Eclipse: raising the bar for the 4X game

I’m a big fan of the game genre called “4X” – “explore, expand, exploit, and exterminate.”. I’ve been playing these ever since the ur-progenitor of the genre in the 1980s, Empire, and I actually still maintain an open-source C version of that game. Civilization is my favorite computer game ever, and by what I hear of it Master of Orion – the game “4X” was coined to describe – would have hooked me even harder if I’d known of it when it came out.

I particularly like SF-themed 4X games. I have previously posted a favorable review of Twilight Imperium (hereafter “TI”), a big sprawling epic of a contending-galactic-empires 4X game. But now I write to report on a game that effectively makes TI obsolete – a new design called Eclipse which I think is going to permanently raise the quality bar in 4X games.

When you unpack the components for Eclipse, you’re going to immediately get the impression that it’s Twilight Imperium lite. Hexagonal starsystem tiles for variable board layout – check. Plastic ship models in different sizes – check. Playing mats, describing human and alien species one per – check. This impression is not exactly wrong, but the differences turn out to be more important than the similarities.

One difference is that the game doesn’t start with all the board tiles down. Instead, player homeworlds are arranged in a broken ring with unexplored space between and around them. Unlike TI, which has exploration only as a bolted-on afterthought with the Distant Suns option, exploration is central to this game and one of the ways to win is to explore more aggressively and successfully than your neighbors.

Another difference is that instead of a huge pile of available ships you have only a relatively small handful. Interestingly, this actually encourages combat, because losing your fleet-in-being isn’t a catastrophe that will take you half the rest of the game to recover from.

But the most important difference is not local to one aspect of the game, it’s a global fact about the style of the entire game. Eclipse is as tightly constructed and carefully interconnected as a Swiss watch. By contrast, TI is a huge sprawling pile of game mechanics that make terrific thematic sense but don’t integrate all that well and in some cases are only half-realized (hello, politics subgame, I’m looking at you!).

Here’s an example of what I mean by tight construction. Your player mat has a track with disc-shaped pieces on it. You have to expend one of these temporarily (getting it back at the end of the round) to take a game action such as moving ships performing research, etc. You have to expend one of these permanently to control a solar system. This matters because the track beneath the pieces has numbers on it representing the upkeep cost for your empire; as you take actions and seize systems, it rises. If at the end of a round you can’t cover that upkeep from your money reserve, you have to give up solar systems (taking back disks to cover numbers) until you can.

That one mechanic (somewhat reminiscent of the resource market in Power Grid) creates a delicate multi-way tradeoff between seizing territory, taking actions, and building a money reserve that you can use to finance a late-game surge. Because it does so with very little state, you can reason about your option tree more quickly and effectively than in a game with heavier mechanics. This nets out as faster turns and shorter overall playing time; where a 6-player game of TI can easily take 8 or 9 hours, I’ve seen a 5-player game with mostly newbies take about 5 hours and a following 6-player game take about 4:30. After another play or two I expect my group will get down to the designer’s estimate of a half hour per player.

Most of the the people in both games had previous experience playing TI with each other, and after the first game the consensus was already becoming clear; this game pretty much obsoletes TI. You give up some thematic chrome; the real draw in TI’s sprawling elaborateness is the way it ticket-punches every trope from battlestars to the Galactic Council in a loving tribute to all those classic space operas you read as a kid.

What you get in return is a much better game – tighter, faster-playing, less vulnerable to runaway-leader effects, packing just as much tactical and strategic depth and multiple paths to victory but with much lower total complexity overhead. Eclipse is elegant in the way a mathematical theorem can be elegant – minimal premises worked to a powerful and satisfying conclusion.

I learned this morning that Eclipse, though only released in 2011, has shot up through BoardgameGeek’s game rankings to make #7 in the top ten. I’m not even a little surprised, and expect that game designers will be studying it as an innovative example for years to come.

Published
Categorized as Games

48 comments

  1. That mechanic you describe sounds like something that would work well for Collusion.

    If I proposed an unholy hybrid of Eclipse and Yggdrasil, would you play it?

    1. >If I proposed an unholy hybrid of Eclipse and Yggdrasil, would you play it?

      Yes. Would you believe I actually had that thought already?

  2. (reading… get to 2nd para… very exciting! I loved MOO, love 4x, and now here is a new game is “going to permanently raise the quality bar in 4X games”…)

    Oh! — it’s a board game.

    Rats. Might have mentioned that in the title.

  3. I quite enjoyed it. And the sheer enjoyment of playing really eases any feelings of resentment and ill will that occasionally crop up in such games. . .

    1. >And the sheer enjoyment of playing really eases any feelings of resentment and ill will that occasionally crop up in such games. . .

      Yes, Phil. But you are still owed a serious stabbing.

      Phil jumped me on the last turn of game 2 and took three of my four systems. It’s a tactic that won’t work twice; he actually stripped the weapons off his ships in order to mount extra-capable drives and neutron bombs, meaning that any system defenses at all would have stopped them cold. At least he actually won with this tactic; if he’d done it just to screw with me I would have been really pissed off.

  4. @Jeff Read
    And you just know it’s going to be broken up into a two part core game plus expansion! Or maybe that’ll be in the computer game adaptation. Pfffff.

    @esr
    I love these games reviews you do. Not a one of my table gaming friends has any interest in board games, strategy-type or otherwise, but I am am a sucker for them. I play them vicariously through your reviews. This one sounds fun!

  5. I played Master of Orion around the time it was originally released, and, yeah, I can confirm that it was pretty addictive. I’d love to be able to play that game again.

  6. …It’s a tactic that won’t work twice…

    All I could think while reading most of this comment was “the enemy’s gate is down”.

  7. (And an off-topic theme note: Your main-text font selection apparently doesn’t cascade into blockquotes.)

  8. @Christopher Smith

    (And an off-topic theme note: Your main-text font selection apparently doesn’t cascade into blockquotes.)

    I think that’s deliberate.

  9. I really enjoy these game reviews. Like others, I don’t always get much ftf gaming time so it’s nice to live vacariously. Also I steal ideas for when I do get to game.

    in the late 80’s to early 90’s I played a *lot* of empire. Had to stop cold turkey. It’s a fantastic and addictive game. If I may ask, what version (codebase) do you help maintain?

    1. >in the late 80?s to early 90?s I played a *lot* of empire. Had to stop cold turkey. It’s a fantastic and addictive game. If I may ask, what version (codebase) do you help maintain?

      VMS Empire.

      HISTORY

      According to A Brief History of Empire The ancestral game was written by Walter Bright sometime in the early 1970s while he was a student at Caltech. A copy leaked out of Caltech and was ported to DEC’s VAX/VMS from the TOPS-10/20 FORTRAN sources available around fall 1979. Support for different terminal types was added by Craig Leres.

      Ed James got hold of the sources at Berkeley and converted portions of the code to C, mostly to use curses for the screen handling. He published his modified sources on the net in December 1986. Because this game ran on VMS machines for so long, it has been known as VMS Empire.

      In 1987 Chuck Simmons at Amdahl reverse-engineered the program and wrote a version completely in C. In doing so, he modified the computer strategy, the commands, the piece types, many of the piece attributes, and the algorithm for creating maps.

      The various versions of this game were ancestral to later and better-known 4X (expand/explore/exploit/exterminate) games, including Civilization (1990) and Master of Orion (1993).

      In 1994 Eric Raymond colorized the game.

  10. What you get in return is a much better game – tighter, faster-playing, less vulnerable to runaway-leader effects, packing just as much tactical and strategic depth and multiple paths to victory but with much lower total complexity overhead.

    I think some of the draw for TI was that complexity overhead. Having said that, this is definately going on my “to look at” list.

    And i’ll admit that my first thought was “how does an IDE raise the bar for 4X games?”.
    Stupid name collisions.

  11. > Stupid name collisions.

    Hate those. Broken bits of perfectly good letters everywhere. Worst one I ever saw actually knocked the capitalization off!

    Yours,
    Tom

  12. I still play Master of Orion (“MOO”) on DosBox. It was half the reason I upgraded to Windows 95 back in the day–Windows95 would do memory management and I was sick of screwing with autoexec.bat and config.sys to squeak out enough memory to get MOO to play.

    The game play is just as addictive as ever. In my opinion, it struck that perfect note that still hasn’t been beaten.

    I’ll have to take a closer look at Eclipse, but for the moment, I just can’t imagine a 4X boardgame that can beat the ultimate 4X computer game.

    For those considering it, several websites have it for purchase for around $5. Best $5 you will ever spend!

    1. >I still play Master of Orion (“MOO”) on DosBox.

      Can I get this to work on a Linux machine? Because I’d really, really like to try it.

  13. Can I get this to work on a Linux machine? Because I’d really, really like to try it.

    DosBox has various linux packages and source available. It requires libSDL and a “decent C++ compiler like GCC”.

    According to their compatibility wiki entry for Master of Orion the game is fully supported in recent versions. You’ll probably need to kerjigger the config file.

  14. Slightly off topic, but are there any games with much shorted playing time? My SF group often likes to dissolve into games (nearly always werewolves but occasionally 7 wonders and the like) and I’m on the look out for an SF themed board game, but a turn around of an hour or maybe a little longer.

    1. >Slightly off topic, but are there any games with much shorted playing time?

      Yeah, look into Eminent Domain. Very light, very fast, decent play value, 45 minutes to an hour. Or Race For The Galaxy which might be an hour when you first play but collapses to 20 minutes with experience. Both card-driven.

  15. I’ve never played the original MOO, but I have successfully run various old DOS games under DOSBox on Ubuntu (10.04). I have also run MOO 3 under Wine.

  16. @Eric
    Absolutely about Linux. I run Ubuntu. Here’s my dosbox configuration from Bash:
    dosbox -c ‘mount C: /home/clay/orion/’ -c C: -c orion

    “-c” allows a dos command to be entered into DosBox from Bash, but if you run dosbox you can then just enter these commands sequentially like you were in Dos

    So, here’s what my command line does:
    (1) I create a “C” drive in Dosbox and mount my orion directory on it
    (2) I switch to my newly mounted C drive
    (3) I run orion.exe

    After that, you’re in business.

    1. >After that, you’re in business.

      Yup. Graphics very blocky and cheesy by modern standards but the gameplay is clearly quite deep. I’m looking at the original MOO, I found it as a free download. I see there are later revisions. Which do you recommend?

    1. >GMT Games, run by some friends of mine,

      You travel in good company. Tell them ESR is a fan, if you think that might mean anything to them.

  17. VMS Empire

    Ah, thanks. Different lineage – what I played was descended from PSL Empire. It’s gone through a number of different incarnations, but interestingly there is still at least one open source C version being actively maintained.

    It’s real-time multiplayer and networked, on a client-server model. Depending on the server settings, single games can last from hours to months. You might find it interesting, but it has a tendency to become an open-ended time sink.

  18. Are there any free pc mmo’s that are similar to the highly strategic wargaming sort of board games you describe on here on occasion? Sounds like fun, but I don’t have a lot of people to play with in the real world.

  19. Haven’t looked at DOSbox for a while … The last time I gave it a look I struck out trying to get Syndicate Wars running acceptably.

    Have been meaning ever since to set up a beefy QEMU instance and run a FreeDOS install, but real life and 3 kids keep getting in the way!

  20. It is sort of confusing how you keep jumping between computer and board games. Apparently they are pretty much the same category in your mind.

  21. Haven’t looked at DOSbox for a while … The last time I gave it a look I struck out trying to get Syndicate Wars running acceptably.

    *COUGH*.

    Honestly it’s almost as hard to get it running acceptably but every little bit helps with SWars.

  22. Master of Orion 1 & 2 are currently $6 with no DRM at gog.com, for anyone who can’t find their old MOO floppy disks (or who can’t find a floppy disk drive!) anymore. I haven’t bothered trying MOO2 yet, but MOO1 actually works better in DOSBox than it did in DOS: the XMS/EMS fiddling is done automagically rather than through a modified config.sys.

  23. I prefer the original.

    The second one had potential, but they increased all the micromanagent in an effort to make the game deeper, I guesd. All they accomplished was making it fiddly.

    MOO1 was a delight, even if the graphics look campy. MOO2 was a chore that may have appealed to tax preparers, but just lacked that indefinable quality: fun.

    I actually want to build my own 4x game that focuses more on what made moo1 so great and cute out

  24. I prefer the original.

    The second one had potential, but they increased all the micromanagent in an effort to make the game deeper, I guesd. All they accomplished was making it fiddly.

    MOO1 was a delight, even if the graphics look campy. MOO2 was a chore that may have appealed to tax preparers, but just lacked that indefinable quality: fun.

    1. >MOO1 was a delight, even if the graphics look campy. MOO2 was a chore that may have appealed to tax preparers, but just lacked that indefinable quality: fun.

      Have you tried out FreeOrion? If so, how much of MOO1 does it capture?

      (If your answer is encouraging I might give it some coding time.)

  25. As memory serves, I agree with Clay’s assessment. Most of the add in MOOII was needless complication, with the notable exception of multiplayer, which was clunky but entertaining.

  26. >Have you tried out FreeOrion? If so, how much of MOO1 does it capture?

    I can’t say how it compares to MOO1 as I’ve never played MOO1, but it’s still fairly incomplete (so that I’m not sure I could say how it compares to MOO3, which I have played). I’ve tried it once or twice and decided to look at it again when it’s further along.

  27. Have you tried out FreeOrion? If so, how much of MOO1 does it capture?

    (If your answer is encouraging I might give it some coding time.)

    I was just looking at their site with the same idea in mind. Looks quite promising. The project still seems pretty active.

    Sorry if this is inappropriate, but I’d like to mention that I am looking around for an open source project to contribute to, so if anybody has a demand for somebody with C and Python coding skills then I would be interested.

  28. I’ve looked at Free Orion’s website but never downloaded or compiled it so I can’t say what exactly is really there and how it compares in game play.

    For some reason, putting the fun in 4X games seems to be harder than in many other genres. There’s a delicate balance between playing a galactic emperor and playing a galactic accountant; unfortunately, most 4X games fail to find it.

  29. Eric, I’ll mention it to Gene, Tony, John K., and the others. I’ll expect a fair number of them are familiar with your work.

    1. >I’ll expect a fair number of them are familiar with your work.

      Well, I’m pretty familiar with their games :-) Not many companies are doing as good a job of addressing hard-core historical gamers as GMT. I’m particularly a fan of Borg’s Command & Colors games.

  30. @clay

    The problem is one of finding the meaningful decisions to be made, and being ruthless about minimizing any recordkeeping and overhead that detracts or distract from those decisions.

    You then need to carefully layer on combinatorics on those decisions. Too many combinatorics and someone has to have too much domain knowledge to access the game at all. Too few and the game has the potential to collapse into a puzzle or “insert recipe” problem.

    I feel that Dominion is a reasonable example of a game that collapses into a “recipe problem” with its base configuration.

  31. Eclipse sounds great, but Amazon says it costs $179. Is that right? One hundred seventy-nine U.S. dollars for a freaking board game?

  32. Somewhat surprisingly, the interface for MOO works very well on android. The only issue is space combat – the mouse over effects don’t work, so you have to know how far your ships can move.

    MOO is a little slow on tegra2, but it’s still quite playable.

  33. I had some problems with Eclipse. The first one is that dice play too much of a role. When taking an Ancient system early in the game, I lost due to unlucky dice rolls, and there was no way of catching up. This is because of my second annoyance with the game. Power development is exponential. Once behind, you stay behind. Thirdly, the game is quite unbalanced with 5 players. The two players next to the gap in the ring have more space to expand and one of them will win in a game of reasonably equal players.

    1. >Thirdly, the game is quite unbalanced with 5 players. The two players next to the gap in the ring have more space to expand and one of them will win in a game of reasonably equal players.

      That is definitely true. Our group looked at it and decided to stick with 4 or 6.

Leave a Reply to esr Cancel reply

Your email address will not be published. Required fields are marked *