I've never played the computer game this is based on, the this game seems to capture some of the idea of the real-time strategy game. You start off with very little, try to build up your production base, create combat units, and go fight the other players. The box is absolutely crammed with plastic miniatures that don't really add much to the play of the game. But they are fun to look at.
I quite enjoyed the production and building aspect of the game. You have three actions every turn. Each action requires a card, and you draw 4 or more cards, splitting between face up cards and more powerful random cards. It is fun picking your actions and deciding whether to gamble on random actions that might be more powerful or might not be useful. The three actions you get is adequate but never enough, just about the balance you want in this sort of game. You can use actions to explore for land to give you better production, or to produce, purchase things, attack, or exchange resources. Each actions seems interesting in its own way. In fact, if you aren't feeling very violent, it is tempting to skip combat entirely and just have fun building up.
Purchasing the combat units is neat. Each player is a different culture and has unique combat units with unique costs and abilities. Some of the units are clearly better than other units of the same cost, but this is not a problem for two reasons. First, each unit requires a different mix of the 4 resource types, so you generally just try to buy what you can afford with what you have. Second, the units have a rock-paper-scissors circular superiority, where each unit is super-good against one or more other unit types. Normally this isn't my favorite game mechanic, but here it works well.
You can send your units to attack other players. What I really like is the idea of limited warfare. Only a portion of each player's forces are involved in a battle, and either side can easily retreat. The attacker can obtain a certain fixed benefit if he wins. So there are definite reasons to attack instead of sitting back, and you can lose a fight without losing the game. So in a three-player game, if everyone is out for himself, you can end up with end up with lots of battles but no one getting knocked out.
On the down side, I'm not sure the game balance goes far enough in this direction. It really helps for the players to want a three-way battle. Otherwise, you could easily end up with players either refusing to fight or teaming up to demolish a player they don't like.
A negative factor is the length of this game. This is a really, really long game, hard to finish in a single play session. Combat takes forever, as you match up the monsters in your army one by one and roll out who wins each fight. Normally, I am not bothered by lengthy combat systems, and I think rolling lots of dice is fun. But this game still gives me pause. You just roll, and roll, and roll. My hand got fatigued from all that rolling; I can't remember that happening to me in any other game. Maybe it was because the dice are so big. Also, before each matchup, each side secretly picks a unit to fight. If you carefully calculate what you think is the best unit to use before each and every matchup, I found the combat quite tedious. I greatly preferred just choosing a unit at random.
There seems to have been a game balance mistake with one type of monster ability. Some Norse units have the "berserk" power, which gives them the option to roll two more combat dice at the cost of losing all ties. This is generally treated as if it were a useful power. But mathematically, losing all ties is approximately equivalent to losing 3 combat dice, so going berserk is normally quite bad. The power is slightly useful if you are badly outmatched, but in general it is a trap for the unwary, and I don't think that was the intent. Conversely, the medusa's ability to win all ties makes it an extraordinary good unit, far better than any unit of its cost in the game.
My least favorite part of the game was the victory conditions. Many modern games try to stand out with unusual methods for determining victory. But this one I just found to be downright unpleasant. Each turn, each player puts a victory point on one of 4 victory conditions. For three of the conditions, the victory points are given out all-or-nothing to the winner of that condition at the end of the game. Near the start of the game, you have to decide what condition to make more valuable, without having any idea who will ultimately win that condition. I don't like making choices like this, which are simultaneously pointless (since you have nothing to base them on) and vital (since they determine the winner). At the end of the game, the mechanic ultimately means that unless one player scores a dominant victory (in which case this mechanic is a moot point), the winner is determined by a sort of vote, by where the players put the victory points. So if two players are vying for one victory condition, they can shut out the third player, unless they tie with each other, in which case the third player wins. I find the whole mechanic weird and don't see how anything fun can come out of it. I'd much rather have a conventional linear scoring system.
In general, I liked a lot of this game. As an inveterate game rule designer and tinkerer, I felt it was a good fixer-upper. I just need to make a new way to determine victory and a few other minor changes. The only problem is that it is so long, it will be hard to get anyone to play it with me.