I was very pleased to hear that the GDC folks posted my 2014 talk for free on the Vault, so everyone can watch it now. I was actually quite surprised during the talk that my content only lasted 40 minutes – I had well over 100 slides, and the presentation was a much expanded version of my 15-minute mini-talk from PRACTICE. The downside is that I incorrectly thought that I wouldn’t have the time to address games that run counter to transparency, including dynamic ones such as Europa Universalis or Out of the Park. The upside is that plenty of people brought up this omission in the Q&A, and I did the best I could on the spot to clarify the value of transparency for different types of games. (I’d point to my answer to Randy Smith – who helped design Thief‘s arrow grammar! – at the 49:00 mark.)
Basically, games like EU are so complex that it’s almost impossible for them to have a transparent rule system because the explanations would probably have to be done as algorithms. Nonetheless, for dynamic, rules-driven games, transparency is still a goal to reach towards; in Crusader Kings 2, the characters’ attitudes towards the player character are broken down in explicit detail (“Long Reign: +8, Liege is Just: +10, Desires Duchy of Ostlandet: -20”) to open a window into a very complicated algorithm. For Civ 4, we similarly exposed the diplomatic AI by detailing all the factors that determine a specific ruler’s attitude towards the player, which increased player comfort significantly. I mentioned that system in the talk, but what I failed to mention is that the designers of Civ 5 tried to move back to a more opaque system by hiding these numerical values but had to patch it back in because players felt too lost dealing with the AI. Ultimately, transparency is a specific game design aesthetic that may not right for all games, especially ones which value mystery, discovery, and perhaps even player frustration, but if a game is meant to be dynamic and replayable, then transparency is a very important goal.
Great talk! For your last question (about if you expose to many rules then it might become solvable), I feel that it would’ve been a good opportunity to tie your answer back into what you saying earlier in the talk about “pre-randomness” (or pre-luck), like how you might have some parts of the game are given to you randomized (like all the tiles and placements of everything in Civ) and how that pre-randomization prevents, or decently delays, the game from becoming solvable.