Wow. Roads and Boats, the second major release from Splotter (along with Bus) represents a Herculean effort in game design. Or, it might be better said in game system design, since the endlessly variable scenario options and wide variety of strategic options make this a game that must be played multiple times and with various approaches before you start to appreciate its depth and sophistication. Said another way, Roads and Boats is not for everyone but those who enjoy a logistical transport and development challenge will find a lot to explore in the box.

The game surface is developed from a series of hex tiles that represent five types of land and one with just water (sea), with some of the lands having rivers. The scenarios included with the game show sample setups for one through five players. The landscape chosen determines how much can be produced over time and how confrontational versus cooperative the game play will be. Since this probably makes no sense yet, let me explain what you do and how you score.

You score for the wealth you control at the games' end in the form of gold nuggets, coins, and stock bonds. The game pace is determined by building a common temple, called the 'Wonder', at the end of each round. In addition to the wealth scoring, each row of the Wonder scores points allocated among the players who helped construct that level. To control the wealth, however, means that you must be carrying it on one of six available 'transporters', three for land and three for water. This brings up one of the main features of this game: you develop the land and spend resources to do so, but you only control what you carry. The buildings, factories, and other producers that are developed are common property, and only those who supply the inputs (if needed) or get there first to take the production can use it. In addition, each land tile can only hold one building so judicious use of the limited resources and building an efficient transport network that can both move goods where needed and hold goods for protection is what this is all about.

Each player begins by choosing a starting hex and gets allocated a basic set of resources: three donkeys (the most basic land transport), five sets of boards, one pile of stones, and two geese. From these humble beginnings, players can build 16 types of buildings or structures, and in turn these can produce eight types of goods along with five transporters (which, when combined with the donkeys, create six.) Along with this, both the donkeys and the geese reproduce if given the right setting, roads and bridges are built to connect the tiles, and walls are created to protect the most desired land or structure from opponents' use.

The best way to describe the variety of development is to walk through the four steps in each round. First, production occurs. Livestock left in pairs on otherwise empty pasture produce an offspring, and each building has the opportunity to produce. Five primary buildings produce automatically each production round, even when no transporter is near and no inputs are added. The eleven secondary buildings produce only if given the proper inputs locally, and each are limited in the total production per round. For example, a Sawmill will produce two sets of boards if fed a single set of tree trunks. This can be done up to six times per turn by the same or multiple players. A Truck Factory will produce a truck if given iron and fuel, but only can be used once per round. The first player to use it gets the truck, and everyone else must wait until the next round or build a factory closer to his or her strength.

Multi-level production is necessary to win, since only gold is produced from an initial building. To turn the gold into coins, you must first build a Mint, and then feed it with fuel and gold. To get a Stock Bond, you build a Stock Exchange and then give it coins and paper. But, to get the paper requires a Paper Mill and, well, you get the idea. In order to produce anything requiring inputs, you must get the appropriate goods to the tile and risk leaving them there while you go back to get the rest (because if someone else gets there first, they can use them,) or bring everything in on your transporters.

After production, the movement phase begins. Each transporter can move a specific distance, with the Donkey being the only land transporter that can move on unpaved land (no road.) Donkeys move slowly and can't carry much, so developing a wagon or truck speeds your delivery and increases your lift capacity but each take time and available resources. In the water, you can start with a raft and upgrade to a rowboat or a steamship with the same consequences. No one player is allowed to build more than eight total transporters or more than five in any one mode (land or water).

Following movement, players build. You are only allowed to build on tiles where you have a transporter, and as said before only one building per hex. Once built, a structure cannot be removed. All buildings are built using only combinations of boards and bricks, making a good supply of each essential. Only Oil-Rigs are built at sea, though many buildings require a specific land type for construction. For example, a Quarry (that produces stones) can be built only on rocks. Certain sophisticated buildings, including water transporter factories, require first that 'research' be completed to gain the construction knowledge. Research is produced with geese and paper (the geese for the quills or the food while studying, we can't decide which,) making it necessary to breed geese and get access to a paper mill.

As the final phase, each player contributes to the building of the Wonder, and every round a neutral brick is added independent of the number of levels constructed by the players. Thus, the game will eventually end even if no one builds the Wonder, although this wouldn't happen since controlling the pace can be used to advantage plus the scoring benefit is meaningful.

The limited resources and production abilities create a natural tension for sequencing. There is always more you want to do but are just short of resources or movement. During each round, play is typically simultaneous although at times there may be conflict as to who gets to use the factory or take the resource sitting on the tile. This is handled in an odd but effective way; before each phase the players can request a change to the play sequence. The game starts by placing a praying figure for each player in front of the temple, and typically the player furthest from the temple moves first. If a change request is made, the player farthest picks the spot they want, followed in order from there. Each player can then "invoke the favor of the gods" to protest the new sequence in reverse order. This exercise adds time to the game and feels like it could be simpler, but it no doubt is effective. At the game end, the tiebreaker is won the person closest to the temple, but that does not typically outweigh being first to act if possible.

Within this vast array of opportunities, natural strategies develop to use the surrounding resources efficiently and gain access to the buildings needed through construction or transport. The land scenario helps determine how conflictive the game will be, since if each player has a reasonable amount of land to develop independently then the need to branch out to use producers built by other players is limited until later in the game. Water forms a natural boundary only until the first raft is built, at which point interaction, desired or not, starts to happen between the lands on each side of the sea. There is of course an optimum score that can be obtained with each scenario, and the one-player version of the game is really a puzzle aimed at getting the most out of the limited resource. Carefully tracking others' building is necessary in order not to duplicate efforts when you can plan for their use.

The open access of resources is something that feels strange, as it is not common in most games. You take time and resources to develop a specific factory or building and it is natural to want its production for yourself. Doing this, though, requires strategic placement and/or building defensive walls, each of which create a corresponding constraint on other activities. There are few games that replicate the delicate balance or opportunity cost of action better than this, and you will find yourself second-guessing previous actions and regretting wasting resource on that Clay Pit, or other similar frustrations.

The game pace is slow at first, but quickens as the networks develop and the development becomes clearer. One minor annoyance is that with your initial resources you must build a Woodcutter and a Sawmill, so in a game where everyone will develop individually at first it makes sense to just start with one of each and fewer raw resources. The timing of the game and the methodic logistics will turn off players who crave more action, but careful planning and networking will reward the player with abundant production and dominance over time. The 'open resources' feature is common to other games by Splotter, notably Bus (also reviewed in this issue). In that game, players construct buildings and lure passengers to the city though all players can use each.

The production of this game is well done. The game includes hundreds of good-quality counters, lots of wooden bits, and a nice plastic case to hold most of the pieces. The land and sea hex tiles are thinner than ideal, but with care they should last. The game is played by placing a Plexiglas sheet over the hexes so that roads can be drawn, and this has the added effect of protecting the paper hexes from spills or tears. For anyone who doesn't enjoy the game play, they can consider Roads and Boats as a 'Binary Arts' type puzzle with the goal being to get all of the pieces to fit in the box and have the lid fully close. This thing is packed!

Roads and Boats is like Ravel's Bolero compared with most games' 1812 Overture; it is the patient building and slow development that leads to the rousing finale. With Gerard Mulder's excellent games, Cwali's intriguing offers, and Splotter producing real substantive efforts, the Netherlands is now bringing the world game development to rival the best of Germany. Perfect time for Knizia to create the 'Merchants of Amsterdam'!