- No clue whether you guys will find this interesting, but I did so I figured I d kick the results out there for everyone else to see. I always felt like thereMessage 1 of 1 , Feb 21, 2012View SourceNo clue whether you guys will find this interesting, but I did so I figured I'd kick the results out there for everyone else to see.I always felt like there was some streakiness to the random number generator in the version of Empire Deluxe I play (it's the 1993 version) - specifically, it feels as though there's an equalizer build in such that the game will balance out kill-death ratios on a per-player basis, resulting in some wacky streakiness in small/tight games (where such a bonus could get reinforced by looping). Gut feeling. Definitely wouldn't rule out confirmation bias, but I have seen combat streaks that were simply to long to be believed.Since I was curious, I decided to keep track on a series of solo games. I settled on 20x20, wrap enabled, standard ED rules (so including armor units/bombers/air fields), default production/combat bonuses, five expert level computer players. As an end point, I was to keep playing until I hit a six-game streak (satisfied by wins or losses). Results below. Spoiler - I failed to determine whether there is an equalizing element to the random number generator, though I still suspect there is.
· it took 266 games.
· In the course of doing this I played five elimination games (ie, five streaks of five): three 5-wins, two 5-losses. I finally won a game six on the third and final win streak. The mean duration for an exercise like this (assuming a true coin-flip/longest-case scenario) is approximately 2^n (where n is the consecutive wins/losses typically needed to end it); in this case, just 64 games. It took me more than 4x that number.
· final record was 132-134.
· At one point I was -23 in W/L. I went 74-53 over the final 127.
· I made a very small adjustment at around the halfway point - switched from outside-edge to optimal-reveal searches during early exploration with land units. It made a big difference.
· I had 132 first place finishes, 43 sixth place finishes, 31 fourth place finishes, 31 fifth place finishes, 15 third place finishes and 8 second place finishes.
· My mean "territory explored" value was 55.6%. Mean on wins: 90.0%. Mean on losses: 21.6%
· My mean kill/death ratio was 3.6 to 1 on a per game basis. Overall it was 3.2 to 1. In wins I was 5.4 to 1. Even in losses I was 1.8 to 1.
· My highest kill/death ratio game had me killing 360 production-turns worth of enemy units at the cost of a single (1) infantry unit, for a kill/death ratio of 60 to 1. This was accomplished with capital ships over the course of 201 turns.
· An average game was 137 turns. Wins averaged 196 turns, losses 79. My longest game was 551 turns. The shortest was 8 turns (happened four times).
· In games over 200 turns I was only 56-13 (each of those losses was infuriating).
· Quickest win was 45 turns. I started on an 8-city island with all but one of the computer players, who immediately resigned once I took the starting island.
· My units had a survival rate of almost precisely 1 in 3 on a per game basis and 1 in 2 overall (that is, the mean per-game survival rate was 33% (due to nearly half my games ending with 0 units), but my overall survival rate was 50% when you take total units produced vs total units lost).
· On four different occasions I lost games with two cruisers in the water. In one of those losses I had two cruisers plus an empty transport.
· I never built a carrier, nor an airbase.
· I didn't track my losses on a per-unit basis (wish I had), but I did keep track of what my cities were working on at the end of games. Here are the totals: 453 armor units, 354 nothings (often to mitigate the production penalty), 153 infantry units, 35 transports, 38 cruisers, 14 fighters, 11 bombers, 11 submarines, 10 destroyers, 9 battleships. Note that these numbers are not representative of my overall production, but rather the units I was producing at the end of winning games.
· I averaged just under 1.4 in production-ticks-per-turn on a per game basis (put another way, I'm averaging 1.4 city-turns of production overall - that's lumping together the wins and losses and considering an 8-turn game equally as relevant as a 551-turn game). The actual rate on a per-turn basis was 1.51 production turns (I'm deriving this value from total-unit-cost-produced-divided-by-turn-duration-of-game, but the ratios I'm presenting are slightly deceiving - in both cases we're losing all units expended in city capture (as far as the game stats are concerned those units cease to have existed), but we're also gaining a 1/6th production bonus from units subsequently produced in a given city - I suspect it's approximately a wash once you add in the end-game incomplete units).
· Keeping in mind the hinky record keeping, I recorded 55,016 production-turns worth of units (a little over 9k worth of infantry units) and destroyed 87,366 production turns worth of enemy units (14.5k worth of infantry units, though more like 7.3k worth of bombers).
· In losses, my production-ticks-per-turn was just under 0.8. You wouldn't think it'd be possible to average this much under 1.0, however the AI loves bombers. Even in wins my production average was only 2.02. There are several reasons for this. Many of the games (110 of them) had me starting on 1- or 2-city islands. Since everyone is so close together, it's often necessary to build numerous planes, cruisers and infantry units before the bombing raids settle enough to allow for transports. Most of my wins involved treading water with two or three cities for 150 turns before taking over the rest of the map 15 turns after producing my first transport (many of my losses went the same way except with me finally succumbing to the nonstop bombing raids).
· Every 20x20 game starts with 6 occupied cities and 4 neutral ones. Of my 266 games, 71 had me starting on single-city islands. Of the remainder, 39 were 2-city, 29 were 3-city, 20 were 4-city, 26 were 5-city, 19 were 6-city, 27 were 7-city, 19 were 8-city, 10 were 9-city and 6 were all-10-cities.
· In 98 games I started alone on an island. I was 66-32 in those games. I managed to lose lose 4 of 19 games in which I started alone on 2-city islands.
· In 6 games where all the cities on the map were on a single island I went 1-5.
· An average start had me on an island with 3.95 cities on it, 1.30 of which were neutral.
· The most populous no-enemy island I started on had three neutral cities. I won that game in 228 turns.
· Aside from a few moments of violent nerd-rage, at no point did I stop having fun.