Caveat: This review was written over two years' time. Everything up to the "Topology + Logistics" section was written a long time ago. I have left it mostly as is, with only minor revisions.
Additionally, the modern section was written mostly at 3:00-4:00 in the morning after feeding a 6-week old baby, and the very last bit was written this morning with him in my arms screaming in my ears.
Vouchsafe me any inconsistencies or ambiguities due to either of these causes.
Errors, syntactical and rhetorical, have surely been made as well; point them out.
A recent Ludology episode had Ryan Sturm making the claim that "Caylus is the most important game of the last 10 years". I'm not prepared to debate that, but I want to argue this:
Thesis [strong form]: Roads & Boats is the most important, most pivotal, and most interesting game in the entire history of modern resource-management Euro games.
Roads & Boats is the “missing link” in the evolutionary chain that runs from old-style topological, route-planning games like 18XX, Empire Builder, or Merchant of Venus, through heavily positional, directly interactive border-fighting games like La Città, Settlers of Catan, or Löwenherz, through the pseudo-spatial, indirectly interactive claim-jumping games like Age of Steam, Keythedral, Power Grid, or Neuland until we reach today's mostly non-spatial, nearly non-interactive resource-churning games like Caylus, Agricola, Le Havre, and all the rest. One cannot easily trace a line of ideas or mechanisms through this chain without encountering Roads & Boats.
In fact, it is hard to find a more traceable lineage of games except in instances where a clear series is identifiable [e.g. the 18xx series, __-oretto games]. While those of us who like to talk about games a lot throw around terms and phrases like “innovative mechanisms,” “mixes old ideas together in new and interesting ways,” or “creates a new genre,” these terms and phrases rarely stick. In the case of Roads & Boats, they are all true. I will argue, perhaps unprovably or even demonstrably false, that the existence of the genres of “Worker Placement”, “Cube Pushing”, and “Resource Management” as we know them now is due to the genius of Jeroen Doumen and Joris Wiersinga's seminal creation.
Roads & Boats is a large and a long game. It is also among the most mechanically pure and designedly succinct games I have encountered. There is essentially only one mechanism in the game, and the game title gives it away: Move things, by road or by boat, from Point A to Point B, where they will change into something else that you can move from Point B to Point C, and so on. The more astute among you might catch a hint in that description of the underlying, unifying theme of the game: Movement and exchange. In short, logistics: What to produce and where, what to process and where, how and when to best move raw materials to processing centers--these are the puzzles the game provides.
In the game, this takes the form of three unique elements: producers, goods, and transporters. Each producer, all 20 or so different types of them, will provide output each turn of the game in the form of goods. Goods, in turn, take almost as many different forms, from raw materials [clay, wood, fuel, gold, iron, &cetera] to processed or semi-processed materials [stone, lumber, paper, coins, &cetera] to finished goods [stock certificates, new trucks and ships and other transporters, bricks toward construction of the wonder, &cetera]. [I have written out “&cetera” in each category, because the expansion, by that name, adds about 50% more stuff in each category.]
Raw-material producers simply provide their output for free; other producers require an input good, in the form of a raw or semi-processed material from a different producer, before providing a processed or finished good. So, you may have the woodcutter or the clay pit piling up fallen trees and lumps of misshapen clay over in one area of the map, and the sawmill and brick factory sitting idle waiting for tree limbs and clay to come in to be transformed into something more usable. You also have those anxious folks at the paper mill, the mint, or the wagon factory waiting for their respective inputs. Stuff in one place needs to get to somewhere else, and here's where you, the player, come in!
In Roads & Boats, you as a player neither own goods nor the means to produce or refine those goods. All you have to your name is a transporter fleet. Your various transporters are how you will pick up output goods from one producer and deliver them as inputs to another, then turn around and take those outputs to yet another producer, and so on. Any goods not held on one of your transporters are fair game for other players to swoop in and pick up, and any surfeit input goods to one of the secondary producers in the game [due to two or more players bringing raw materials in for refining] will just have to sit on the truck 'til next turn while the factory services that no good lout that got there first.
Surprisingly, for such a long and involved game, there are extraordinarily few rules and even fewer rules exceptions or ambiguities. Every turn, all the factories and producers will supply output goods if they have the requisite inputs [some require no input]. After that, you will be able to load up goods on your transporters if you have them positioned on the tiles with the goods [each transporter can hold a certain amount of goods]. Then, you will move your transporters along the roads and rivers [each can move a certain number of spaces]. Finally, you will use the goods on your transporters to construct new buildings so that in future turns you can refine even further your finished goods [eventually for points!].
It is incredibly intuitive, and hearkens back to something simple like playing with Tonka trucks as a kid. If the game had pieces sufficiently elaborate enough, it would almost literally require no rules. There are a few wrinkles to learn, such as the process for moving goods on water [you've got to dock and load the boats!], the ability to build walls and additional roads, and the last big element of the game... the construction of the wonder, which serves as both a time-keeper and a way to earn points. All of the real meat of the game is contained on the player aid card, which lays out all of the production functions [inputs -> outputs] and carrying capacities / distances of your various types of transporters.
So, what is the meat of the game? Simply put, it is finding the quickest way to navigate the twisty, turny production functions to get from a few measly sticks and bricks to a hoard of gleaming coins and a monument that celebrates your excellent logistical skills. Of the dozen or so goods in the game, only 3 earn you any points. These 3 “wealth” goods are related, and can be progressively refined into each other, such that you double your score by reaching the 2nd level and triple it by reaching the 3rd. 2 gold is worth 20pts alone [10pts each], but you can convert 2 gold -> 1 coin = 40pts, or, in turn, 2 coins -> 1 stock = 120 pts. 4 gold [= 40pts] -> 2 coins [= 80pts—doubled] -> 1 stock [= 120 pts—tripled].
The only other way to get points [outside of &cetera] is to build in the wonder, which is kind of an upside-down pyramid shape of layered rows of bricks. Each brick you build in a row costs you 1 of any good [2 later], and the most points you can get per row is 10pts [you typically have to share these 10pts with other players]. So, if you're keeping score as you follow along, you'll note that a row in the wonder is worth as much as one piece of gold, but there's no way to “double” a wonder row. These two point sources [wealth and wonder points] are in conflict with one another, both in terms of resource needs [you can't put resources into refining gold if you're making bricks out of them!] and game timing.
Suffice to say that a player who is throwing a bunch of bricks into the wonder is going to really make life difficult for a player [maybe themselves!] who is hoping to convert a bunch of gold into coins and eventually into stock certificates. I will speak in much greater detail about the wonder later, but it's important to know now [before we start comparing games] that there are two ways to earn points in the game—refine resources [gold -> coins -> stock] and consume resources [throw it all into the wonder]—and that the two ways of earning points are in conflict both in terms of individual player resource use and in terms of inter-player relationships through game timing and majority scoring.
In the game, your crucial victory-related decisions come down to whether to focus your infrastructure and logistical network on amassing wealth or on accomplishing wonders of architecture. Any winning strategy will consist of elements of both, but the two goals are in tension, and require significantly different types of infrastructure which are also in tension should you try to build both simultaneously. The key to winning is creating an infrastructure that supports one of these goals very well and has the amicable by-product of creating ancillary pathways and back-alleys for the other. Lest you think there are “only two” strategies [wealth or wonder], let's start to talk about strategy in the game.
To get to stock certificates, you have to bring raw materials through a grand total of 6 transformations. First, you need lumber and stones to build the mines that provide raw gold. To get lumber, you first must get tree trunks, from the Woodcutter producer. You bring the trunks to the Sawmill and get your sticks. To get stones to build, you either need a Quarry, or some raw clay to put into the Stone Factory. After you've got the Mine, you'll start pulling out gold. To get coins, you need gold and fuel. Fuel can come from an Oil Rig, but you have to research that [takes geese and paper; paper takes wood], or from a Coal Burner [put in wood, any kind, take out fuel]. Finally, to get stock, you need coins and paper.
So, trunks to sticks and clay to bricks [two transformations, four buildings], then you can build a Mine and get some gold. Then, you need either geese+paper for the Oil Rig [just to get the right to build it; you still need more sticks and bricks to actually make the thing] or a Coal Burner to throw wood into; either way, this is one more building and transformation [wood->paper, or wood->fuel]. Take the gold and the fuel, throw them in the Mint, and you get some shiny coins [yet another building and transformation]. Finally, you take some leftover paper [transformation] and write out a promissory note against your wealth [coins] to get your first stock certificate [final transformation and building].
To sum it up, that's 7 or 8 buildings [Woodcutter, Sawmill, Clay Pit + Stone Factory or just Quarry, Oil Rig or Coal Burner, Mint, Stock Exchange, and a Paper Mill—needed for the stock certificate transformation, at least, and also possibly for the Oil Rig research if you go that route] at a bare minimum [usually you'll have multiples of some, or both of one step—Oil Rig AND Coal Burner, for instance—to maximize output of a crucial resource]. Each building has a construction cost in addition to its turn-by-turn resource input demands for its production function. This is also a minimum of 6 or 7 transformations / productions to get from sticks and stones to stocks, the ultimate status symbol.
“So what?”, you may be asking. The “so what” is that the game time is limited to a set number of turns, and can be shortened [but never lengthened]. Each turn only provides you with one transformation as you follow each embodied resource through the supply chain. That is, you can't transform a bunch of gold to coins and those same coins to stock on one turn, though you can further transform previously transformed gold—in the form of coins, presently—to stock while you're in the middle of transforming different pieces of gold into new coins, to be transformed themselves, in turn, into stocks later. Your goal is to shove embodied resources through subsequent refinement steps as fast as you can manage.
To do this effectively, you need to stack the different stages of production on top of each other as efficiently as possible. Once you've got some coins and are shipping them off to the Stock Exchange, you can't just forget about your Mint; you need to immediately gear up production for additional coins. This requires you to work all the way back through the supply chain on every turn to maximize the mobility of your goods to get everything to where it needs to be on time. And, you need to do this simultaneously at all levels of the supply chain. You will have, at any given time, goods in every phase of production on your transporters, and the puzzle is to get all of them where they need to go ASAP.
Strategy-wise, the game space positively explodes as the game progresses. You start with maybe two or three producers, a simple little network of donkeys and roads, casually bringing trunks from the forest to the Sawmill for refining. About a third of the way through, you're to the point where you begin to wonder if your feeble brain can even handle the amount of information you are supposed to be processing. By the end, you will have a dozen or so producers, each with goods piling up that need to go to at least two different places each, and half a dozen or more transporters with which to attempt the job. Leaving aside the actual logistics, the sheer math is scary. “Snowball” is an understatement.
To make it all work, you have to plan out a leverageable spatial layout for the producer network you'll be using and then make efficient use of—surprise!—roads and boats to get the goods physically from one producer to the next in their chain. There are a number of ways to do this, but it goes back to the basic wealth vs. wonder dilemma: you can only contribute to the wonder in one hex [we haven't even talked about the board yet!], whereas a wealth production chain requires at least 6. A heavy wonder builder will create networks that can rapidly push goods back to that one home hex; a heavy wealth builder will create networks that can efficiently move goods through a mostly linear chain. Clump vs chain.
Somewhere along the way, you'll find out that you need to do everything if you want to stay competitive. The key is to leverage your network so that one transporter can do the work of two, or more, picking up goods from one place and simultaneously delivering them to two separate refining facilities in one turn [drop one off, move again, drop the other off]—or the reverse, taking goods from two places to the one place they need to go [pick up, move, pick up, move], or even some combination of the two. You have a second, auxiliary choice to make here; you can build similar producers close together, or you can upgrade your transporters to move further and carry more, or some combination.
Wealth or wonder? The answer is “Both”, but to what extent? Here's where things get interesting. You should already have some idea that the process of getting the big point-scoring wealth goods in the game [stocks] is lengthy and difficult. I'll tell you now, because it's beginning to be relevant and meaningful, but there are at most 33 turns in the game. I don't know what the theoretical minimum is, but common knowledge [and personal experience] seems to be that you'll be making your first stock [if you're lucky] somewhere around turn 25-28. It would not be uncommon for a beginner to not make a single stock. Even one is an accomplishment. Multiple stocks is probably a game-winning blow.
“What about the wonder?” Here's what: It shortens the game and requires no transformation of goods through the supply chain. Why is this important [and genius, I think]? As an example, in the 2p game, if one player builds to the wonder uncontested [the other player does not build at all], they can rack up a total of 120pts [in games of higher player counts, there are slightly more wonder points to go around]—exactly as much as a stock—before the game ends. The wonder can shorten the game if the players build it very quickly, perhaps by as much as 4 or 5 turns or more. This can be the difference, easily, in the wonder-builder's opponent getting to convert their coins to that stock, or even their gold to coins.
What did I say earlier, that the wonder is both a source of points and a time-keeper? Not to get into too much detail, but the way this works is that there are 33 “neutral” bricks for the wonder; one goes on each turn, in addition to any player bricks. There are spaces on the wonder marked with a player count [like the turn track in Genoa, Age of Steam, et al] that will not be reached before the neutral bricks run out if no players build to the wonder, but might be reached if the players contribute heavily. Depending on exactly how many bricks the players between them build, the game can go the full 33 turns, or can be shortened considerably [if 2p each build just 1 brick per turn, the game will last only 21 turns].
Wonder construction requires no special care in daisy chaining production centers together aside from getting as many resources as possible to your home hex [where the wonder may be built] each turn. This can be intentional or unintentional, but is [as most things in the game] usually a sort of mix. If your network starts to look ugly, you might want to consider switching gears and try to end the game early before your more efficient opponent gets his engine running and can start churning out stock certificates. If you have excess resources for your beautifully planned supply chain, and plenty of time to run it through a few cycles, you can drop them off at the wonder for additional points.
Winning the game depends heavily on controlling and predicting exactly how long the game will last and making sure that your production goals and capabilities are aligned accordingly. If you get in a tight spot, and are on the high-end of the turn 25-28 range for getting your first stock [it's certainly possible not to get a stock until turn 30 or higher], you either need to have a backup plan in place to churn out points from the wonder or be watching to make sure your opponent isn't going to bum-rush you into an end-game that comes out of the blue when you weren't expecting it, long before your sloppy supply chain job can materialize into anything even remotely resembling wealth or the pursuit of happiness.
Why the wonder is genius is that it provides a substantive way for a player falling behind in the stock race to put pressure on the leaders of that race, without resorting to awkward catch-up or slow-down mechanisms that are so often used in newer games to prevent “runaway leaders”. The wonder is not a “gimme”; a player who ignores wealth cannot win with the wonder alone. What the wonder serves as is instead a sort of antitrust / regulatory compliance board. If everyone is in the competitive hunt, it can be ignored; if a clear leader begins to emerge, the less efficient player can start pushing the wonder and the leader will have to respond, cutting into their efficiency, or risk an early game end and no wonder points.
Topology + Logistics
One of the most interesting features of Roads & Boats is the difficulty of planning your network of roads, which you physically draw on the board in the manner of something like Crayon Rails or Dampfross, so as to make the most logistical success from your limited transport fleet. Every tile on the board can only hold one production building, so to navigate your way through the extraordinarily long production chains previously described, you'll have to literally navigate your transport fleet through a system of interconnected stops. Traditionally, in gaming, we'd call this “route-planning” or “network-building”; in the larger world, we call this topology, the mathematical study of the shape of systems.
Fair warning: I am not a topologist. I do know, at least, that the discipline began with Euler's exposition of the famous “Seven Bridges of Konigsberg” problem. There are seven bridges in the city that connect two large islands and the two banks of the mainland on either side of a widely forked river in which the islands sit. The task, as defined by Euler, is to define a path through the city such that every bridge is crossed once and only once. It turns out, due to the specific configuration of islands and bridges, that this feat is both practically and [this was Euler's spark that birthed the field] mathematically provably impossible. Go ahead and prove it to yourself by attempting to trace such a path on the map below.
From Euler's inspiration, a massively important mathematic discipline was built up that presently is applied in things as diverse as protein folding, cosmological physics, and computer network design. In no small part, the reason the Internet works as well as it does is traceable to Euler. For our purposes, the aspects of topology that are most interesting are the simplest, and the ones closest to Euler's original problem [thank God, since I don't understand competing space-time structures]. In basic topologies like this, there are “nodes” [islands] and “links” [bridges], and various paths that can be traced. “Nodes” and “links” are also called “vertices” and “edges” because a topological graph may be conceived of as a polygon.
One can imagine an infinite number of different, and increasingly complex, possible configurations of nodes and links [islands and bridge]. What if there were a bunch of islands in a row and we built bridges between each successive pair? What if we also made a massive bridge at the end of the row that went back to the very first island? What if we built a single long bridge alongside the row of islands, like a highway, and had “exits” so we could stop off at whatever destination we liked? What if we worked for the government and had the bright idea to “create jobs” by contracting bridges to be built between every possible pair of islands, adjacent or not? These questions represent different topological graphs.
Each possible system of bridges we might build between our imagined islands has its own unique topological structure, and its own unique efficiencies for path-tracing. Take a look at the “ring” and the “star” shapes. If the goal is to be able to move quickly [as few “bridge crossings” as possible] through every one of the nodes, the “ring” is the clear winner; we can start at any node and travel through all of the others in only 5 crossings. To accomplish the same task under the “star” structure, we'd need at least 8 crossings [if we started at any of the outer nodes], and maybe as many as 9 [if we started at the central node]. This doesn't mean, of course, that there is no benefit or comparative advantage to the star.
What the star does well, comparatively, is to minimize the distance between any given pair of nodes [and particularly pairs that involve the central node]. Start at the bottom-left corner of the “ring” and “star” and label each node, going clockwise, with a letter [bottom-left = A, left = B, etc; the central node in the star is F]. Now count up the number of crossings needed to get from, say, A to D in either structure. In the ring, it takes 3 crossings [as it would for B to E or C to F]. In the star, however, it only takes you 2 crossings to get from A to D. More importantly, the distance between any of A, B, C, D, or E and the central node F is only 1 crossing under the star, whereas it is anywhere from 1-3 with the ring.
Even more efficient at connecting nodes is the “fully connected” structure. Any given pair is only 1 crossing apart! This added efficiency might be redundant, though, and so this structure isn't always preferable [especially if there is a cost or even a negative value to building the extra “bridges”]. Let's put all this in some real-world terms now, to see if we can make sense of these competing structures. One good example of a real-world “ring” is a traffic circle / roundabout. From any starting point, the distance to your desired ending point is relatively low [the fact you have to travel in a particular direction is an important consideration in topology, but some “magic” roundabouts even remove this restriction].
Why not just connect every pair of entrances and exits directly in a traffic situation where we might otherwise use a roundabout? The answer should be obvious. Does this mean the roundabout / ring structure is always desired? Of course not. A real-world example of a competing structure, the “star”, is in the design of hospital nursing wings. Typically, the patient rooms are distributed in roughly a star-like shape around the central “node” that is the nurses' station. The reason for this is clear, and is the chief benefit of the star: it minimizes the distance between the outer nodes [patients] and the central node [nurses]. We don't need to “fully connect” this graph, because patients don't interact with other patients.
Empire Builder and Dampfross challenged players to build a hugely interconnected network of cities [nodes] such that they could trace a path, and continue tracing it, between successive randomly selected pairs of cities in as little distance as possible. The networks ended up extremely bushy and complicated, looking like some combination of the “bus” with the “tree” or “mesh” shapes. The “bus” baseline serves the purpose in-game of providing the option to begin tracing a path not knowing where it will end, and still end up with a minimal overall travel distance. The “tree” or “mesh” secondary shapes allow a wide variety of overlapping paths, which helps since the game gives players three pairs of nodes to work on at a time.
What the “bus”, or “bus / tree” and “bus / mesh” hybrids, do not do well is prevent back-tracking, which is a necessary evil [and time-waster] often encountered in Empire Builder, where the “holy grail” of card draws are ones that let you string pick-up and delivery locations back-to-back so that your inevitable back-tracking [retracing the same path in the opposite direction, East Coast to West Coast, then West to East] allows you to actually make progress in the game [picking up something from the nurses' station on the way from Patient 1 to Patient 2, since you have to stop off there anyway]. The game is won, race-style, purely by minimizing back-tracking and maximizing efficiency when you do have to reverse direction.
18XX games present a more restrictive path-tracing challenge. The goal is to connect as many as of the most valuable cities as possible, and then trace as many paths as possible between them, but the game prevents you from back-tracking or even re-using any link [like Euler's islands]; points are scored every time an “island” is visited. The way to maximize path-tracing points, then, is to start and end as many paths as possible at the higher valued cities, using every possible path into or out of them that is available. Networks tend to look something like a “star” or multiple “stars” with a “tree” or “mesh” connecting them. The benefit is that players can trace many non-overlapping paths using high-valued cities as starting and ending points.
Where the 18XX games demonstrate their notorious meanness in the path-tracing portion of the game is in allowing players to interrupt the connecting “tree” or “mesh” between an opponents' high-value “stars”. This forces players to use the “star” centers in the middle of their paths rather than as end-points, which reduces points dramatically if the center is a high-value city, since players can count a city every time it serves as a starting or ending point of differently traced paths, but only once for being in the middle of a path. It also prevents players from tracing a path from one high-value “star” center to another elsewhere on the board, and especially [in the most extreme cases] from being able to do this twice with two paths.
One of the earliest influences on Roads & Boats were surely these heavily network-focused games of efficient path-tracing. Roads & Boats even carries over the literal “drawing paths on the board” mechanism from Empire Builder and Dampfross. The goals and efficiencies particular to Roads & Boats are more akin to Empire Builder as well, but the Splotter design team have repeatedly declared and demonstrated that 18XX are among the games that most inspire and influence them so there is at least a subconscious influence from that source, too. The thing that sets Roads & Boats apart, and is its chief mechanical innovation, is that path-tracing along the network doesn't have an immediate impact on scoring.
From the idea of route-building / path-tracing as an end in itself [rewarded immediately with VP gains], Roads & Boats developed a richer experience that used topological efficiency to power a resource management game. This particular combination, resulting in a type of supply chain management feel, hasn't really been replicated since [there's one exception I know of I'll mention briefly in a later section], but certain characteristics and niceties necessary to pull it off have been used and reused in umpteen-thousand resource management games that followed. We'll look at the resource management aspect in detail later, but let's focus now on the purely topological concerns in Roads & Boats. A typical late-game board might look like this:
From the pictures, a number of topological patterns for path-tracing are apparent. Much of these players' networks are “fully connected”, with the outskirts following primarily a branching design around the connected central core, like a sort of “super-star”. These patterns aren't accidental, but as in Empire Builder and 18XX are the players' attempt to tackle the particular path-tracing problems presented by the game. There is a relatively high cost, in resources and especially action efficiency, to correcting poor network design at a later point in time, and the game's central challenge is making the right decisions in designing your production network. This entails two elements: 1) Where to build production buildings in relation to each other; 2) How best to connect those buildings.
One of the things that makes Roads & Boats' topological game tick is that there are two contrasting resource models at play in the game, which each require different network shapes to function efficiently. The primary model, the long conversion chain running from mines to gold to coins to stock certificates, requires a lot of back-and-forth path-tracing between all the different producers. This would usually be best accomplished by a network where nodes are connected sequentially / serially, like a bus or a ring, but the game doesn't simply move resources through the chain assembly line style. Instead, there are frequent required injections of supporting resources [fuel to smelt gold into coins, paper to print stock certificates on] that necessitate something more like a mesh or a tree.
The other resource utilization model in the game is concerned with infrastructure, building the factories and refineries needed to make the production chain happen. This model is also tasked with building the actual paths to be traced along that production chain. Under this second model, there are just two primary resources to contend with: wood and stone. Every production building in the game costs some combination of wood and stone to build, most secondary producers require wood as an input, and every road built costs stone. There are only so many places where a player can pick up wood or stone, and the challenge is to get the resources to increasingly far-reaching destinations efficiently. This is best accomplished by positioning your wood / stone producers centrally, as in the star network shape.
Once you start trying to play these two resource games simultaneously [and you have to], the demands on your network-building and path-tracing efficiency increase tenfold. This is further complicated by the many sub-games you need to play: Getting the raw materials of timber and clay to the woodcutter and stone factory, increasing the size and capacity of your transport fleet, staying up-to-date in the research game so you have access to the advanced producers you need, and making sure you don't fall behind in wonder construction. All of these mini-games have their own topological efficiencies to concern yourself with. Getting all your
ducks geese in a row and making sure the trains ferries run on time are extraordinarily puzzling challenges unmatched in difficulty in the path-tracing genre.
Terrain + Territory
When Settlers of Catan appeared on the scene in 1995, the primary impact it had on the terrain of the hobby was to popularize the idea that terrain on a map could provide resources that could be used elsewhere in the game to do things. There had been games before where terrain and your position on the map interacted directly, of course, in the form of wargames where terrain affects movement or defense or whatnot, or in the form of things like Kings & Things or Titan where terrain affects what creatures you're able to summon in an area. There had also been resource management games before, like the old AH Civilization, but in Civ your resource collection is only based on your overall presence on the board and not on your position relative to any particular terrain. Settlers did something different.
Nearer to the Settlers model of resource collection is the progenitor of its die-rolling mechanism, McMulti / Crude, but your resource collection is only determined by what you have built and not where you have built it relative to terrain. A few obscure titles from the old AH are variously near to Settlers' resource system: In Outdoor Survival you gain resources necessary for the eponymous survival by moving over different terrain; in New World there is a blind-hex variant where the type of terrain determines the general productivity of a region; in Source Of The Nile your discoveries may depend on the terrain you're in. Faidutti's Valley Of The Mammoths is sort of a cross between Kings & Things and Outdoor Survival; there's only one resource, food, but terrain affects how you collect it and whether it spoils. The old SPI title, After The Holocaust, has hexes which provide various resources if controlled.
Where we find the nearest analogue to Settlers, however, and probably a joint inspiration of both Settlers and Roads & Boats [which it shares other mechanisms with] is in the forgotten Eon [maker of Cosmic Encounter and Dune] game, Borderlands. At the start of a game of Borderlands, which is essentially a Diplomacy-style game of world-domination, resource production chits are distributed pseudo-randomly across the board and determine what resources [timber, coal, iron, gold, horses], if any, are available in each territory. Every turn, as in Roads & Boats, a resource chit is generated by these producers, and players controlling the territories can use those resources to build weapons, boats, or cities [which are needed to win the game]. They can also use horses, Roads & Boats style, to move goods.
Nearly as far  behind Settlers as we are beyond it, Borderlands implemented the idea of positional control on a map as a determinant of what resources were available to players in a resource management game. Borderlands was more focused on territorial control than on resource management, but its resource concept is something that was developed in Settlers and many future games, all using different methods of on-board positional play to determine who would get what resources. The heart of Settlers is choosing the correct locations on the board to build your settlements so as to get the best resource production, and then making the best positional attacks to expand your own territory and block off everyone else's. It may have had the first incarnation of the blocking interaction that Euros are known for.
Worth talking about here is La Città, which, though published later than Roads & Boats, has a history which pegs its development as taking place concurrently with Settlers in the early 1990s. The first prototype was created in 1992, while later prototype images show a Settlers like hex grid with play on the vertices [and Settlers roads on the hex sides!]. In its final incarnation, La Citta gives players resources based on what their budding cities are next to [either gold or wheat], and then allows players to expand their territory and eventually take over population from neighboring cities. Once again, the interaction is about highly aggressive positional play and siphoning off opportunities from your opponents. This interaction isn't necessarily central to Roads & Boats, but is included for those who want it.
You never “own” territory in Roads & Boats [you don't even own the roads your transport fleet moves on], but you are able to carve out a tenuous amount of control over territory by building walls and developing “your” road network in such a way that it is difficult for other players to link up with it. You can also use walls aggressively by linking up with an opponent's network and building literal roadblocks in their way. Managing walls, knocking down barriers, and pushing forward into your opponents' space are key tactical maneuvers in more aggressively played games of Roads & Boats. This element of the game is reminiscent of Teuber's other, more confrontational, positional game, Löwenherz. An extremely aggressive game of Roads & Boats can end with long stretches of wall marking off clumps of territory as if you scored for controlling it!
What is more central to Roads & Boats than this kind of territorial expansion is the underlying idea of terrain as a resource. It is not as direct a resource as it is in Settlers or Borderlands, where you just control a territory and get its product; instead, players are required to build up the infrastructure that can harvest resources from each type of terrain and have some level of freedom in determining exactly what the output of the terrain will be. In building the production buildings, players choose between primary and secondary producers. Primary producers are restricted to certain types of terrain and produce the appropriate type of resource for that terrain [lumber from woods, clay from quarries, etc]. The catch is that only one producer may be on each tile, so to build a secondary producer [unrestricted by terrain] you must give up a tile's primary product.
The first step in the long logistical planning process of getting raw goods turned into wealth or wonder points is to decide where you will get each type of raw resource [what terrain tiles you will choose to turn into primary producers] and where you will give up production of primary goods in order to build the more complex elements of your supply chain. Because you will have to physically move goods across the map, the layout of terrain tiles in each game has similar importance and strategic impact as it does in a game of Settlers. The most influential decisions you will make in the game come in this initial planning stage where you decide how your network will lay out on the map and how you will make the best use of the terrain before you. There are two primary elements predetermined by the map that help you make these decisions: mountains and rivers.
Where you get the raw gold necessary to make any headway along the game's primary goods conversion path is in the mountain terrain hexes, and nowhere else. These are easily the most important terrain tiles on any Roads & Boats setup, and in well-designed scenarios, they are centrally located or otherwise positioned so as to create a competitive race to connect to and control these figurative and literal “gold mines”. They are often the terrain tiles on which players' independent networks make “first contact”, allowing transports to “invade” into opposing territory, and in less aggressive games might be the only place where such connections are made. As you plan your road layout and the position of secondary producers, especially coal and coin factories, you will need to take into account the distance from the mines where gold [and iron] is produced.
The second strategic terrain consideration is the layout of the river and sea terrain hexes on the map. Rivers are preset paths that can be used by water transports and are equivalent in every way to roads, except that you don't get to decide where they go. Seas may only be traversed by water transports, obviously, but require transports to make extra moves to “dock” with adjoining land in order to pick up goods. The tradeoff for the restricted positional flexibility is that water transports can carry more and move faster than their equivalent land transports. It is not strictly necessary to use water transports in a game, but a player who includes them in their plans and develops their supply chain such that producers are positioned along rivers and coastlines whereever possible will be able to be more efficient and do “double duty” by transporting goods by land and by sea.
Water transports and terrain are probably the most difficult entities in the game to use effectively. You don't have the freedom to build new paths if you find you've screwed up in the positioning of some production facility, so there is no Plan B to fall back on. When used well, however, they are a deciding factor in the game, and allow you to move more goods with greater efficiency than a player stuck being a landlubber. Additionally, water hexes can support one of the more important primary producers in the game, the oil rig, which gives you fuel directly without having to burn wood [and transport it in]. Fuel is needed both to smelt gold into coins at the mint and to create the biggest and fastest transports, in both the land and water categories. Late in the game, a steady supply of free fuel releases transporters from having to support a coal burner so they can do more important things.
The terrain in Roads & Boats is of significant strategic importance, not only in terms of what each territory produces [as in Settlers or Borderlands] but in terms of the distance between terrain features and in terms of secondary features like rivers and seas. More than in just about any terrain-dependent resource game preceding or following [Antiquity is a notable exception], the terrain in Roads & Boats provides a “landscape” on which the game plays out differently every time. Many resource or path-tracing games that followed that have modular [Keythedral, Attika] or alternate boards [Age of Steam, Power Grid] don't have nearly this variability. The modular terrain is either only locally relevant rather than of strategic import, or else the new map plays nearly the same as the old one aside from the added expansion mechanisms. Different maps in Roads & Boats truly present new puzzles to players.
Claim Jumps + Cube Churning
If anything comes to mind when you think of a modern Euro game, it is worker placement and cube pushing. I would argue that Splotter essentially created both genres when it released Roads & Boats [and Bus] in 1999. At the very least, these games are perhaps the earliest examples of these two mechanisms, if not the actual impetus for their development and current popularity. The central ideas from Roads & Boats that you do not own a resource until you have claimed it from a commons and that a resource has no value until converted into something better are at the mechanical and ludological heart of the present resource management sub-genre that dominates Euro games. They are foundational to the way modern Euro designers approach the process of game design, and they came from Roads & Boats.
Two games released in 2004, one year before the firestorm that was Caylus [which we'll get to soon], took key mechanisms from Roads & Boats and focused them in a way that portended a change to come in the prevailing ludographic landscape: Keythedral and Neuland. In both games, there is a map of terrain tiles that produce [or refine] resources which are commonly available for any / all players to lay claim to [just as in Roads & Boats]. The shift from Roads & Boats' focus is that these games are highly compacted, and players don't have even tenuous control over an identifiable production network / supply chain. Rather, players' competing access to resources overlaps from the very beginning of the game, and the focus is not on development of a production system to optimize output / throughput but on claim-jumping.
In Keythedral and Neuland, players will position their workers on the map each turn [through various mechanisms that look a lot like worker placement] so as to lay claim to particular resources for the turn. These claims are exclusive but transitory; in Keythedral, players get to store up resources between turns so there is at least that level of permanence, but in Neuland a resource claimed on a turn must be used immediately for something else [there are not even any resource tokens or cards]. In Keythedral, since worker placement is iterative / additive [happens one-at-a-time around the table], access and precedence are the important concerns; Keythedral “borrows” the wall-building mechanism from Roads & Boats that allows players to block off certain producers for their exclusive use. In Neuland, “worker placement” is a little more nuanced.
The placement of workers in Neuland happens all at once on a player's turn; each player has a certain number of worker tokens to assign to claims on their turn, and they position them in a sort of chain across the board [through a rather awkward “movement” mechanism] and then “fire up” the chosen production engine by “collecting” virtual resources and “converting” them through claimed secondary producers in order to build more production buildings to the shared map and eventually [the goal of the game] lay permanent claim to certain prestige buildings [which offer no productive use]. The game has a bit more of the “this is my supply chain” feel of Roads & Boats [despite neither game offering true “ownership” of production facilities] because you can maneuver your workers in such a way as to block of resource / production buildings for multiple turns.
Immediately after Keythedral and Neuland were released, another game that incorporated and built on similar claim-jumping mechanism took the gaming world by storm: Caylus. Caylus' worker placement mechanism has been hailed as the first of its kind by some, and is certainly among the clearest interpretations of the concept, but it was preceded by at least three games [Keydom, Bus, and Way Out West] and I think owes a bit to Roads & Boats, too [especially as mediated by the likes of Keythedral and Neuland]. In a recent interview, William Attia describes the generative spark for Caylus' design as being that of a “line” of “effects” that players chose from in an action draft and then activated in order, with the pool of available effects growing over time. I don't think he was directly influenced by Roads & Boats, but it and Neuland have very similar game structures.
This sequential stringing together of actions / effects related to resource gathering, refinement, and eventual use toward victory is a page right out of Roads & Boats. Perhaps not Attia himself but a playtester who had played Roads & Boats or its derivatives offered up some of the early changes to the game, like players adding buildings to the board, based on that experience. In style, at least, if not in actual derivation, Caylus follows much of the trail blazed by Roads & Boats. These similarities are admittedly not extreme, and were Caylus [or even the much less appreciated Keythedral or Neuland] the only game to show the influence of Roads & Boats, I wouldn't have much of a case here to argue its importance. But, I think there are quite a few other games as seminal as Caylus that show the influence of Splotter's original cube churner.
In particular, I think the influence is also evident in Rosenberg's “landmark game”, Le Havre, and to a lesser extent in Agricola [which Rosenberg has acknowledged was influenced heavily by Splotter's other giant, Antiquity; Rosenberg rates both Antiquity and Roads & Boats as 9s, by the way]. There are two evidences of Roads & Boats' influence on Le Havre: 1) The accumulation of goods mechanism [also seen in Agricola]; 2) The resource conversion chain game structure. The accumulation of goods is a key feature of Roads & Boats, where every primary producer just piles up goods every turn until they are taken [sound familiar?]. I don't think Rosenberg has ever identified where the idea for this mechanism [as he has for the harvest mechanism], but I don't think I've ever seen it tied to resources [and not, say, currency like in Puerto Rico] except in Roads & Boats and Rosenberg's games.
The resource conversion chains in Le Havre also bear similarity to Roads & Boats. There are a lot of “cube churning” games where you gather resources and turn them directly into points, but there surprisingly aren't many where you have to convert primary resources to secondary or tertiary resources before they're worth anything or can be converted to points. Most of these games that I've identified are Splotter designs or ones I've already argued were influenced by them [Keythedral and Neuland, for instance].... or Rosenberg titles. The size and scope of Le Havre's “goods cycle” is similar to that in Roads & Boats, and considerably more involved than any other non-Splotter game until Ora & Labora last year. Get some cows, then get wood and bricks to build a butchery, then kill the cows for meat and hides, then turn the hides into leather, then sell the leather [and meat] for gold. Etc.
I hope I have shown how Roads & Boats winds it way through the morass of modern Euro resource management designs while it picks up and delivers the best mechanisms from prior genres of the same. It is truly a seminal work of design with an influence that can be seen in nearly all of the most successful resource management and logistics games that followed. Indirectly, through its co-creative effect [with Bus] on the two largest worker placement titles, Caylus and Agricola, it has roots that run even deeper than the heavy cube pushers that it exemplifies, to games as diverse as Stone Age, Troyes, and Dominant Species. Additionally, as a logistical / engine building game par excellence, it has a similar indirect role in shaping designs like Homesteaders, Through The Ages, and 51st State.
Roads & Boats stands, I think, at the intersection of two eras of design, the 90s era of singularly focused games that develop one simple mechanism out into a full game, and the current era that began sometime around the release of Puerto Rico in 2001 and continues to produce multifaceted games with many interlocking and overlapping mechanisms that give rise to multiple paths to victory. With one foot on each side of this divide, Roads & Boats not only rocketed a new publisher and design team toward still increasing levels of innovation, but inspired a multitude of designs by dozens of other designers, many of whom don't even know that the ideas and mechanisms they use so often were previously tested, developed, and exemplified in this wonderful design. It is truly a game of great importance to the hobby.