In addition to Realms Beyond Civilization events, very occasionally I play a solo game of Civilization IV. This is a report of one.
The genesis of this goes all the way back to the first Sid Meier's Civilization. I played Civ 1 starting in high school, after someone had put a copy on one of the PCs in the school computer lab. (DOS 486 machines, they would be. ) Naturally I copied it to take home, and learned to play and love it. A few years later I also found and downloaded a copy of Civ 2 from a local BBS. (Hey, I was young and stupid. I buy my games now.)
Now, the only thing I knew to do with Civ games was to play for a high score. That's just what video games are naturally there to do, dating from the likes of Space Invaders and Pac-man, and even before that to pinball games. So I gradually figured out how to work the highest scores. In a nutshell, population was the biggest factor: the secret was to get your cities into We Love The King Day (at least 1/2 citizens happy and none unhappy) which makes them grow every single turn. Conquer the AIs down to one city and milk forever, with no domination win condition to interrupt you.
The only problem was the massive amount of time and micromanagement involved. Civ 2's engineer units (workers) had the ability to transform any tile between peak - hill - plains - forest - grassland, meaning that the highest score meant converting every single tile on the map to grassland, plus farming and railroading. And you had to check every city every turn to see if it would grow an unhappy citizen and fall out of WLTKD.
I spent most of summer 1997 attempting this on a single game of Civ 2, easily 200 hours on one game... and still only completed about half the map before burning out. I also attempted the same thing in Sid Meier's Alpha Centauri, with predictable burnout results.
Fortunately, Sirian and Realms Beyond emerged to show how much more Civ games can offer besides simply high scoring. And Firaxis themselves fixed the score milking problem for good in Civ 4, introducing the normalization formula to rate skilled gameplay by extrapolating.
Still, I've never really played for a high score game in Civ 4. I've kept a halfway curious eye on my score results, with the highest coming in games that didn't count fairly (350k with Advanced Start in Arise Ethiopia, 322k with Sid's Sushi in Something Fishy.)
And I've wanted to deliver one magnum opus game of Civ 4 ever since Pax Americana didn't really execute on that front for me. I was simply overwhelmed trying to click through 500 turns, coupled with some severe PC stability problems, and didn't give that huge map game the effort it deserved.
And I happened to come across again and re-read the famous guide by VirusMonster on high scores, which demonstrated the first recorded 500,000 score and paved the way for players later topping 1,000,000. Finally, there was a break for me in the Realms Beyond schedule when I wasn't looking to jump into Fall from Heaven, so I embarked on this instead.
The general approach is clear: conquer up to the domination limit, milk population for a short time, as long as the victory score continues to grow, then claim the victory. I decided to follow the CFC HOF rules for map setup and game options, which seems like a reasonable baseline, even though I'm not modding the game to fit under their rules.
For the game setup, I decided to simply follow most of VirusMonster's guide suggestions. Julius Caesar of Rome, since the Praetorian is broken and Organized is easily the best economic trait for a warring game on high difficulty. Immortal difficulty does seem like the sweet spot for score multiplier (Deity seriously roids the AI bonuses for only 1/9 more score.) And yes, Marathon speed for all its well-known advantages. I am only going to do this once ever and I think I can tolerate one game of Marathon.
Pangaea, low sea level, natural shoreline (more water in play means more population). Tropical climate for more jungle grassland and less ice. And yes, huge map. No Barbarians, since that is allowed by HOF rules - the most important effect is to allow stolen workers to reach home safely. Maximum AIs by HOF rules (17) - more AIs means closer targets to conquer, more tech in play, and more start-normalized spots.
I am certainly not trying to milk for every single possible point - build cities in every tiny little crack, as densely as possible to get corporation food in more cities. I'm happy to aim for a showing of "pretty close" on high score. I think I can match VirusMonster's 500,000 score, and maybe push a bit beyond that, 750k would be certainly satisfying. (See VirusMonster's thread for a detailed discussion of the normalization formula.)
The Report