What's missing to get realistic timelines

alexman · June 5, 2002, 18:20

Firaxis has said that they test and balance the game at Regent level. They have done an excellent job in that, IMHO.

At levels other than Regent, however, the play is not as balanced. Spaceship launces by 1200 A.D. at Deity anyone? This tech acceleration at higher levels spoils the Civ3 feel. You hardly have time to build a unit and it goes obsolete.

The reason for the timeline being so out of whack is, of course, the fact that the AI gets bonuses at higher levels.

There is an extremely simple way for Firaxis to fix to this problem: Just multiply the cost of each technology by the ratio of the AI Cost Factor at Regent to the Cost Factor at the current level.
This should be done internally, without the need to add yet another parameter in the editor.

Example:
At Deity, the AI Cost Factor is 6.
At Regent, the AI Cost Factor is 10.
When playing at Deity, all techs should cost 1.66 (10/6) times more.

This assumes that the turns played up to a given point in the tech tree are directly proportional to the tech cost. Is this true in practice? Probably not exacly, because of more aggressive AI trading at higher levels. It is also not accurate for lower levels, especially Chieftain, because in those cases the human is usually the tech leader, so the rate goes faster than what the AI would do alone without the human player. But in any case, it's certainly better than what we have now!

Lets assume a Regent-level game that enters the modern age at 1950 A.D. That's 440 turns worth of research from all civs. For a Deity-level game, the AIs would get the equivalent of 440 turns of research in only 264 turns: year 1320! Sounds familiar? Using the same assumption, you would enter the modern age at 1774 on Emperor, and at 1862 on Monarch. Is this about right for you? If it is, then the proposed fix would work perfectly.

Spiffor · June 5, 2002, 20:55

Very good ideas. Plus, the weaker players (chieftains and warlords) currently suffer from a too slow tech rate. That would help them, by giving them a true boost in an otherwise regular tech race.

Sava · June 6, 2002, 07:25

The problem is with the turn based system in general. EU2 has a better flow of time. Maybe if Civ had a turn based system like EU2's where you issue orders and then the units move in the turn.

DeepO · June 6, 2002, 08:17

I also posted this idea in some earlier thread, so of course I'm in favor of it

I don't know whether your formula will work in every case, I think it is a bit too simple. But in general, you're abslutely right.

Other factors you might consider: the more AIs, the faster research is going. Taking the amount of AIs into account (e.g. the square root of the # of AIs or something, to avoid extreme costs), and possibly the mapsize or better yet the ideal number of cities (for corruption) on that map size should better balance it on higher levels.

Doing this, you would ensure that a good, non-corrupt AI city always gives about the same portion of research, e.g. 1/50 of total needed beakers for a tech, in each era. Plus, trading between AIs would be accounted for.

One downside to all of this is that on higher levels, on bigger maps, and with many enemies, it becomes harder for the human to research something on his own. Gaining a considerable lead is already hard, it would become nearly impossible. It would call for even harder early warlike expansion, as it won't be possible to keep up when your empire size is mediocre. Now, people can still win a space race when they're small...

DeepO

alexman · June 6, 2002, 09:49

DeepO, I agree that this won't fix faster research with an increasing number of AIs. But I personally can live with that. It's actually realistic.

As for map size, it's already taken into account. There is a "Tech Rate" setting in the editor (one per map size), which effectively multiplies all tech costs. Firaxis has done a good job in balancing tech rate for each map size at Regent, I think. It's balancing the difficulty levels where they apparently didn't put any effort. The proposed fix is very easy to implement. It could easily be part of the next patch, and doesn't need extensive playtesting. Simple is better, sometimes!

Jawa Jocky · June 6, 2002, 10:51

small maps or more ocean coverage on normal maps will give you a more accurate time line. It also speeds up the game

Jawa Jocky · June 6, 2002, 10:52

small maps and increased ocean coverage on normal maps give an accurate time line. It also plays quickier

how did I post this twice

Adagio · June 6, 2002, 11:28

The idea is great

If it works I don't know, but something needs to be changed, and this might be the way to do it...

DeepO · June 6, 2002, 11:55

Quote:

Originally posted by alexman
DeepO, I agree that this won't fix faster research with an increasing number of AIs. But I personally can live with that. It's actually realistic.

As for map size, it's already taken into account. There is a "Tech Rate" setting in the editor (one per map size), which effectively multiplies all tech costs. Firaxis has done a good job in balancing tech rate for each map size at Regent, I think. It's balancing the difficulty levels where they apparently didn't put any effort. The proposed fix is very easy to implement. It could easily be part of the next patch, and doesn't need extensive playtesting. Simple is better, sometimes!

Aha, I wondered if there was something to balance mapsize. In that case I totally agree, they should add a factor to make research harder on higher difficulty levels. I don't know if it is needed to make it so that lower levels are speeded up, but this could easily be capped.

Faster research with more AIs: I know, it is a characteristic of the game. And as such it can be realistic. But still it feels not quite all right, maybe a little factor would help, which does not negate the whole effect... this is, just like the 'bigger cost if higher level' something that can be very easy implemented in a patch, and as such worth looking into.

DeepO

aebrahim · June 6, 2002, 18:59

Quote:

Originally posted by alexman
There is an extremely simple way for Firaxis to fix to this problem: Just multiply the cost of each technology by the ratio of the AI Cost Factor at Regent to the Cost Factor at the current level.
This should be done internally, without the need to add yet another parameter in the editor.

Wow, what a wonderful idea. Hope they implement this.

Mannamagnus · June 7, 2002, 04:38

My timeline is a chronological sequence the dates are just figures not related to the real world

Panag · June 7, 2002, 08:15

hi ,

there should be the ability to add more turns in the editor ,....

more years , meaning , in stead of five years , turns like one year , the rate and the amount of time for research should also be looked at again , ....

have a nice day

alexman · July 19, 2002, 18:10

Soren just said (in the chat) that this change made it in the 1.29f patch.

Thank you Firaxis for listening!