The ideal length of a game

[balparmak]

Yeah 450 average is too high. I think if you go by anything higher than 425 as the avg. point, you're getting too close to time victory. On the issue of balancing victory conditions, I think they should aim for the same average, otherwise one will be simply more accesible than others. Sure each has different counters and boosts, like the DV is very much affected by whether the game rolls more than one DV civs, but I don't see any way to move forward without a baseline, counters&boosts can come after we get different victory conditions within acceptible timeframes.

 

[ma_kuh]

I made a graphic from the data in the 2.7 AI games thread. Low and high points of the boxes are 1st and 3rd quartiles of the data.
(https://forums.civfanatics.com/thre...update-now-with-summary-list-of-games.679819/)

Median win turns:
Cultural: 380
Diplomatic: 406
Domination: 426
Science: 431
Time: 500

Diplomatic seems like it's in a good baseline spot. Culture wins 20 turns early by general consensus in this thread, and Science is 30 turns late. I wouldn't pretend like the single Domination win is meaningful, but I have a suspicion the Time wins might be somewhat related to Domination victories that never quite converted in the end (since no other alternative win condition kicked them out).

 

[psparky]

While personally I would prefer a higher number, if the average at king is above 400, then players at lower difficulties will often hit the 500 limit before being able to complete the game. This is because in addition to these players playing less efficiently, the AI's offer less help e.g. have no gold for you to acquire, don't give a discount on research because they haven't researched the techs either, etc. This is bad.

To achieve 400 turns without making the later eras fly by, it would be necessary to shorten all the earlier eras, again something that I wouldn't like.

One possibility is to make the game pace depend on difficulty level, so that the average is 400 turns whatever the difficulty level. Or perhaps we could extend the game - it's trivial to add 100 turns with a 2100 end date.

 

[azum4roll]

We don't have to keep time victory at 500 turns by default. The game doesn't break after turn 500 anymore.

 

[KlHannibal2]

I think balancing victories condition based on in-game length (turns) isn't a good approach. It would be better if we balance it based on the amount of work needed (aka real-time game length, not in-game turns).

For example victories that constantly give you more things to do like domination (manage warfare) or science (both research and production of new building/units that come with new techs) should come earlier than victories with less things to do like culture or diplomatic (mostly just passing turns while repeating the same thing/waiting for specific timing like GM bomb or next congress), and the slowest one would be the most boring ones (time).
Reason being the more active victories should give faster result in perfect condition, but since they're usually harder to achieve they will get delayed and end up with similar timing of other easier but slower victories (plus civ aiming for easier but slower victories can spend more time trying to sabotage civ with earlier victory instead of sitting tight passing turn)

Based on score/time victory at turn 500 standard speed, I would say if we can balance culture/diplomatic/science victories around 450/425/400 it would work best. Domination would need the new proposal passed, but aiming for 375 would be reasonable. Reminder this is perfect condition (when you're running away), the actual turns to get the corresponding victories would get delayed and generally would happen around the same time for all victories (450-500).

I don't agree that domination victory should by design come at an earlier turn than peaceful victories. That would make you lose out on late game units and they constitute a much different type of warfare. Also, a competitive domination game is usually a race against an AI's peaceful victory (if your are not killed by a knight rush).

If you can easily win domination before other victory conditions kick in, the domination game will fell less challenging and players might not be interested in finishing it. I find it more interesting if you still have to slow down the leading AI by world congress or war with other AIs before you can take it out yourself in order to win a domination game.

 

[nekokon]

You're missing my point. The idea of making domination victory come earlier than peaceful victory when you're already running away (very important part) is because in practice the closer you're to a domination victory the harder it becomes, and you will be slowed down to the point of only able to achieve domination victory around the same time with other peaceful victories.

The order of victory condition is based on perfect condition where you're already running away and nothing can stop you, and if at that moment you decide to pick a victory to go for you would get the order above (domination comes earliest). In a real game you won't be in that perfect condition (else you're playing at lower difficulty than you should). This order is made based on how much harder it becomes the closer you get to a certain victory condition.

 

[azum4roll]

The ideal length is 450-500 turns for me. I think it's necessary if you want to have a chance to make use of each unit at their prime.