Njet!, a 1997 design by Stefan Dorra, is yet another trick-taking game with a twist. The twist in this case is that the game includes a small board depicting icons for each player (you are provided a matching identity card at the start of the game), each suit, each possible "supertrump," each possible number of cards that can be discarded before playing a hand (0-2), and each possible number of points that each trick and point card can be worth (1-4). Wooden markers are provided, enough to cover all but one icon in each row. At the start of each hand, after cards are dealt, players take turns covering an icon with a marker, thus insuring that that option will NOT be effective for that hand (hence, the title).
The deck consists of 40 cards in four Soviet-themed suits - two 1s and 2-9 in each suit. All cards are dealt each hand (with three players, the 2s are removed). After all the markers are placed, the player whose icon is left uncovered (the start player) gets to make a decision. In the four-player game, he or she chooses a partner - the other two players are automatically partners. In the three-player game, the start player can choose to play alone instead (and if he or she chooses a partner, the remaining player plays alone). A lone player doubles his or her score for the hand. After this decision is made, players discard cards if necessary, and then the start player leads to the first trick.
The rules of trick play are fairly standard - follow the lead if possible, otherwise trump or renege at your option. If a trump is led, one must follow with trump if possible. There can also be supertrumps, which are always the two 1s of a suit (not necessarily the same as the trump suit, though they can be). These essentially form one big suit together with the regular trump suits - leading one calls out the other, and the supertrumps are no longer considered part of their normal suit (Mu players will find this all familiar). One additional twist here is that the second 1 of the same suit played to a trick wins, not the first (which is the more usual rule).
At the end of a hand each partnership gets a base score of 1 for every trick taken AND for every 1 in those tricks. This base score is multiplied by the number chosen on the board (and further doubled for the lone player in a three-player game). Both members of a partnership get the same score. Eight hands are played, and the player with the highest score wins (but you were expecting that).
The scoring is the game's biggest problem. In the eighth hand, unless some of the players happen to be tied, it is guaranteed that one (in a three-player game) or two (in a four-player game) players will have no shot at winning, because their partners have higher scores than they do. I've certainly seen this happen in other trick-taking games, but in none of them will it inevitably happen every game.
So, is there any way to fix this problem? Maybe. I can think of several possibilities:
I think the last two would probably have to be combined - if you use the first only, a player partnered with someone who has a higher score and could potentially win on that hand has an incentive to sabotage the partnership. If you use only the second option, the start player will always choose a partner on the last hand if there's anyone lower-scoring around.)
Bear in mind that I haven't tested any of these, and I probably won't. Even if the scoring system were fixed, I wouldn't be that excited about the game. It seems clever at first blush, but after a while I found that the marker-placing system seems to result in fairly mediocre trick-play. If you have a hand strong in, say, sickles (I said it was Soviet-themed), it's a fairly good bet that someone else will eliminate sickles as trump, since the other players must be relatively weak in it. Now this certainly means that someone who by luck gets a hand which is very strong in one suit won't be able to run away with the hand, which is good, but I prefer other games' methods of controlling this (bidding for contracts, so that you have to take on a lot of risk to get a high reward, no matter how good your hand is, to name just the most common example).
In summation - Njet! is a game with a clever idea, but that idea seems to work better in theory than in practice. There are better trick-taking games out there, certainly.