Interesting Game Data

[Simple Machine]

Hello! I tried something interesting in an immortal game: recording turn-by-turn game data, because what else am I to do quarantined at home .

Game Settings:
- Immortal Difficulty
- Pangaea
- Standard World size
- Normal Speed
- Six AIs
- Temperate climate
- Medium sea level
- No huts, no events

I'm attaching the 4000 BC worldbuilder save and the raw data itself. Each turn, I recorded the following:
1) Turn #
2) Beakers per Turn (BPT) @ 100% Slider
3) Gold per Turn (GPT) @ 100% Slider
4) BPT @ 0% Slider
5) GPT @ 0% Slider
6) Score
7) Population
8) City Count

Despite the manual entry, it actually wasn't that bad! Only took about 15-20 seconds extra per turn. From that I could use formulas to calculate several other interesting fields:

9) Breakeven Slider (exact) -- can be calculated from (3) and (5)
10) Breakeven Beakers (exact) -- can be calculated from (2), (3), (4) and (5)
11) Five-turn moving average of Breakeven Beakers (10)
12) Five-turn moving average of Breakeven Slider (9)
13) Five-turn moving average of score (6)
14) Turn-over-turn growth rate of (11)
15) Five-turn moving average of Population (7)
16) Turn-over-turn growth rate of (15)
17) Turn-over-turn growth rate of (13)

Here is what the start looked like:

Spoiler Interesting takeaways from the data :

So I went for a Hwacha rush against Egypt, vassalized him and later vassalized Ethiopia too. I haven't actually finished the game yet, only going up to turn 225 or so. I'll highlight a few of the points that I think are interesting:

Breakeven beakers per turn (exact value):

Breakeven slider (exact value):

For both charts, the blue line is the actual turn-by-turn value, and the red line is a 5-turn moving average. Black lines indicate when I founded cities, bold Yellow/Green lines indicate when I captured cities from Egypt/Ethiopia, and the purple lines indicate golden ages.

Early game was perhaps not optimal, but I developed quickly and founded cities at each of the solid black lines in the chart above: T41, then T53, T68, T74 and T83 for a total of six cities. It took 77 turns to first hit 30 BPT at breakeven, but then 50 BPT came at turn 89, and 75 BPT at turn 98. The noticeable jump upwards at T103 came from completing Currency, which in one turn increased BPT by about 40%! The high water mark before invading Egypt was a lofty 112.8 BPT (at breakeven slider of 75%) on turn 106.

As I was whipping up an army and spending money moving troops into enemy territory, tech rate declined, but not as rapidly as you might expect -- thanks to being financial and having three happy resources. The low water mark was a dismal 14.1 BPT at breakeven 10.2% slider on turn 130, roughly an 85% beaker decline in just 24 turns. Economic recovery was extremely rapid though, and the capturing of a holy Shrine helped a lot. It took just 9 turns (!) to get BPT back above pre-war levels, and the pacification of the Shrine city (and the +29 gold from it) by itself increased breakeven BPT by 76% in one turn!

The economy really took off with the Taj Mahal, which combined with a great spy gave a 16-turn golden age. This enabled me to build Seowons and Oxford which fully tripled the beaker rate -- not a bad result from just one great wonder and one national wonder. Even after the golden age expired, BPT continued a steep upward trajectory, which I guess was due to me building lots of markets, grocers and observatories in my high-commerce cities.

Anyway, TLDR version:
- Currency is a very significant boost to early game tech rate
- Powerful economic growth happens by reducing your breakeven slider rate while having a "bigger pie" to work with: i.e. 75% of 30 BPT is less than 50% of 100 BPT, which itself is less than 40% of 200 BPT.
- Financial + 3 happy resources = Easy economic recovery for early game war
- At any point, your economy seems to be either "growing" or "at capacity". E.g. BPT was essentially stagnant for about 30 turns right before the GA kicked in, while during and after the GA, the economy grew rapidly. When it hits the new capacity, it will probably be a good time to DOW on somebody and take more land.

If there is any interest I'll probably be doing some more stuff like this, so will be happy to share more data.

 

[dankok8]

I just realized I never replied to this thread. I'd like to see data from more games.

In fact if all games are played on similar settings (i.e. Normal Speed, Pangaea, Moderate Climate, Medium Sea Level) etc. it would be nice to collect benchmarks that we can compare by turn 50, turn 100, turn 150. It will give newer players a sense of where they should be. The turn by turn data is invaluable of course but too time-consuming to record in every game.

 

[FlyinJohnnyL]

I'd love it if we could ban year identifiers altogether here, but I realize that's a personal preference. I have a much better idea of what point of the game someone is at by turn number rather than year. If you're at 375 AD, I don't have a good feel for when that is. But if you say it's turn 130, I'll have a much better idea.

 

[soundjata]

Thanks Simple Machine for the good read

Spoiler somewhat related :

As you speak of covid crisis, I can't help but put your analysis in perspective with... How our lords and rulers also operate on the assumption that growth = win and that going down = good because down means up...
Play with the commerce slider and things will be fine!

Anyways... No citizen = No fun.

 

[Simple Machine]

Thanks Simple Machine for the good read

Spoiler somewhat related :

As you speak of covid crisis, I can't help but put your analysis in perspective with... How our lords and rulers also operate on the assumption that growth = win and that going down = good because down means up...
Play with the commerce slider and things will be fine!

Anyways... No citizen = No fun.

Lol, with my background in economics it is always nice to see that good old-fashioned V-shaped recovery after invading Egypt