Review: Agon (episodes 1, 2, 3) / The Mysterious Codex

Official game site; Private Moon Studios (creators).

Review written by Andrew Plotkin

Pretty good
Writing and dialogue
Pretty good, although English translation is imperfect
Board games
Fairly easy
Forgiveness rating
You cannot die or make a fatal mistake.

The promo quote on Agon's web site says, "The AGON Adventure Game Project is the world's first adventure game series." Which is confusing on first bite: how many games do you already own with "Two", "3", or "IV" in the title? But what they mean is, the world's first episodic adventure series. Planned as a unified arc of fourteen chapters.

Not actually the world's first, of course. You'll remember The Forgotten. Or maybe not. But that never got past the introductory episode, and Agon is now up to chapter three.

(The Quibbler-Within wonders if Zork 1, 2, 3 counts as an episodic series, playing out the unified story arc of Dungeon. Pardon me while I go administer some clobbering to the Quibbler-Within.)

Perhaps the Agon developers learned from The Forgotten's mistakes, because they've set up a very clear formal structure for their game. You are a professor of antiquities at the British Museum, in 1903. In the introduction (episode one), you receive a mysterious letter about some mysterious artifacts which have just arrived. The clues you discover in that episode send you scurrying across Europe... and each succeeding episode takes place in a new country. Fragments of background information are doled out as you proceed, but essentially each episode is a complete mini-adventure, whose goal is "find the next artifact".

To keep players interested in between episodes, the designers have added a twist: each episode (except the first) ends with an original board game. You can play against the computer, and you must win at least one match to actually get your hands on the artifact and finish that episode. There's also an Internet play application which lets you play the games against other humans. (Unfortunately, the Internet app is Windows-only. I haven't booted up the PC to try it.)

Does this structure work? Mostly. I'd say the episodes aren't hung together very well -- the designers went a little too far in making them stand-alone games. Yes, there's a background story to discover; but the fragments are so small and obscure that they fade into, well, the background. You wind up focussing on the scenery and puzzles of the episode you're playing, which don't tie at all into the preceding or following episodes.

And the board-game gimmick, while interesting, isn't really tied into the adventure game either. Sure, it's the central McGuffin. But nothing seems to explain why you're hunting for board games, as opposed to ping-pong paddles or samples of exotic fruit.

Mind you, we're only up to episode three. No doubt more will become clear in time. And I certainly have no objection to a sequence of short adventure games alternating with board games -- both aspects are fun. I just wish there was more of a sense of mystery and thematic cohesion. The Forgotten had those in buckets.

The adventure games, of themselves, are reasonably well-constructed little ditties. The puzzles are a nice mix of mechanisms, NPCs, pattern-matching, ciphers, research, common sense, and exploration. The full range of adventure game design, now that I think about it. The designers are clearly paying attention to their field of art.

The only flaw I see is a tendency towards the "search problem". Important stuff can be scattered in any nook or cranny of the landscape; which means you have to look around, everywhere, with great care. If you miss a spot, you'll probably miss an object -- which means going back to look everywhere again -- because it could be anywhere. Agon isn't the first game with this failure mode, and it isn't as bad as some, (and I like searching landscapes). But it's worth noting.

Oh, and the third episode has sound puzzles. (There's a sound mechanism in episode 2, but you don't actually have to hear the computer to solve it.) Again, worth noting.

Agon looks fine. The graphics aren't incredibly detailed, but the artists have taken the time to work in lots of background and detail animations. So the environments come off as modern-if-modest, as opposed to cheap or old-fashioned. (I particularly liked the gimmick of allowing your reflection to appear in polished surfaces -- just slightly animated. Most first-person games go to great lengths to avoid showing your reflection.) The background sounds are good; the background music is quite excellent, particularly in the second and third episode.

The interface is what you expect. Standard panning interface, with inventory bar and dialogue menus. I found a few places where hotspots appeared for plot-contrivance reasons, and a few more where an interactable spot wasn't marked with the proper cursor. Again, not the first game with these problems; and they weren't endemic.

A bigger nuisance was the walking "animations" -- which were not full animations; just fades to the new location, with a bit of zoom to give the effect of movement. They look fine, actually. But they're as slow as full transition animations, and there's no way to skip them. Each game requires a lot of walking around. I soon found myself wishing I could click-click-click through the locations, instead of waiting for each fade to complete. You, too, will learn to curse that stairwell in episode 1.

But, in counterpoint, I must admire an interface idea which I believe is original to Agon. And, as I so often do, I will admire it in the form of a verbose tangent about game design.

Many of Agon's puzzles, as I said, involve research, ciphers, and hidden messages. That is, they require you to discover information which will lead you further into the game. But information puzzles are an enduring pitfall in adventure game design. How do you ensure that the player actually gathers the clues and discerns the information -- as opposed to guessing, reading a walkthrough, or using brute force?

We have a handful of stock solutions to this problem. You can make the guessing space so large that brute force is impractical. (A six-digit combination lock, for example.) You can randomize the combination for each player, to foil walkthroughs. You can block the action at the interface level. (A voice-over says, "I don't know the combination! I'll have to search for it.") You can ignore the problem. (In Myst, you can actually skip all the clue-gathering that makes up the body of the game. If you know what to do, you can start it up, perform the necessary actions, and jump straight to the endgame.)

Indeed, Agon uses all these tactics at one time or another. (Except randomization, I think.) But there are some situations where none of them will serve. Consider a story in which the protagonist receives an encoded telegram; deciphers "MEET ME IN PUB AT SUNSET"; rushes off to fateful encounter. A perfectly good narrative incident, but how do you render it in an adventure? The pub has to be a location which the player can reach. Exploring the game's locations is the mildest of brute force; every player will eventually wander into the pub, whether he's deciphered the message or not. Saying "The sender won't be present unless the telegram is deciphered", or "I won't let the player inside", only highlights the problem. You can know when the player has acquired all the information he needs to figure out the puzzle; but you can't know when he's actually done the figuring.

(Note: this pub example is not a spoiler. It's my made-up example. Just mentioning.)

In Agon, when you pick up the telegram (or whatever), the game pops up a text entry prompt, and an on-screen keyboard. You get to type in the solution, the message, or whatever information you were supposed to discover. (The on-screen keyboard is annoying, but by Episode 3 the designers have wised up and allowed real keyboard typing also.) When you enter it correctly, the game proceeds.

So here we have an interface mechanism which cheerily violates one of the basic rules of computer gaming: it lets the protagonist diverge from the player. In most puzzles, you identify with every move the protagonist makes. You are performing those actions. The game does as little as possible to interfere with that sense of identity. But with these form puzzles, the protagonist explicitly doesn't decipher the message at the same time you do. You solve the puzzle first; then you type it in, and hear the protagonist say "Aha, the pub! I'll go there right away!" And then of course the messenger is at the pub, solely because you filled in the form, which is a further mimetic hole.

And is this a problem? Not at all. It's not a perfectly smooth game mechanism -- you can see the plot levers, as it were -- but it didn't feel unnatural either. It certainly didn't wrench my sense of immersion. I said "Aha, now I must solve this puzzle," which is a common feeling in an adventure game, and I got on with the solving. Perhaps in an extremely immersive, solid-feeling game, the forms would have bothered me. But Agon already has this game-like structure with the chapters and the board games and all; the forms didn't stick out.

(For the text adventure fans reading this: I think this interface would be perfectly workable in text IF. We've already got text input prompts, right? Having the game switch from the usual action prompt to "Enter deciphered message:" should be possible. But I can't think of a game that's tried it. Someone experiment!)

Enough about adventure game design. Let me comment a bit about board game design, and then I'll let you go.

Two of the Agon games are now available (those in episodes 2 and 3). Both seemed adequately challenging; I lost three or five matches, and then managed to win. That's about as much time as I wanted to spend on them. (There is an "easy" board-game mode, which I assume makes the computer player stupider.)

Both games are pure strategy -- no randomness, no hidden information. Moreover, they are both... not very different from chess. That is, they both give the feeling of taking the assumptions of chess-like games for granted. Two players; taking turns; alternating; each player owns some pieces; pieces are points; the board is a fixed grid; pieces move; pieces can capture enemy pieces.

I certainly don't mean these games are similar to chess. One is a fox-and-hounds sort of game, where you are trying to move your king off the board and the computer is trying to stop you. The other is closer to checkers -- capture all the enemy pieces -- but with a high-power capture mechanic which can decimate the board very quickly. Both were fun. But I wonder if the authors will look at the very wide range of strategy board games out there, and allow their designs to push more limits in the future. I could name games that violate each of the assumptions I mention above.

You could certainly try both of the board games in real life. Find a friend, draw out a board on paper, plunk down some colored beads, and hit it. I'm not sure how well they'd play, though. (As I said, I haven't tried the Internet person-to-person Agon application.) Both games are asymmetrical. Which I suppose isn't surprising, since they're designed to give the computer a strong shot at beating a human, without hours-long move computations. The fox-and-hounds form is inherently asymmetrical; the second game has symmetrical rules, but the computer always moves second, and the second player seems to have a hefty advantage. In real-life play, I'd suggest a pair of games, and trade off who goes first.

I must cover one more topic, which is a sad one: the frequent bugs that I ran into playing the Mac version of episode 3. Something about the conversation interface was very flaky. Sometimes the voice-over sound would cut out; sometimes the dialogue window would get stuck on-screen after a conversation ended. No matter what the immediate symptom was, the game would wind up in a broken state. The hotspot, interaction, or dialogue option needed to continue would be missing. No way to keep playing; I had to reload an earlier saved game. (The nastiest instance was right after winning the board game at the end! The game got stuck on the penultimate conversation, and refused to show me the ending.)

Even worse, the game allows you to save when it's in this broken state. The resulting save file is corrupt. So the lesson is, if you're playing on the Mac, save frequently in episode 3, and keep as many saved games as possible. Save before every conversation. If you see evidence of the bug, do not save, but quit and reload the last clean save.

I hate having to write paragraphs like that.

So, will Agon succeed? Naturally I have no clue. I played three short adventures for $10 each, spending part of an evening on each. It adds up to a medium-sized game's worth of adventuring, which is certainly worth thirty bucks. With two original board games thrown in.

The Agon designers released two episodes at the end of 2003. Their original plan was to take the revenues from those, hire more developers, and start releasing an episode every two months for the next two years. That was a "challenging" schedule, as project managers like to say. Unfortunately (but probably not surprisingly), it didn't work out. A note posted in January 2004 indicated that they had not been able to hire more people, but episode 3 would still appear in spring 2004. Which turned into summer, which turned into September. Episode 4 does not yet have a release date.

So, realistically, you can expect an episode every nine months, for as long as they can hold it together. I hope they manage to finish the series -- even if it does take until 2012. I'll have spent $140 by then. (Assuming the US dollar doesn't collapse first, that is. The developers are Hungarian, and are actually charging 2000 Hungarian forints per episode. Who knows what that'll be in ten years.) If they get that far, it'll be the world's first successful episodic graphical adventure game series, and they'll certainly deserve the money.

Summary: A promising start to a long puzzle-game series. Even if episode 4 never appears, the first three are a satisfying experience.

Game Reviews