Map size and the cost of research

[samson]

The cost of research varies with map size when selecting the New World or Customize World options. The cost difference is noticeable on the second researched tech. For small map this is 18 beakers, medium map is 22 beakers and large map is 26 beakers. This relative difference maintains throughout the tech progression. On the 20th tech and above the minimum cost on a small map is 19X, medium map is 24X, and large map is 30X.

However, this cost difference does not hold true for Pre-made Worlds (edited maps) only for random maps. Oddly (and this is undoubtedly a bug), when playing on a Pre-made World the cost of research is the same as was set by the last random map generated. For example, if you start a game on a small random map (paying the low research cost) then quit and start up a game on large pre-made map, you will pay the small-map research cost on the large map. Similarly if you play a large random map game then start a small edited map game, you pay the large-map research cost on the small map game. This feature (or bug) exists in both Civ2.4.2 and the latest version of the MP Gold edition. If the first game you play upon starting Civ2 is on an Pre-made World, then you pay the "medium map" cost regardless of map size.

This explains, I believe, the research cost anomalies that have been reported in comparison games. Since each player's most recent game play history affects the research cost they pay on edited maps, it is quite possible for one player to pay the small map cost while another pays the large map cost. Any comparison of game play on Pre-made maps, even standard ones such as Earth, are therefore suspect.

Also, this can be used as another "tell" for distinguishing random and edited map starts. If the research cost does not match the map size, it is an edited map.

This bug appears to affect, at least partially, MP games on Pre-made worlds as well. Whether this could be exploited to the advantage of one player over another, I have not yet determined. Perhaps someone would be interested in checking that out.

Assuming this bug is caused by a "sticky" mapsize flag, then there could be other mapsize-based game factors besides research cost which are affected. Does anyone know of any?

[Caesar the Great]

We learn something new each day....

Do you think this affect scenarios as well?

[Marquis de Sodaq]

Other general game parameters are also sticky:

Cheat mode: If you use cheat mode in a game, then quit and load another, your second game score will still show a cheat. To disable it, you must quit the software.

City names: The list of city names used is run thru once per software load (for random starts). If you start as the Romans, settle Rome, quit, restart as the Romans, and settle, Veii will be your capitol. The name Rome was "skipped" because the software thinks it was already used.

Random start menu selections: If you change these (male/female, civ, and such), your new selections are retained the next time you start a game, not the defaults.

I think this happens universally because the game menus were designed to retain your latest choices for that entire run of the software. Only a startup will set them to default. Your map size selection from the first game is retained as a variable that the software apparently references for tech cost.

One wonders if also your barbarian level selection would be retained for scoring? We all probably play every game on the same level - If this is sticky, you could start at one difficulty level, quit, then restart at another level, but reap the benefits of the first selection (frequency, more gold for leaders, weaker/stronger barb attack values). I've never tested this, has anybody else?

[samson]

Caesar,

Yes, scenarios are affected by this bug but at creation time, not game time. Since most scenarios use edited maps, the research cost rate of the scenario game will be inherited from the mapsize of the last random map generated before the scenario was created. (Or the default medium map cost, if there was none. ) This may or may not match the mapsize of the scenario itself.

In a quick check of several OCC fortnight games, I found most small map scenarios on edited maps were using medium map research costs. This is to be expected. Since most scenario creators would be unaware of the bug, the default of medium map cost gets assigned regardless of map size. To properly match the research cost and mapsize, the scenario creator would have to generate a random map of the same size used in the scenario just prior to loading the edited map and saving the scenario. Smash's 2 whale/2 gold scenario (a small map) is properly matched.

Saved games using edited maps are similarly affected. OCC23 (the river map) is large while OCC25 and OCC2 are on small maps yet the research cost of all of them is that of a medium map.

Marquis,

The Cheat mode and City name bugs would seem to be the same as the mapsize problem. The other "sticky" features you mention are different. They are preferences which are saved across startups. Further, options such as the Barbarian setting can not be bypassed on a new game start and therefore could not affect play in an unforeseen manner.

What I am wondering is: are there any other game elements (like research cost) which are determined, in whole or part, by mapsize? If so, these would also be affected by this bug.

[Ribannah]

I'm thinking it may have an effect on AI behaviour, like exploration, the number of cities it is planning to build and diplomacy.

[Julius Brenzaida]

It also has an effect on the number of turns on lower levels. For example, if you play a small map at warlord, you will have less turns than if you play a large map.