As gamers, we are often quick to chastise the industry for too much sameness and a "me too" risk-aversion mentality among publishers. At actual game development studios, of course, game designers often want to innovate, but are careful not to do too much differently, because everything new (and thus untested) carries some design risk. Unless you have an infinite budget, you simply can't take too many risks, because each one has a very real cost in testing and iteration time. We rarely see games that are too innovative, because those games generally run out of money before release.
In the boardgame world, costs are three orders of magnitude lower, so in theory we can see the occasional game that innovates a little too much for its own good. Duel of Ages is a case in point: a game with a lot of very interesting experiments going on, some of which succeed while others fall flat, resulting in a game that is worthy of study by game designers even if it's not necessarily worth playing purely for entertainment.
The premise of the game is that in the future, there is this Holodeck-inspired battle arena where anyone, real or imagined, past or present, can be brought to life. I suspect it is the inherent anachronisms that were the original seed for the idea, and the game carries it well: imagine Ghengis Khan with a rocket pack facing off against Annie Oakley with a laser pistol, and you get the idea.
On the surface, the mechanics look like a simplified chits-and-hexes wargame: You've got individual units with equipment on a hex, there's movement, terrain, a simple zone-of-control system, and attacks are handled through a color-coded stat system where you compare one of the attacker's stats to one of the defender's and this gives you a target number on 2d6. Most characters are hardy and can take a few hits before dying, but kills are possible. For the most part, the combat system takes a fairly intimidating set of math formulas and makes it a bit more accessible using a universal lookup table that you'll have memorized by the end of your first game anyway.
The object of the game, however, is not to kill the opposing team (well, in theory you could win that way, but combat is balanced in such a way that it is exceedingly unlikely). Rather, there are 5 potential victory points, and "having the most people alive at the end of the game" is one of them. The other four are these independent scoring tracks on the board that can only be advanced by maneuvering your characters to certain key locations, and then attempting a challenge (another die roll against a random stat, and win or lose, your character is teleported away so they have to walk all the way back to try again). Whoever is further ahead on each track gets one of the victory points. It is therefore control and advancement of these four strategic locations that has more of an effect on the game than the actual combat, making this play like a really strange military almost-non-combat game. While I appreciate a wargame with an alternate goal to "just shoot until everything around you is dead," the sheer amount of die-rolling (particularly against arbitrary stats) makes the game feel like the focus is more on luck than strategy.
Game length is not actually stated in the rules, other than that players should agree to a set length in advance. You can therefore play for half an hour or four hours or whatever, and when the timer reaches zero, you count up victory points and determine a winner... which again is different from most games that have a more explicitly-defined victory condition. (This also means a player could theoretically stall the game intentionally, so I preferred playing to a set number of turns rather than a set amount of time.) While this does let you tailor the length of the game to the amount of time you actually have to play, it feels a little awkward to be struggling towards a goal and knowing that no matter what, you can't possibly reach it until a timer runs out. It also means that unless the game is very close, the endgame feels like the game is already decided; victory points just don't transfer from one player to another that quickly or easily. This would be one of those new things that didn't quite work out.
The player structure is a stroke of brilliance that I hope to some day see in other games. Rather than the more typical free-for-all of tabletop board games, players are divided into exactly two teams, with each team given a set number of character tokens on the map. A team collectively wins or loses, and all players on the team work together; they can collaborate in deciding which characters move where (a "play-by-committee"), or more likely, they can assign specific characters to individual players and you each move your own. This has the side benefit of free entry and free exit to the game. Say a friend walks by and asks what you're playing; "here, let me give you one of my characters and we'll explain as we go" is a perfectly valid response. Likewise, if you have to leave early, it's not a game-killer; just hand off your characters to your teammates and they can continue where you left off. As another side effect, the two-team structure essentially means that Duel of Ages is one of the few games that could reasonably claim "number of players: 2 to infinity."
The last area of innovation is the business model. Apparently, the game that the designer really wanted to make would have retailed at around $200 or so... a steep price for any tabletop game that isn't owned by Wizards of the Coast or Games Workshop. Rather than scale down the gameplay or reduce the core rule set, however, the game was instead broken down into a single base set and a pile of expansions, all co-released from day one. As such, playing the core set only gives the feeling that something is lacking... largely because it is. On the other hand, playing the game the way the designer intended requires an expense that most gamers would only be likely to pay if the core set was sufficiently satisfying, so it becomes something of a catch-22. It's not unlike a board game equivalent of a "crippleware" video game, trying to find that fine line where you give the player enough to make them want to buy everything, without giving such an entertaining experience that the "demo" satisfies the player's need. Except that in this case, the demo still costs forty bucks. Oops. So... interesting business model, but a swing and a miss in perhaps the most critical area.
The big takeaway I get while playing this game, is the importance of only innovating in one or two key chosen areas when making a new game. Too many new things means some of them will not work the way you hoped they would, and the result is a game with a lot of potential, a gold mine for new ideas, but an unsatisfying overall experience because half of it just doesn't seem to work right.