The most active of the producers is Chris Lawson, so I'll start with the kits he has produced. The first was 1899, designed by Ingo Meyer and Dirk Clemens. Designed may be too strong a word, since it is largely 1830 on a different map, in this case eastern China, Korea, and parts of Russia. There are some innovations, though. To compensate for the huge terrain costs, companies earn money from mail runs, rather as in 1853, and the Outer Mongolia off-board hex rises in value as the number of connections to it increases, providing a measure of developmental interest in the longer term. But apart from that the game feels a lot like 1830 on a map which is much more constrained - most of it is two or three hexes across - and where the relative merits of the various corporations aren't at all well balanced. Bankruptcies are much more common in 1899 than in 1830, which weakens it in the eyes of many. The second version improves matters somewhat, though the real problem - the map - isn't addressed at all. Accordingly, it only tends to get played once in a while, as a change.

1876-30 and 1876-35 were designed by Peter Jacobi to see how small a playable 18xx game could be. They are, as the names suggest, based on 1830 and 1835 respectively. Each is set in Trinidad, with a map containing only 12 buildable hexes, and three major corporations. These games are playable, just about, though one is reminded of a dancing bear (or, given the size, a rabbit), where the issue is not how gracefully it waltzes, but that it dances at all. They can be played in a couple of hours or even a trifle less, though the games, especially in track building, get more than a little stereotyped. Like 1899, these games cannot be considered a complete success, and are played only infrequently, usually when time permits nothing longer.

1841, by contrast, is the 18xx game of choice for many players, though some hate it. Designed, and originally published as 1839 by Federico Vellani, whose name we'll see again, it is set in northern Italy. The McGuffin here is that 1841 is the 18xx game which is least about actually building a railway and most about manipulating the stock market and managing the extremely fierce train obsolescence. Route development has its place, but unless the players can get the rest of the game about right they are doomed, and it is this which polarises players' views of the game. When it first became available it whetted the slightly jaded palates of some players since the dividing line between success and abject failure is so narrow. 1841 achieves this by allowing companies to buy and sell shares in each other, and to merge with one another. These factors result in frequent and hard-to-predict changes to the order in which the companies operate. It's exciting and dynamic, and the only real downside, at least for "the financiers who've had enough of the engineers", to misquote the 1853 box art, is the length of the game. At typically seven hours it cannot sensibly be played in an evening. Even so, the game is usually well worth the investment of time, though it's certainly not for the novice.

Chris Lawson not only produced 1851, but he mostly designed it as well. The game started out as the largely unsatisfactory 18TN, by Mark Derrick, but virtually all that remains of the original design is the selection of companies. The McGuffin here is simplicity; many of the fiddlier bits of other games have been discarded. The game starts, not with an auction, but simply by dealing out one private company to each player; this mechanism is taken from 1825, but unlike that game it is made to work by ensuring that the private companies are balanced. The map is simplified by discarding all of the small cities, and none of the remaining cities have special tiles. It takes from 1841 the property that share prices rise only when earnings exceed the share price, and from 1849 comes a limited form of financial manipulation. 1847's long-distance run bonuses are adapted to promote route development (although Chris maintains that although he'd seen 1847 before he designed 1851 the bonus runs were not intentionally borrowed). Examined in detail, one is reminded of the words of the Tom Lehrer song, Lobachevsky, "Plagiarise, plagiarise, plagiarise - but remember always to call it, please, `research'". Despite this unflattering summary, the game works remarkably well, and, at about three hours, it is the shortest truly playable 18xx game currently on the market.

1837, by Leonhard Orgler, is an 1835 variant, set in the Austro-Hungarian empire and expanded until the pips squeak. The McGuffin here is sheer size. Instead of six minor companies it has 21, allowing everyone to run several companies. Ten of the minors eventually fold into six of the "normal" major corporations while the other 11 form the basis of three different "state railways" rather like 1835's Prussian. The map barely fits onto a sheet of A1 and contains more buildable hexes than any predecessor (though it is of course dwarfed by 1827 and 1831). It has 15 different types of train, at least ten of which survive to the end game, though often there is only one or two of each type available. As a consequence, it is often hard to predict just what track is going to be useful in the long term. Despite this it vies, in the UK at least, with 1841 for the title of most popular 18xx game kit. It usually takes about seven hours to play, though Dirk Clemens' 18xx amanuensis program can usually trim a couple of hours off that (suggesting that we normally spend two hours per game just manipulating bank notes) and players at the brisker end of the spectrum can trim yet another hour off the game; as a result 1837 can be played in an evening.

1842, by Wolfram Janich, is a game set in Schleswig-Holstein with a distinctly 1835ish heritage on about the same scale as 1830. It introduces three new ideas, none of which, unfortunately, work at all well. Some of its minor corporations can be sold, for arbitrary amounts, to major corporations, resulting in unlimited looting of corporation treasuries. It's the same sort of thing that's done in 1830 with private companies, but without any checks or balances. The map's main city, Hamburg, can only be developed by a special corporation, using mechanisms which look OK on paper but which in practice involve too much accountancy for little gain. Finally, other cities can only be upgraded if they lie on a route connecting the North Sea to the Baltic, a restriction which causes no end of annoyance when the pettifogging details slip the mind. These problems could no doubt be overcome, though there seems little incentive since they're the only features of the game which stand out enough to be called a McGuffin.

1847, also by Wolfram Janich, is another 1835 variant, this time stripped almost bare. The map, set in the Pfalz region of Germany, fits on an A3 sheet, and there are but six major corporations. For reasons probably concerned more with historical accuracy than playability a third of the map is the private reserve of one of the corporations, while the others get to build track in the rest. The result is a game which plays in two to three hours, so it's short enough to fill those annoying little gaps in a gaming session, but it isn't very interesting, unfortunately. It also has a few flaws, the most annoying of which is the fierce train rush in the mid-game. (This rush is common to a lot of the small games, for some reason.) One idea which is superficially attractive is the loco works, a private company which receives a cut of each train purchased. Unfortunately it just leads to excess administration.

1869: The Golden Spike, by Colin Barnhorst and Kris Marquardt, has elements both of the large and the medium-sized. Its map, covering the western half of the United States on an A1 sheet, has about 200 hexes, and there are 12 major corporations. On the other hand, the corporations can lay up to three tiles per turn, so the map fills up quite rapidly, and the early track lays are tightly constrained by land grants. The companies tend to merge together in the later stages of the game, so the scale of the game isn't anything as large as it looks. On the debit side, though, the game isn't finished, despite what its designers thought when they released it. Several of the rules can, and therefore probably should, be discarded without losing anything except needless complication. The map is riddled with errors. The end-game is complicated by "classification yards", a nice idea which needs refining. If a train passes through a classification yard, it gets to make more stops than it might otherwise do. This is fine, except that things get out of hand when you're running a 5-train through 16 stops, and entirely ridiculous when you're running two (or even three) such trains. Despite these flaws, the game shows promise. It does appear to recreate the race to build transcontinental railroads and this makes a decent McGuffin. This is one to watch out for in the future.

18GA, by Mark Derrick, is a small-scale game set in the US state of Georgia. Following an unfortunate experience with a nine-hour game of 1870 at a convention in the US, Mark is a convert to the idea that most 18xx games take far too long, a viewpoint with which it's hard to argue. He is attempting to rectify matters by designing ever-smaller games. 18GA's map fits on an A3 sheet with six corporations, and the game can be completed in a couple of hours. It very nearly works, too. There is a tendency, in 4-player games at least, for the game to pause over a "poisoned" 2-train as well as the more usual poisoned 4-train, but apart from that things flow fairly smoothly. The Civil War (or "late unpleasantness", as I believe it's usually called in that part of the world) is rather cutely represented by Atlanta being worth less in the green phase than it is in the yellow or brown. The McGuffin here is simply the small scale of the game.

1849, by Federico Vellani, was originally entitled 1850, and in its earlier versions it boasted, if that's the right word, the most seriously broken start mechanism in any 18xx game, kit or otherwise. Set in Sicily, its start was designed to reflect the fact that various entrepreneurs bribed the King of Sicily to allow them to build railways on the island, and he took their money but denied their requests. In the first three versions there was a tremendous amount of luck at the start and it completely dominated anything done with the companies once started. However, the latest version introduces private companies sold as in 1830 and the companies which start in favourable locations pay heavily for the privilege. The result is a short (three hour) game in which much of the flavour is provided by the extreme cost of building track, and without any of the mail runs which make 1899 tractable. It's the first game produced which had "H" trains; these can travel a fixed distance (halved for narrow-gauge track) and visit as many cities as they can reach. This is certainly a better model for how real-life trains work but at some cost in playability. Chris Lawson is, as I type, hard at work producing a version which should be ready shortly - watch for announcements in the usual places.

Last, and by no means least, is 1831, by Carl Burger. It's not really a kit, in the sense that there is no need to laminate and cut out the components, though it's not really of commercial quality. The McGuffin here is the array of (metaphorical) kitchen sinks included in the game. 1831 covers more or less the same ground as 1830, that is, the north-eastern US, but on a somewhat larger map with much smaller hexes. There are some 450 buildable hexes on the map (six times 1830's count), up to 30 companies (about 8 of which are coloured green) each with up to 40 station tokens. Trains can carry freight, passengers, or both. The stock in each company is divided into preferred and ordinary shares, with somewhat different rules governing them; in particular, the company president is the largest preferred stock holder. Companies can merge to make systems, and systems can lay up to seven track tiles per turn. All in all it's a monster, requiring well over 12 hours to play and, at $250 for an original edition, comfortably the most expensive 18xx game ever produced. There was talk of Winsome Games producing a commercial version, presumably more cheaply; it's not clear whether that will come to anything. While it's true to say that you certainly get a lot of components for your money, the game is just too large to play at a single sitting. To cap it, there is also a rules problem whereby it's possible to double your investment in a company in one round without actually running any trains, and the game is heavily slanted towards running in the New York area and starting a company elsewhere almost certainly means you won't win. As an exercise in taking 18xx games beyond all natural limits it succeeds, but there doesn't appear to be a game in the box, which is a pity.

When 18xx devotees gather together, the choice of game is now rather wide. My own collection includes nearly 20 different games, including the commercial ones. In practice, though, we tend to play the same few titles most of the time - 1830, 1856, or possibly 1837 in an evening, perhaps 1851 if we're starting late, or 1837 or 1841 in an all-day session. Why is this? Well, one side effect of playing lots of different 18xx games is that most of the players find themselves uncertain of the rules of the game at hand. In order to save wear and tear on the "little grey cells" there is a lot to be said for a game where the number of differences from a "core" 18xx is limited, but each new feature is strongly themed, critical to doing well, and simple to describe - in short, a McGuffin. The games we play score well in this respect, and we can overlook small flaws.

So, what of the future? David Hecht's 1826, set in France and Belgium, is in the very late stages of playtest and may well be published soon (probably by the indefatigable Chris Lawson after he's done with 1849). Its main features include a complex set of transitions from minor to major companies, the formation of state railways like 1835's Prussian, and rather complex rules governing electric trains. Federico Vellani is working on 1827, a monster covering the whole of the US, and 1827Jr, a more reasonable-sized game covering the same ground but with enormous levels of bookkeeping. Progress on both is reportedly slow. Tom Lehmann is more or less satisfied with the current state of 1834, set in New England; whether it is produced as a kit, commercially, or not at all depends on the politics of licensing. Its features include regional bonuses, a flexible stock market, and takeovers. Mark Derrick is working on 18AL (Alabama), which is apparently rather like 18GA on a different map. If that's true, I doubt that it will be well received, but we'll have to wait and see. Chris Lawson receives prototype 18xx games with surprising frequency, and one or more of these may well be produced when Chris gets some time, though most are obviously hopeless. Chris himself is working on 1852, which uses the same map and many of the same components as 1851, but with more companies to cater for more players. Provided Hasbro, the new owners of 18xx, don't decide to get heavy with producers, the future is bright for the 18xx game kit.