Ra is an ancient-Egyptian-themed, press-your-luck auction game by Renier Knizia. The rules says that you influence Pharaohs and pay homage to gods, but Ra is thinly themed. Regardless, the simple game mechanics and tough decisions make this game a perennial favorite.
There are three Epochs (rounds) of play. Each Epoch you have three uniquely numbered Sun (bidding) tokens, ranging from 2 to 16, that are fairly distributed according to number of players. For instance in a 5 player game, one player will get tokens 2, 7, and 16, while another will get 3, 8, and 15. Every turn each player makes a bid on the existing pot of tiles (call Ra), or draws blindly and adds one tile from the bag. If a player calls Ra, a Ra tile is drawn from the bag, or all eight tile slots are full, then a once-around English auction begins, the active player having the final bid. An Epoch ends when 10 Ra tiles are drawn.
You need to collect elaborate sets of tiles from eight Monuments, five Civilizations, Pharaohs, Nile, and more. There are several negative tiles mixed into the bag that can destroy your existing tiles, while having the low quantity Pharaoh or Civilization tiles will also give you negative points. Another twist is that the bid token you use is added to next pot of tiles. If you bid on a small pot you can get it at low price. On the other hand, if you grow the pot, you may get the specific tiles you are looking for; however, larger pot also tempt bids from more players. Thus you must make this choice each turn: grow the pot (risk) or bid now (safe)?
Ra is available as a well-implemented free fanware by Gabriel Rocklin; with an online multiplayer version on Gourmet Gaming by Kevin Bohlmann; or as a universal iOS app by Codito Development.
Ra!
Ra is clearly a good game... but it has never quite sat right with me. I think the problem is how well it rewards careful consideration. It seems like to play seriously you need to consider out how much benefit each player would receive for a set of tiles, and I find doing so time consuming and not fun. But if I don't do that, the game feels random and unsatisfying.
For me, it has kind of been replaced by 7 Wonders. 7 Wonders is lighter clearly, but it's still a reasonably deep strategic game and the returns for mega-analysis diminish much faster (especially if you're playing with enough players that most cards don't make it all the way around). I also find there's much more clarity in the game state and scoring (scoring in 7 Wonders was, I assume, influenced heavily by Ra - but it feels more streamlined to me).
So yeah, I appreciate it objectively as a good game, but for me to enjoy it I think it would need some sort of information barrier preventing analysis paralysis.