's Hill on the right. They have to hold all three to win. Each hour on the first day that the Confederates have the preponderance of force on the map, the Union can move one of these objectives back one area, generally to more defensible terrain that is closer to the Union reinforcement areas and further from the Confederates. This is evaluated turn-by-turn, and because the exact contours of future reinforcements are unknown, you're never quite sure where the line is going to end up. This is a big source of tension for the Union player. You'll need to deploy your initial forces pretty far forward to hold the line, but you'll probably also have to withdraw as the Confederates do enjoy better reinforcement odds on the first day. Timing and executing that withdrawal is the key to surviving the initial contact as the Union.

Second is that the design recognizes the fact that Gettysburg was a battle of sharp engagements and long pauses to reorganize, extend lines, and gather strength. This causes problems for games with fixed time scales – a lot of the time not a lot is happening. Guns of Gettysburg gives us variable length turns. If one side is attacking, turns are an hour each. If both sides are holding, though, turns telescope out to as many as 5 or 6 hours. These long turns allow the player with the initiative to execute longer line-extension marches that the defender has to try to anticipate. Bundling these long turns really speeds up these periods of maneuver and focuses the game on important decisions.

Lastly is the way that the game takes a lot of complexity and designs it into the physical components – in this case the mapboard – instead of writing a lot of rules. I've always said that one of the things that made Hannibal: Rome vs. Carthage so powerful is that it took a lot of what would traditionally have been special "chrome" rules that you would have to keep in your mind all the time (Sicilian grain, Syracuse's mixed loyalties, Rome's manpower advantages) and puts them into the card deck, with dramatic improvements in the game's playability and flavor. Guns of Gettysburg does something similar, elegantly offloading complicated terrain, line of sight, and fire zone rules onto the design of the game board itself. This is not to say that the rules for ridges and fire zones are simple, or that the map is visually easy to parse; they are not, and it can definitely be confusing until you get the hang of what it's trying to do (it doesn't help that the rules are not written as clearly as one might hope). However, once you've figured out the design, it packs a ton of interesting terrain detail into a very clean system. No more stringing line-of-site threads, calculating height differentials and slopes and blind zones, counting ranges, or arguing about exactly how many tree symbols are in a hex. The map itself takes care of everything.

These are the three big things that make Guns of Gettysburg a fascinating game. What makes it a great game is that everything surrounding these big ideas is also carefully crafted. The abstract system for artillery is simple yet captures a lot (when you withdraw you give up artillery positions, for example, and it takes time to reestablish them). The rules for command posture (attack, hold, or withdraw) are simple but make for tough, authentic choices. Going into detail on all the little things the game does right would involve writing more than you likely want to read, so I'll leave them for you discover as you learn the game.

This is both the games greatest strength, and – to the extent it has one – its greatest weakness. Because Guns of Gettysburg resembles the best eurogames, designed from the ground up for a specific purpose and to create a specific effect, and not particularly beholden to any prior wargame design conventions, it can be hard to come to grips with. With 13 fairly dense pages of rules it's considerably more complicated than any eurogame, and you really need to understand all the rules and their implications to understand the whole game. Usually when I learn a new wargame like France '40 or FAB: Sicily or Crown of Roses (just to pick a few at random), I can just glance at big sections of the rules since I know how ZOC bonds or step reductions or ops cards work in general and I just need to learn the particular game's quirks. Even if you've played Napoleon's Triumph, Guns of Gettysburg is a largely unique game that has to be understood in its entirety. The game is quite intuitive in a lot of ways because of how effectively it models many things, and once you understand the premise and motivation of the design it all follows quite logically from them. Nonetheless, there are tricky details to nail down and it'll probably take you a game or two to feel like you're playing it correctly, understanding it, and bending it to your will.

It's totally worth it though. Guns of Gettysburg for me stands alongside top-of-the-line games of any genre and shows the potential of the game not just as as an interesting and engaging hobby but as a truly expressive form. For me it joins games like Modern Art, Beowulf, Hannibal: Rome vs. Carthage, Sekigahara, Rommel in the Desert, and Ashen Stars – skillfully-executed games that have great and compelling game systems that serve narrative purpose and are fully melded to the presentation of interesting, relevant, and enlightening ideas. Games like this are the reason I'm in this hobby, and a joy to discover.