I've been musing over a moderately complicated alternative to the standard Treachery implementation. I don't like the standard Treachery rules at all, basically because of poor balance.
None of the decks of cards in Descent are particularly well-balanced (i.e., at a similar power level to each other), but in most cases it doesn't really matter much, since the cards are drawn randomly and you have to make do with what you get. The only exception, other than Treachery cards, is Shop Items. The Shop Item deck is quite small, and with a slight blip in the Ranged Weapons, is mostly one of the best balanced decks in the game. It's also only really used in the initial stages of the game, before the heroes get more powerful weapons.
Treachery cards (and indeed Overlord cards in general) are horribly balanced, however. They range from completely useless (a surprising number- I reckon somewhere of the order of one third of all Overlord cards in existence are so bad you'd play them less than 5% of the time you drew them) to ludicrously powerful (absolute must-plays if drawn). Treachery cost has next-to-no correlation with power level, either: compare Lone Golem with Lone Troll for a particularly stupid example. Consequently, the Treachery mechanic is extremely dull, as the same small number of cards get used over and over again (Crushing Blow, Danger, Ambush, Elite Beastman War Party, Lone Troll, Drinkers of Blood, Poison Pit to name some of the cards I consider the most powerful: I don't think anyone else's list would be radically different).
Proposed variant, therefore:
Separate all Overlord cards (with and without Treachery costs) into three piles: Traps, Spawns and Events. Power cards go in the pile corresponding to their Treachery cost; Evil Genius goes into Events, Brilliant Commander and Hordes of the Things into Spawns). Shuffle all three piles separately.
Go through each pile in turn. From each pile, randomly draw, and then turn face up to inspect, cards equal to your Treachery value for that type of card times 4. So with 3 Spawn Treachery, draw 12 Spawn cards. Sort these cards into two piles: cards you want to keep, and cards you want to discard. You may not keep more than 16 cards, even if your Treachery is 5 or higher, but you may discard all cards if you wish. Having done this, increase the size of the "keep" pile to 16 by dealing further cards randomly from the top of the corresponding shuffled deck. Remove all discarded cards and the remainder of the shuffled deck from the game.
You now have three piles of 16 cards each, each pile partially chosen and partially random. Shuffle the three piles together to form your Overlord deck.
Note that the actual Treachery cost of each card is completely ignored: this is intentional.
You will, of course, end up with a deck containing far more cards with Treachery costs than normal (probably the majority). I actually consider this a bonus, for the following reasons:
The average power level of Treachery cards is not, I believe, much greater than that of non-Treachery Overlord cards (witness, for a start, that the best Power in the game has no Treachery cost)- there are of course a small number of exceptions, but it's exploitation of these exceptions that I'm trying to avoid.
You get a *much* more varied Overlord deck each game.
Thoughts? "4" may not be the correct multiplier, I admit: I'm wondering whether 3 or even 2 would work better. I'm also consider whether some sort of limit to the number of Power cards selected would be useful, not because I consider Power cards particularly powerful in play, but because they do generate a lot of Threat when discarded.