v148.6

[Txurce]

I just played an extremely encouraging game. The AI was competitive scientifically, competed appropriately for CS, and built an excellent mix of units, making warfare much tougher.

My only quibbles is that the years move too fast now. I'd suggest slowing down the progress sooner.

A comment on the next version: I don't think it's a good idea to remove trading for DoWs. The AI does not always accept any deal. In fact, they decline more often than they accept. What does happen is that militaristic AI will accept any deal against someone they'd be happy to DoW in the first place... but the latter is definitely not always the case. Removing the option outright takes away too much flexibility, in order to avoid a relatively infrequent exploit.

There may be a courthouse bug: I built one in a conquered city, and still had the "occupied" symbol.

 

[Thalassicus]

I'd seen posts that it's possible to get huge sums of gold from AIs if a DoW agreement is added to the deal... is this true? I honestly haven't ever used the DoW part of deals.

 

[Txurce]

I'd seen posts that it's possible to get huge sums of gold from AIs if a DoW agreement is added to the deal... is this true? I honestly haven't ever used the DoW part of deals.

In my experience - very extensive in this area! - I've never acquired huge amounts of gold, probably because the AI that used to rack up large gold reserves weren't the ones who go for this sort of deal. In the worst cases it's usually a combination of luxuries and gold. (Also keep in mind that the AI no longer has that much gold available, anyway.) To recap, you can only exploit militaristic civs who would otherwise be willing to DoW the target anyway. (I just tested this by asking Japan to DoW Siam, and they refused no matter what I offered.) My sense is that the exploit is narrow enough and contained enough that it's not worth erasing the entire mechanism.

This issue aside, try a game if you haven't played a full one in a while, Thal. It is really humming now.

 

[Thalassicus]

What numbers would you recommend for the year rate? This is the current setup:

Turn|Date|Years per Turn
0|-4000|50
80|0|25
120 | 1000 |20
140|1400|10
170|1700|5
190|1800|2
240 | 1900 |1
340|2000|0.5
500|2080|End

The benchmarks are the Medieval era (should be turn 120 at 1000 CE) and Modern era (turn 240 at 1900 CE). If we can get those two right, the rest should fall into place.

 

[Txurce]

What numbers would you recommend for the year rate? This is the current setup:



The benchmarks are the Medieval era (should be around 500-1500 ce) and Modern era (1950-2000). If we can get those two right, the rest should fall into place.

I was thinking the same way. Your present schedule looks right - it just didn't work out that way for me. I tend to hit Steam Power about when I should (or a little early) with a good science civ, but not otherwise.

Given that the dates can't move much slower after a certain point, why not try 40/20/10 for the first three stages (as opposed to 50/25/20)?

If you don't mind diddling with this, we can keep fine-tuning it. Seek, you're playing right now. What do you think?

 

[Thalassicus]

That would delay the mid-medieval to turn 150 (was 120), and the modern era to turn 290 (was 240). Would these be more accurate?

Turn|Date|Years per Turn
0|-4000|40
100|0|20
150 | 1000 |10
190|1400|10
220|1700|5
240|1800|2
290 | 1900 |1
390|2000|0.5
550|2080|End

Edit: I just noticed I forgot to reduce later turns to compensate for more early turns, resulting in an overflow of 50 turns. It's tricky to adjust this while keeping three restrictions:

• Human-friendly year increments which decrease.
• Same max turns.
• Same max date.

Edit 2: I also realized those numbers would shift steam power from 1800 to 1400, so that wouldn't work... hmm...

 

[Txurce]

Why don't we give this a shot? I think it's that middle window that we have to watch. I'll try to remember to take notes in my next game, playing with a mid-level science civ like the Aztecs.

 

[Thalassicus]

Note the edits - I realized those numbers won't work. We need to know:

• Average turn # of the point where 50% of Medieval techs are researched.
• Same thing for Industrial era.

If we know the turns, and know the real-world year they should represent, it's easy to line them up.

 

[Txurce]

Playing with Korea I now reach Steam Power between t190-210. I think it would be slower with a less-focused civ.

Trying to think back, and just guessing, I would say mid-medieval would be close to t100.

Mid-modern is harder to say accurately, because I usually go Scientific Theory/Replacement Parts/Plastics.

 

[Thalassicus]

Right; I'm thinking Medieval & Industrial would make better benchmarks than Medieval & Modern. In the current version of the mod, 50% medieval is turn 120, and 50% industrial is turn 240. This sounds accurate based on your estimates, but like you said, we need data from more players.

 

[Txurce]

Yeah, I think so. And the Aztecs will be a good test, while giving me the chance to bully the Bronze Age folks.

 

[lordrune]

My current game - in version 149.0 - is on Prince, my usual difficulty, playing England, continents map, standard size, standard speed. The start was pretty good, a peninsula start, the game is more successful than my average. My policies went as follows - tradition, liberty, commerce, freedom, and i've just started rationalism.

I got 5 out of 11 medieval techs at turn 135 (1300), and 6 out of 12 industrial techs at turn 262 (1922). One of the AIs was keeping pace well (they beat me to the Education wonders) but I've just now pulled ahead and am bombing their cities, cannons, and musketeers with Ironclads (which are useful for the first time ever in a Civ game ).

 

[Txurce]

I took my standard science route with the Aztecs, who are easy to dominate with. I have the most beakers by far with double the pop, the GL and PT, one or two RA's, and the Patronage science boost with a lot of CS alliances. With that:

Civil Service (mid Medieval): t94.
Scientific Theory (end Renaissance): t197.
Replacement Parts (mid Industrial): t213.

That all seems about right, don't you think?

 

[wobuffet]

That all seems about right, don't you think?

For a Science game, yeah I think so.

 

[Seek]

I got around the same numbers as Txurce in my cultural Gandhi game.

When I won at t228, AI capitals were between size 7 and 15 (! - but that might have been partially because I was trying the Emigration mod) while I had three cities sized 26, 32 and 33. I dominated in a way I've never seen in any civ game.

Denmark pulled through to take all of siam's cities but the cap towards the end, but it was too little too late. It felt like I was playing on Settler.:/

 

[Zaldron]

What numbers would you recommend for the year rate? This is the current setup:

The following represents my opinion only and I have no concrete proof of this. Take it with a grain of salt.

I don't think this is a solvable problem worth spending time on. No matter how much we tune, we'll only be close in maybe 20% of games. The rest of the time we'll be faster or slower and the years STILL won't look right no matter how much we tweak. I really only look at the current turn to judge my progress anyway. I suspect there are plenty of actual functional issues that we be higher priority than a display-only number that will never be right most of the time.

 

[Txurce]

Zaldron, I think you're right in terms of priorities. Fortunately this didn't take long, and it's much more in synch now than it used to be.

 

[Jaybe]

Zaldron, I think you're right in terms of priorities. Fortunately this didn't take long, and it's much more in synch now than it used to be.

In my 147.2 game I just finished it was much more OUT of synch than usual; getting to infantry in the 1980's, ....

Of course, a great work-around would be to have the dates NOT displayed, only the turn number. To reference dates at potentially different game speeds, specify 'turn number/game length' (e.g., 439/750).

 

[Txurce]

In my 147.2 game I just finished it was much more OUT of synch than usual; getting to infantry in the 1980's, ....

I know what you mean, but... infantry in the 1980s is much closer to dead-on than infantry a few centuries earlier.

 

[lordrune]

Zaldron, I think you're right in terms of priorities. Fortunately this didn't take long, and it's much more in synch now than it used to be.

I have to agree with this.

Most Civ mods (whether 4 or 5) get this balance entirely wrong for the following reason: they buff up yields and modifiers to the point where larger city sizes, combined with the increased yields, makes science and production so large that you start flying through techs, and also building units and buildings instantly - in just about every city. And rush building can become mandatory because your science is going up so fast.

Counteracting this effect fully is really tough but a little bit of effort can greatly improve the experience for people who actually do want to spend a little bit of time in the later eras before the game is won.