As a game creator and lifelong player, I am pleased beyond expectations with FFG's new LotR: The Card Game. I have enjoyed engaging its challenges with friends and alone and want to up the ante. As they only primed the pump for those of us interested in an "expert game", I thought I'd share what I have come up with. There have been a couple of threads glancing along the surface, but I've given this a lot of thought. (I'll be exhaustive up front, with a summary at the end.)
My key considerations were deck states, board state and player threat. I'll show the logic behind my thoughts.
Deck States
I like reshuffling the discard pile back into the deck. The passage of time and the reduction of dangers along the way (Ungoliant's Spawn being slain) eases the arrival more allies and supplies. I do not, however, shuffle player hands or cards in play back into the deck. Why would your allies head back just to be recruited again? Why would Gimli drop those awesome axes he found? I also, therefore, do not draw a new hand of six or allow mulligans.
I then set up the encounter deck, by taking all unused sets out of the discard pile and deck and adding the new sets. I then shuffle the discard pile back into the new scenario's deck the same way as with player decks.
Board State
We've handled the player side of the table state, but there are cards from the encounter deck about. To err on the side of simplicity I simply remove cards not built into the next scenario (ie. the Orc set sticks around the whole game; all others phase out as you leave their region). How can you "Journey Down the Anduin" if one or more of your heroes is "Caught in a Web" back in Mirkwood? Certainly your party will secure each other's safety rather than leave someone behind. As a benefit to beating a scenario, your characters are freed from the penalties and perils unique to it.
Player Threat
Three scenarios without hitting 50 threat? Yeah right. Using the same logic as with "Cards in Play" I assume some player threat will drop as the foes of the previous scenario are no longer chasing you. It's hard to pick a number, since player threat can vary so much from game to game and deck to deck. In trying to find a consistent and reasonable rule of thumb I decided to use an equation similar to questing. I take the combined will power of all characters in play minus the total threat in play from the encounter deck (both engaged and staged).
Each character's "will" to be rid of the threats around and get back to their main objective is only counteracted by the current threats acting as a beacon to enemies yet to come.
Good players may be able to come out of Mirkwood with a lower threat score than they started with. But, since threat is really just a hard cap to losing the game (ie. not the only way) it doesn't seem terrible to mitigate its effect over the long, "Expert" road. Indeed, this possibility allows for truly expert players to gain an advantage where newcomers would not. And anyways, insurmountable enemies or a lack of a deck will still lead to a slow, but sure demise. A game lost due to player threat is just quicker and more exacting.
In the end, I just can't justify the seemingly unbeatable Nightmare suggestion of leaving player threat intact from scenario to scenario while maintaining a loss cap of 50.