Starting location pretty much has two possible ways to affect.
1. By limiting what the game considers acceptable when picking locations for the capitals. We have limited options here because the game freezes and fails the map generation if the map does not have enough good spots to fill all requirements at once for every wizard.
In particular, we can have the "max pop" and the "distance from other wizards" appear here and nothing else (the rest are obviously needed restrictions such as "must be on land").
With the newest - not yet released - update, the system changes so that the game drops the food and wizard distance requirements simultaneously if a higher requirement cannot be fulfilled in a certain amount of tries. (30000)
We need to discuss what the lowest acceptable amount for these two (enemy distance and food) is.
We also need to discuss how the correlation between the two should work - for example for each 2 lower food we drop distance by 1, or for each 1 food, or the other way around, we drop 1 food for 2 distance, etc.
And finally we need to discuss what the starting amounts should be, the amount where the game doesn't even try a higher because we don't expect it to happen or we don't care about it. For example we can say "We don't care to try for 18 food, and are fine with at least 12".
Based on a few experiments, a distance of 16+ is unlikely to happen and I usually get maps with the lowest distances being around 12 on normal and larger maps. On Tiny, even a distance of 6 is not unheard of.
Do note that if placing the cities fails for the lowest amount the game will now move lairs (which don't allow cities next to them for AI reasons) and tries again. So even if the map would end up a successful "6 wizard distance", if we set the minimal distance at 10 it might work, but would take longer to generate until lairs end up in a position where it works out. However, as Nodes and Towers are unmovable, it's also possible that repeated attempts fail as well and the game freezes. I don't know what the highest safe "minimal value" for each stat is. I think food should be safe up to ~10, but distance might not be. If we do want high minimal amounts, we'll need to do some extensive testing to see if it works or not. Also note that the minimal amount, maximal amount and the correlation aren't independent, if we decide on any two, the last one will have to be calculated from that.
2. As discussed in the starting mineral thread, it's in theory possible to make the game check tiles and add or remove minerals after the city was already placed. I believe it's also possible to change terrain itself, as at that time the tiles are still "raw" and not connected which means only the 10-15 basic tile types are used, not the 600+ various subtypes. So if we want any sort of restriction on production, minerals, etc, we need that to apply here. However, this a questionable, as it artificially changes the availability of good or bad starts and would take a huge amount of time for marginal gain. Honestly, as good as it sounds in theory, aside from early rush tactics, nothing else cares about the starting city to this level of detail. Sure, the starting city might not have adamantium, but the first hamlet next to it does, if you plan to strike on turn 90 it makes zero difference. Yes, the starting city matters a bit more, but still I rather start on a pop 4 swamp tundra that has pop 25 half a screen away with 3-4 ores than on something perfectly average. I also think it'll be close to impossible to agree on anything here as some people will ask for a maximal others for a minimal quality in the same or different resources, so I am tempted to just go with "I rather not open this can of worms". More importantly I now see the "[Starting Location] - Unable to improve this due to the order of steps in map generation." entry in the current tasks list which doesn't sound promising. I remember we had some sort of a discussion about it this spring which didn't go very well and concluded on "can't be done", but I don't remember the details. It might have been for some different goal but now I think it probably was the same. So I strongly believe we are better off staying away from trying to add this feature, game balance or not.
1. By limiting what the game considers acceptable when picking locations for the capitals. We have limited options here because the game freezes and fails the map generation if the map does not have enough good spots to fill all requirements at once for every wizard.
In particular, we can have the "max pop" and the "distance from other wizards" appear here and nothing else (the rest are obviously needed restrictions such as "must be on land").
With the newest - not yet released - update, the system changes so that the game drops the food and wizard distance requirements simultaneously if a higher requirement cannot be fulfilled in a certain amount of tries. (30000)
We need to discuss what the lowest acceptable amount for these two (enemy distance and food) is.
We also need to discuss how the correlation between the two should work - for example for each 2 lower food we drop distance by 1, or for each 1 food, or the other way around, we drop 1 food for 2 distance, etc.
And finally we need to discuss what the starting amounts should be, the amount where the game doesn't even try a higher because we don't expect it to happen or we don't care about it. For example we can say "We don't care to try for 18 food, and are fine with at least 12".
Based on a few experiments, a distance of 16+ is unlikely to happen and I usually get maps with the lowest distances being around 12 on normal and larger maps. On Tiny, even a distance of 6 is not unheard of.
Do note that if placing the cities fails for the lowest amount the game will now move lairs (which don't allow cities next to them for AI reasons) and tries again. So even if the map would end up a successful "6 wizard distance", if we set the minimal distance at 10 it might work, but would take longer to generate until lairs end up in a position where it works out. However, as Nodes and Towers are unmovable, it's also possible that repeated attempts fail as well and the game freezes. I don't know what the highest safe "minimal value" for each stat is. I think food should be safe up to ~10, but distance might not be. If we do want high minimal amounts, we'll need to do some extensive testing to see if it works or not. Also note that the minimal amount, maximal amount and the correlation aren't independent, if we decide on any two, the last one will have to be calculated from that.
2. As discussed in the starting mineral thread, it's in theory possible to make the game check tiles and add or remove minerals after the city was already placed. I believe it's also possible to change terrain itself, as at that time the tiles are still "raw" and not connected which means only the 10-15 basic tile types are used, not the 600+ various subtypes. So if we want any sort of restriction on production, minerals, etc, we need that to apply here. However, this a questionable, as it artificially changes the availability of good or bad starts and would take a huge amount of time for marginal gain. Honestly, as good as it sounds in theory, aside from early rush tactics, nothing else cares about the starting city to this level of detail. Sure, the starting city might not have adamantium, but the first hamlet next to it does, if you plan to strike on turn 90 it makes zero difference. Yes, the starting city matters a bit more, but still I rather start on a pop 4 swamp tundra that has pop 25 half a screen away with 3-4 ores than on something perfectly average. I also think it'll be close to impossible to agree on anything here as some people will ask for a maximal others for a minimal quality in the same or different resources, so I am tempted to just go with "I rather not open this can of worms". More importantly I now see the "[Starting Location] - Unable to improve this due to the order of steps in map generation." entry in the current tasks list which doesn't sound promising. I remember we had some sort of a discussion about it this spring which didn't go very well and concluded on "can't be done", but I don't remember the details. It might have been for some different goal but now I think it probably was the same. So I strongly believe we are better off staying away from trying to add this feature, game balance or not.
