We will meet at the Hour of Scampering.
What Led to this Crazy Project?
My affection for Silverton started before I even bought the game – I just didn’t realize it at the time. I had played John Lueke’s Rails Through the Rockies dozens of times – I had even “pimped” it, with color-coded pieces of plastic I-beam to mark what mines were open. RttR had some great things going for it – varying terrain difficulty, mountain pass and tunnel construction, mines opening and closing, narrow- and standard-gauge rail choices, and winter turns. These were all new ideas for RR games, at the time.
But RttR also had some problems. The major one was its predictability. All players always started in Denver, and there were only 3 practical routes going into the Front Range. (In fact, for having a hex-based map, it had the largest portion of unusable hexes I’ve ever seen.) The mines could not be predicted or anticipated in any functional way, and their output (i.e., profit) was unvarying. Too, it was very easy to find yourself out of the game after a bad winter, and an event card or two, or a pass-construction injunction freezing your progress. (The injunction cards were the nastiest thing in the game, because they were, simply, game-changers, but in a negative way.)
When I first opened Silverton, I was disappointed because it lacked a grid. Instead, it had routes that players claim. (And that’s part of the story that comes later.)
But what Silverton did have more than made up for that:
Locomotives improving over time
Mines becoming available for anyone to claim, with numbers for output and duration, on them, that made it possible to perform some risk assessment
Snow removal considerations
Fluctuating mine output – but not completely random
Fluctuating goods prices, based on units sold
More options for paths through the Rockies
The inclusion of Utah (and later, New Mexico)
Mines, mines and more mines – gold, silver, lumber, copper and coal.
By 2000 (about eight years after I bought it), I was at the point where I wanted something nicer-looking than the original, Two Wolf components. Considering it was one of my favorite games, I felt it merited the effort. So, I made a new board, new mine cards, new passenger route cards, and new locomotive cards.
The First Frankenstein
And I was happy with that, for most of the last ten years. However, somewhere around 2005, I stumbled across another train game – TrainSport: Austria. It came in a small tube, and the board was a wipe-off board. But it wasn’t a crayon-rail game in the same sense that the Empire Builder series is. Instead, the board is divided into regions, costing from one dollar to five dollars to build in, reflecting the difficulty (aka, elevation) of the terrain. The crayons are used to mark presence in those regions where you’ve built.
But the system – while being very abstract – is clever, resulting in a game that is almost chess-like. I felt it merited a cosmetic upgrade, and it seemed painfully obvious to equate the five degrees of regions with five levels of elevation. So, that is what I did, using the foam mats you can buy at craft stores. I re-drew the map increased in size, and then printed the successive levels, with terrain coloring and texturing to thematically match the levels of elevation. It came out quite well (see below). But there are a couple of things I don’t like about it. Due to the use of the foam mats, the board is heavy, and it is flimsy. So, when I started toying with the idea of doing a three-dimensional treatment for Silverton, I decide I did not want to use the same material, again. I needed something lighter and sturdier.
Trainsport: Austria (2005)
And the back of my mind had been toying with the idea of doing this for Silverton for several years, now, but I never pulled the trigger, until it dawned on me that I had been doing modifications for any number of games, over the last ten years, but not the one modification that I thought of as being the “most cool” – the Rocky Mountains in 3D, as a playable game board.
When I first started taking the idea as a serious, priority endeavor – about six months ago - I just assumed I’d be using the routes found on the original Silverton board, and printing them up on the new board. And the more I thought about it, the less I liked it. For one thing, that meant these lines cross the different elevations, interrupted by the vertical increments between levels – not pretty. But then an even more germane question came to mind – why spend all that time on a 3D map, if it was going to serve as nothing more than eye candy? No, the elevations had to actually mean something, as they did in Trainsport: Austria.
Making the Board
I had an idea about what I wanted to do – something like the hex-based costs found in the venerable Rails through the Rockies. That meant using hexes – areas would be too big, and require multiple mine sites co-occupying the same areas. I didn’t want that – it likely would make for some messy rules about presence and ownership. So, hexes it would be, and hexes that covered a small enough amount of terrain that one would never find two mine sites in the same hex. That turned out to be something smaller than about 40 miles per hex. (The scale I finally settled on was 17 miles per hex, across the flats.)
Maps were what I needed next. In 2000, I had done some elevation artwork on my first Silverton map, but it was just something to look at, to get a feel for the terrain – it had no impact on game play whatsoever. But now it was 2010, and here comes Google Earth to the rescue. (Ain’t technology great?) I quartered each state (about 12 maps in all), and used the “terrain” option on the web page images, saved each of them, and brought them into Corel Draw. I then overlaid a 5/8-inch hex grid, and got to work assigning elevation values to each hex, and dropping names in for the cities and mines.
Mind you, this process isn’t “fudge-free”. Terrain maps do not give you the overall elevation for a given, hexagonally-shaped chunk of real estate. And elevation isn’t the whole story, either – judging how expensive and labor-intensive a section of land is, for building rail through it, has as much to do with how rough the terrain as, as it has to do with the elevation. So, making this kind of map is a “fudge-rich environment”. (Why am I suddenly hungry?)
It took time, and a lot of it was like playing Dominion; that is, leaving me largely free to do some deep thinking. How, exactly, was this rail-construction system supposed to work? Clearly, I couldn’t use crayons. I could use the markers that came with Silverton (and those I made, in addition to the ones that came with the game), but flat markers would likely get lost in the varying elevations of the board. So, I decided cubes were probably the best way to go (as I previously had, in Trainsport: Austria). The costs for each hex would be the value of the elevation, and I had decided to go with 7 levels of elevation. Too many, I wondered? Perhaps, but rather there be too many than too few, was my reasoning. If construction cost was a factor of elevation cost, though, the scale had to start at higher than “1”, or else players would be scooting all across the lowlands of the board with the same effort that got them essentially nowhere at higher elevations. So, I set the lowest level as being elevation “2”, and the highest as “8” – with “8” being tentatively considered impassible. (Not going to be building straight over the top of Pikes Peak, for instance – rather, building around it.)
While building the map, I also decided that above a certain elevation would be where snow removal was required. At first I figured this should be elevation “6”, but I later brought it down to elevation “5”. As the higher the elevation level the darker the hex would be filled in, I went with the elevation number inside the hex being black for levels 4 and lower, and white for levels 5 and higher, making it easier to tell at a glance which hexes were going to be snow-bound during winter.
My construction cost model went through many mental variations, while making the map - and in fact, as I write this, is still not quite set in stone. I first figured on the costs as being as follows:
For building rail in a level 1, 2, 3, or 4 hex, the cost would be the cost of the target hex. For elevations above that the cost would be the sum of the target hex and the hex of origin. And in all cases, the cost would then be multiplied by the difference in elevation. Furthermore, if the elevation was “5” or higher, the cost would always be the sum of the two hexes. (Due to the land be that much rougher and uneven, at that elevation.) So, this amounted to costs:
Not without some flaws, right? For starters, it’s cheaper to build between elevations 6 and 7, rather than between elevations 7 and 7. I rationalized this would be due to the increased requirement for large spans of bridge, and cuts for mountain-side grades, but it is just a rationalization. So, this formula has undergone some “normalization”, and probably isn’t done, yet.
But none of this kept me from building the map. And so I initially built a “flat” map, with which to test the rail construction system, before moving on to the more ambitious seven-layer cake. So, I’m building a rail construction system from scratch – and like just about every other core mechanism for a themed game, “scratch” is relative. In other words, I have a pretty good idea of what’s in the store, for current construction systems, and whether I find them suitable or preferable will dictate what I use, or whether I’ll have to pull something out of thin air.
I like the idea of a somewhat random distribution of construction points, so I decide to try it with each player getting 2D6 worth of construction points, for starters.
This goes nowhere, fast. Literally. So long as you want to run around on the east Colorado plains (also known as “western Kansas”), you’re fine, but trying punching into those mountains, and watch them laugh at your puny two dice. (See numbers, above.)
So, then I jack it up to four dice. That’s better, but still not quite there. Even so, I know I’m probably on the right path, so I press forward with building the map layers in Corel Draw. (The elevation color scheme goes through an evolution.)
The first level is a breeze – it’s the entire map, with every hex above the first level filled with white. I print it (as I print all the levels) on full-page label paper. I’ve chosen black foam board for the medium (found in Office Depot, among other places), as it is light, sturdy, easy to cut with a razor knife, and the black looks good for the vertical facings.
In the process of making the map on the computer, I’ve added a coordinate grid along the perimeter. This will be referenced on the mine cards – not for me, as I know the map almost by heart, at this point, but for newcomers to the game, reducing the time they spend hunting around for the mines that are available.
I’ve also left off the price tables, as I intend to make this a stand-up piece, using dowels and spools to indicate current prices. This will not only be easier for all to see, but will also reduce the chance of all the price markers getting misplaced by a bump of the board.
Silverton 3D – First layer
After applying the first level, the real work starts. For each of the remaining six levels, I’ll be spending a great deal of time cutting all the hex edges. And it’s surprising, how fast foam board can dull a blade. I use X-Acto #11 blades, and I went through about half of a large pack of them, just completing this project. Roughly, I would get about one hour of cutting out of each blade. You know it’s starting to dull, when the cuts start pulling little bits of foam along with them, and it just gets worse from there.
Mounting the second layer (the first elevated layer) presented a new problem I hadn’t anticipated. It’s next to impossible to make your cuts through foam board at a perfect, ninety-degree angle. Too, you don’t want anything more obtuse than that, because it will obscure part of the adjacent hex at the lower level. Yet when the angle is more acute than ninety degrees, it leaves a bit of those white-out hexes showing up, from the lower level map. Yes, I could have started over, and gone through a few black ink cartridges, making all those hexes black, but I didn’t do that. Instead, I took a Sharpie fine point black marker, and ran about a 1/8” border along the insides of all the lower-elevation hexes being covered by the next elevation. Time consuming, yes, but the board looks a lot better this way. The black of the pen matches up perfectly with the black foam board.
First Elevation, viewing from oh, say, west Texas
As I built the board with subsequent elevations, the total number of hexes being printed decreased; however, the number of cuts having to be made remained almost constant, because the number of contiguous hexes of the same elevation also decreases. You find yourself cutting more and more single hexes, and single lines of hexes. I hadn’t really figured on that, either. Once more, the project took longer than anticipated.
But I’m pretty happy with the final product; overall, it looks pretty good. And it’s sturdy as hell, what with being seven, glued-together levels of foam board. This sucker isn’t going to be bending or warping. My one regret is just how thick foam board is, when you place seven levels of it together. But that’s a minor annoyance, really.
Completed board, viewed from the east
Evolution of a System
As mentioned before, I had a few assumptions going into this, which gave me a starting point for working out the mechanisms involved. First, that construction would be hex-by-hex, and each hex would have a cost related to its elevation. Second, that each player would have X amount of construction points per turn, with which to build. Third, that such points could be used to lay multiple, new segments in a turn, but that the longest segment would have a maximum length (initially, six hexes). There are corollaries to these, but I won’t dwell on them.
This left a great number of questions requiring answers. For example:
1) Should the construction points available be random or fixed?
2) Should more than one railroad company (aka, “player”) be allowed to occupy the same hex?
3) What construction cost formula would best balance the relative proximity of mines to Denver (at higher elevations), with the more remote mines (at lower elevations) found in starting in El Paso or Salt Lake City?
4) Should there be a cost associated with maintaining the labor for construction, in addition to the cost for the construction itself?
5) What role should surveyors play – or should they have any role at all?
6) Should construction be performed by a fixed asset, such as construction crew markers?
7) If the answer to 2 is “no”, what should be used to determine order of track placement (fixed turn order, roll-off, etc.)?
8) Should a player be able to accumulate construction points from one turn to the next? (And possibly more to the point, should the construction of a given hex of rail be allowed to spread across more than one turn, for those really tough hexes?)
And the closer I got to a model for implementation, the more questions arose. And once I tested a model, still more would come up, even as others started being answered.
An early play test on the new board, with an obvious conclusion – building into the mountains is taking just too, damn long. This is evidenced by green’s prolific expansion almost to the Colorado border, while red and white are still futzing around at Canon City and Georgetown, respectively. A little harder to make out is black, coming out of Salt Lake City, at the far end of the board.
I tried to be clever, initially. I suspect all designers do this, and I also suspect that the better designers are the ones who know how much “clever” to use, very early in the development process, as opposed to such as myself, who gets to attend remedial classes in the school of hard knocks.
I had a role for construction crews, surveyors, and prospectors. Prospectors were (and are, so far) essentially unchanged from their functions in Silverton. Surveyors were now going to be purposed for establishing right-of-way, and reducing construction costs when traversing elevation changes. Construction crews were going to restrict the player to exertion of effort at the point where the crews were placed. Too, the amount of effort would be commensurate with how many crews were placed in a location.
To heap cleverness on top of cleverness, I made up 5 cards for each of the unit types, for the commitment to using 1, 2, 3, 4, or 5 of each unit type at increasing cost. The player would secretly allocate these cards on their current turn, for their next turn’s efforts, thus forcing everyone to plan ahead.
In my desire to keep things interesting, I decided that each construction crew used would give the owning player one six-sided die worth of construction points. I was even clever enough to imagine the player cursed with a spate of bad die rolls, and smugly added a chit-system, whereby players with an average roll total of 3 or less per die would draw a chit that would, on future turns, be turned in for an extra die roll.
And that right there should tell you (or me, in this case) something is a bad idea – when you’ve already come up with a method to compensate for it. You not only have two rules where you should have only one, but you have two things (and their interaction) to test, where you should only test one.
I’ve subsequently stepped back from the purely-random nature of construction points, and I know the final system will be a compromise, with part of the total being fixed and part random. (A completely fixed number of points will lead to all sorts of “perfect plan” mentality, and I want to avoid that. Better if each playing of the game grows some of its own personality.) At the moment, I’m thinking 12 points fixed plus up to 3 dice that can be purchased. And I’ve dispensed with the Construction Crew markers, allowing the player to simply state where he will be starting new construction on the turn.
I’ve dispensed with the cards altogether. I like the idea behind it, but it’s simply too much of a headache play testing by oneself, and likely will just add time to a game that already goes on for quite a long time.
I’ve also dispensed with the bevy of surveyors that players could use, earlier. In their place, I’ve come up with a rule for construction on right-of-way, and a series of player options to select from (as in Age of Steam), at the beginning of the turn:
1) First Build – just what it says - you get to build before anyone else, if you pick this.
2) Extra Construction Die – add one more die to your roll for Construction points this turn.
3) Automatic Claim – in addition to using prospectors for claiming on this turn, you can take one mine or one (eligible) passenger route without using a prospector, and without worrying about competing for the claim
4) Surveyor – you may select one hex on the board this turn, which is reduced in elevation by 1, for the purposes of building into it
5) $500 – take 500 extra bucks. Please.
First Build cannot be picked by the same player, two turns in a row. That just didn’t seem right, as this will often give one player, in competition with others out of the same starting city, an insurmountable advantage – particularly in grabbing passenger routes. The others, I see no such problem with, for the moment. Player order will change based on how many construction actions you take during your turn. In deference to the esteemed Mr. Wallace, he who passes first during construction will go first next turn. Why would you need to go more than once? Because, you may only build one new segment in a construction action. That segment must have a specific start point adjacent to an existing rail segment you own, and cannot exceed 5 hexes in length. If you wish to build other segments, elsewhere on your turn, you’ll have to wait until it’s your chance, again. In the meantime, players with more modest construction ambitions on this turn will be scooping up early positions in the next turn of the game. I love well-rationalized karma in games.
For those extra-hard-to-get-to places, at rarified heights, I’ve included a “banking” system for construction points, in the form of counters (1s, 5s, and 10s). These counters are placed on the hex where you have started construction for the turn, but cannot complete it. Other than this, any construction points you do not use on a turn are lost.
In these aspects, the system works. It is down to getting that set of construction costs fine-tuned, for a given range of construction points available. As I stated, each player is going to receive a fixed number of these points, with the option to purchase dice for the generation of more. This provides the opportunity for spending big in an effort toward aggressive expansion, but by no means guarantees unqualified success when doing so.
There are other elements that still need to be completed – the new mine cards, the price chart, cards for generating price changes (because I hate all those rolls and calculations at the end of every turn), and so on. What applies equally well to the standard version of Silverton I’ll be posting on BGG, too.
If I find that more people have an interest in this project than can fit in a small elevator, I’ll be making a follow-up document.
- Last edited Fri Aug 13, 2010 4:13 pm (Total Number of Edits: 1)
- Posted Fri Aug 13, 2010 4:13 pm
I finally got to try out Silverton for the first time last weekend at GenCon and loved it. This just looks like extra fun heaped on top of the fun.
We will meet at the Hour of Scampering.
Silverton isn't for everyone, but if you want to sink your teeth into a beefy railroad game, it scratches itches other RR games don't.
How is testing/playing going?
Have you worked on any of the other stuff you mentioned? ie mine cards/price charts/price change cards
We will meet at the Hour of Scampering.
Testing has gone well. Although it's in hiatus for the time being, I'm happy where it is - each player getting a base of 12 construction points, and able to purchase up to 3 D6 rolls at $200 each (for additional construction crews). I changed the costs of construction above the "snow line", so it's identical to that below the snow line, which means that it's possible to get to Leadville from Denver in the first year (a rule of thumb I think is a good gauge).
The new mine cards are 99% done. I need to make a QC pass on them, while adding the 3-letter abbreviations to them. (You might have seen some of the abbreviations on the board photos. They just make it easier to read from a distance.)
I've made the card decks for price changes, for gold, silver, and copper. So, still a ways go, with lumber and 2 flavors of coal (as the number sold affects prices differently, for 2 sets of cities).
The new price change board is still in my head, but it's not a difficult thing to make.
However, I'm taking a break to paint the figures in Battles of Napoleon: The Eagle and the Lion, now that all the questions about that game's rules seem to have been answered.
The one question I still have in my mind is whether this "version" of Silverton would merit a separate entry in BGG?