4.5 AI-Only Game Statistics (we love war)

L. Vern · Feb 24, 2024

Going to be a pretty quick one today but should be a cool overview nontheless, wanted to put the numbers out there as I haven't done one of these in a while. I'm doing some more work on this to turn it into a grad course project so here's a quick look into how good the ai does conquering things. More ideas for stats you'd like to see would be appreciated.

These games were run on a version that's more or less 4.5, and the settings used now were default. Notably, that means ancient ruins are now enabled so the game speeds up a little bit compared to previous tests. Map was Communitu_79a v3.1.0 . Map size - standard (8 players).

Spoiler Victory Type / Civ Performance History :

Civ	4.5 Rank	4.5 Winrate	3.10 Rank	3.10 Winrate	3.6 Rank	3.6 Winrate	3.2 Rank	3.2 Winrate	2.7 Rank	2.7 Winrate
Austria	1	0.33	2	0.26	1	0.27	1	0.35	4	0.31
The Netherlands	2	0.31	13	0.15	32	0.08	21	0.11	12	0.19
China	3	0.28	19	0.14	8	0.18	2	0.27	2	0.4
The Inca	4	0.28	24	0.11	14	0.15	30	0.07	30	0
The Maya	5	0.25	1	0.27	34	0.07	28	0.08	11	0.19
Germany	6	0.24	6	0.2	5	0.25	4	0.25	14	0.17
Korea	7	0.23	3	0.24	11	0.17	22	0.1	8	0.22
Morocco	8	0.22	26	0.11	15	0.14	16	0.15	9	0.21
The Ottomans	9	0.22	31	0.07	4	0.26	14	0.16	6	0.26
Songhai	10	0.21	20	0.13	26	0.09	12	0.17	26	0.07
Greece	11	0.21	10	0.16	6	0.24	13	0.17	7	0.22
Siam	12	0.2	14	0.15	21	0.11	17	0.14	3	0.35
Venice	13	0.19	21	0.13	41	0.03	19	0.12	22	0.12
Portugal	14	0.18	12	0.15	28	0.09	25	0.09	27	0.07
Poland	15	0.17	28	0.09	7	0.19	26	0.08	5	0.29
Sweden	16	0.16	32	0.07	24	0.1	31	0.07	41	0
Russia	17	0.15	9	0.18	30	0.08	37	0.04	37	0
The Celts	18	0.15	15	0.15	31	0.08	32	0.07	25	0.09
Byzantium	19	0.13	11	0.16	22	0.11	24	0.1	38	0
Babylon	20	0.13	8	0.19	13	0.15	34	0.07	15	0.16
Brazil	21	0.12	5	0.2	12	0.17	3	0.27	10	0.19
The Iroquois	23	0.11	22	0.13	23	0.11	11	0.17	18	0.13
Assyria	22	0.11	18	0.14	17	0.14	33	0.07	17	0.14
Arabia	24	0.1	4	0.23	2	0.26	7	0.23	1	0.44
India	25	0.1	17	0.15	10	0.17	20	0.12	19	0.13
Egypt	26	0.08	23	0.12	18	0.13	10	0.18	13	0.19
Ethiopia	27	0.07	37	0.05	19	0.13	15	0.16	16	0.14
America	28	0.06	42	0	27	0.09	8	0.21	39	0
Carthage	29	0.05	16	0.15	38	0.05	35	0.06	23	0.1
Denmark	30	0.05	36	0.05	36	0.06	42	0	35	0
Persia	31	0.05	25	0.11	33	0.07	23	0.1	34	0
Indonesia	32	0.05	39	0.03	16	0.14	5	0.23	36	0
Rome	33	0.05	29	0.09	9	0.17	38	0.03	20	0.12
The Zulus	34	0.04	35	0.05	20	0.12	18	0.14	40	0
England	35	0.04	30	0.07	29	0.09	29	0.07	28	0.06
France	36	0.03	27	0.09	43	0.01	36	0.04	43	0
The Aztecs	37	0.02	33	0.07	25	0.09	9	0.21	33	0
Japan	38	0.02	7	0.2	3	0.26	6	0.23	24	0.09
The Shoshone	39	0.02	43	0	35	0.06	40	0	31	0
Polynesia	40	0.02	34	0.06	39	0.05	39	0	21	0.12
Mongolia	41	0.02	41	0.03	37	0.05	43	0	32	0
The Huns	42	0	40	0.03	42	0.03	41	0	42	0
Spain	43	0	38	0.05	40	0.05	27	0.08	29	0.06

Victory share by version:

Victory Type	4.5	3.10	3.6	3.2	2.7
Cultural	28.8	26.2	36.3	50.7	58.7
Diplomatic	30.6	25.2	24	22	29.3
Science	31.3	37.9	19.7	14	9.8
Domination	8.3	5.8	13.7	6.7	0
Time	1.1	4.9	6.3	6.7	2.2

Spoiler Victory Types :

Spoiler Civilization Performance :

Rank	Civ	Total Games	Total Victories	Winrate	Culture	Diplomatic	Domination	Science	Time	Average Victory Points
1	Austria	52	17	0.327	8	7	0	2	0	1967.85
2	The Netherlands	51	16	0.314	2	14	0	0	0	2075.76
3	China	50	14	0.28	10	0	0	4	0	2401.02
4	The Inca	47	13	0.277	2	1	0	10	0	3074.4
5	The Maya	51	13	0.255	4	5	0	4	0	2064.96
6	Germany	42	10	0.238	3	4	0	3	0	1592.12
7	Korea	48	11	0.229	5	1	0	5	0	2097.75
8	Morocco	63	14	0.222	4	8	0	2	0	1727.03
9	The Ottomans	45	10	0.222	2	3	0	5	0	2225.91
10	Songhai	61	13	0.213	0	3	7	3	0	3594.43
11	Greece	39	8	0.205	2	6	0	0	0	1373.92
12	Siam	46	9	0.196	0	8	0	1	0	1662.35
13	Venice	53	10	0.189	4	5	0	1	0	2068.11
14	Portugal	44	8	0.182	3	1	0	4	0	1586.14
15	Poland	53	9	0.17	8	0	0	1	0	1884.24
16	Sweden	63	10	0.159	1	2	4	3	0	3168.46
17	Russia	74	11	0.149	1	5	0	5	0	2468.39
18	The Celts	48	7	0.146	4	2	0	1	0	1678.06
19	Byzantium	53	7	0.132	2	2	0	3	0	1713.23
20	Babylon	47	6	0.128	3	0	0	3	0	1761.49
21	Brazil	50	6	0.12	4	0	0	2	0	1592.22
22	The Iroquois	53	6	0.113	0	1	4	1	0	2952.17
23	Assyria	53	6	0.113	0	1	0	5	0	1990
24	Arabia	50	5	0.1	4	0	0	1	0	1291.7
25	India	51	5	0.098	2	0	0	3	0	1512.29
26	Egypt	53	4	0.075	4	0	0	0	0	1126.66
27	Ethiopia	42	3	0.071	0	0	0	3	0	1532.14
28	America	50	3	0.06	1	0	2	0	0	2151.86
29	Carthage	55	3	0.055	0	2	0	1	0	1805.16
30	Denmark	61	3	0.049	0	0	2	0	1	2306.75
31	Persia	63	3	0.048	0	0	2	1	0	1872.6
32	Indonesia	44	2	0.045	0	1	0	1	0	1745.09
33	Rome	44	2	0.045	0	1	1	0	0	2094.45
34	The Zulus	50	2	0.04	0	0	1	1	0	2544.4
35	England	53	2	0.038	0	2	0	0	0	1613.83
36	France	61	2	0.033	0	0	0	1	1	1973.59
37	The Aztecs	48	1	0.021	0	1	0	0	0	2133.52
38	Japan	50	1	0.02	1	0	0	0	0	1737.92
39	The Shoshone	52	1	0.019	1	0	0	0	0	1567.58
40	Polynesia	54	1	0.019	0	1	0	0	0	1622.13
41	Mongolia	59	1	0.017	0	0	0	0	1	1786.39
42	The Huns	57	0	0	0	0	0	0	0	1654.89
43	Spain	41	0	0	0	0	0	0	0	1499.68

For the next sections I wanted to explore how often the AI actually does really combat-wise but chooses an alternate path to victory, and what paths it chooses, so made up the concept of a "Pseudo-Domination victory", which is when a civ takes over most of the world but wins via a different victory type. I have arbitrarily defined it as when they win a non-domination victory while owning more than 2/3 of tiles and more than 2/3 of cities.

Spoiler Victory Types With Pseudo-Domination :

Spoiler Civilization Performance With Pseudo-Domination :

Spoiler Pseudo-Domination Wins Breakdown :

Spoiler List of Pseudo-Domination Games :

Spoiler Per-Civ Combat Related Stats :

Rank	Civ	Total Games	Total Victories	Winrate	Cultural	Diplomatic	Domination	Science	Time	Pseudo-Domination	Average Win Owned Territory Percent	Average Win Owned Cities Percent	Average Victory Points
1	Austria	52	17	0.327	8	7	0	2	0	0	0.193	0.172	1967.85
2	The Netherlands	51	16	0.314	2	13	0	0	0	1	0.268	0.256	2075.76
3	China	50	14	0.28	10	0	0	3	0	1	0.367	0.351	2401.02
4	The Inca	47	13	0.277	2	1	0	10	0	0	0.307	0.313	3074.4
5	The Maya	51	13	0.255	4	5	0	4	0	0	0.274	0.27	2064.96
6	Germany	42	10	0.238	3	4	0	3	0	0	0.159	0.146	1592.12
7	Korea	48	11	0.229	5	1	0	5	0	0	0.235	0.238	2097.75
8	Morocco	63	14	0.222	4	8	0	2	0	0	0.174	0.162	1727.03
9	The Ottomans	45	10	0.222	2	3	0	5	0	0	0.198	0.197	2225.91
10	Songhai	61	13	0.213	0	1	7	2	0	3	0.789	0.791	3594.43
11	Greece	39	8	0.205	2	6	0	0	0	0	0.214	0.2	1373.92
12	Siam	46	9	0.196	0	7	0	1	0	1	0.216	0.198	1662.35
13	Venice	53	10	0.189	4	5	0	1	0	0	0.216	0.204	2068.11
14	Portugal	44	8	0.182	3	1	0	4	0	0	0.248	0.242	1586.14
15	Poland	53	9	0.17	8	0	0	1	0	0	0.251	0.235	1884.24
16	Sweden	63	10	0.159	0	2	4	2	0	2	0.701	0.711	3168.46
17	Russia	74	11	0.149	1	5	0	5	0	0	0.381	0.37	2468.39
18	The Celts	48	7	0.146	4	2	0	1	0	0	0.238	0.237	1678.06
19	Byzantium	53	7	0.132	2	2	0	3	0	0	0.18	0.166	1713.23
20	Babylon	47	6	0.128	3	0	0	3	0	0	0.237	0.24	1761.49
21	Brazil	50	6	0.12	4	0	0	2	0	0	0.258	0.252	1592.22
23	The Iroquois	53	6	0.113	0	1	4	0	0	1	0.751	0.743	2952.17
22	Assyria	53	6	0.113	0	0	0	5	0	1	0.434	0.432	1990
24	Arabia	50	5	0.1	4	0	0	1	0	0	0.149	0.145	1291.7
25	India	51	5	0.098	2	0	0	3	0	0	0.24	0.227	1512.29
26	Egypt	53	4	0.075	4	0	0	0	0	0	0.143	0.138	1126.66
27	Ethiopia	42	3	0.071	0	0	0	3	0	0	0.158	0.167	1532.14
28	America	50	3	0.06	1	0	2	0	0	0	0.667	0.664	2151.86
29	Carthage	55	3	0.055	0	1	0	1	0	1	0.511	0.512	1805.16
30	Denmark	61	3	0.049	0	0	2	0	1	0	0.709	0.72	2306.75
31	Persia	63	3	0.048	0	0	2	1	0	0	0.711	0.715	1872.6
32	Indonesia	44	2	0.045	0	1	0	1	0	0	0.352	0.343	1745.09
33	Rome	44	2	0.045	0	0	1	0	0	1	0.86	0.868	2094.45
34	The Zulus	50	2	0.04	0	0	1	0	0	1	0.861	0.849	2544.4
35	England	53	2	0.038	0	2	0	0	0	0	0.303	0.303	1613.83
36	France	61	2	0.033	0	0	0	0	1	1	0.633	0.615	1973.59
37	The Aztecs	48	1	0.021	0	0	0	0	0	1	0.835	0.836	2133.52
38	Japan	50	1	0.02	1	0	0	0	0	0	0.327	0.295	1737.92
39	The Shoshone	52	1	0.019	1	0	0	0	0	0	0.233	0.246	1567.58
40	Polynesia	54	1	0.019	0	1	0	0	0	0	0.209	0.182	1622.13
41	Mongolia	59	1	0.017	0	0	0	0	1	0	0.323	0.354	1786.39
42	The Huns	57	0	0	0	0	0	0	0	0	0	0	1654.89
43	Spain	41	0	0	0	0	0	0	0	0	0	0	1499.68

Also, here is a video of the first 225 games so you can see a bit more qualitatively how AI games go on the territory front. If you love watching colors change on a map you are going to love this video lol

Spoiler replay compilation :

Based on these numbers and my experience with an emperor level game going on right now I would say the AI is pretty strong on a combat front, they are quite aggressive and take cities really well now I think. I've already heard rumblings of a Songhai nerf coming...

ryanmusante · Feb 24, 2024

Can I see statistics on the variations between versions?

Stalker0 · Feb 24, 2024

The pseudo don stat is interesting but still seems a minor portion of the winning civ strats.

I think the more interesting (and likely harder to get) stat is, if you assume your pseudo don as a win condition…how often would a civ actually have “won” over the civ that actually won the game?

I’m other words you are saying “this civ won, but they kind of won by a different method” but I want to know “this civ won, but if pseudo dom was a thing this other civ would have won”

L. Vern · Feb 24, 2024

ryanmusante said:
Can I see statistics on the variations between versions?

It's not really practical to do so for most of the metrics, but here's a history of the big overview metrics, victory share and civ rank/winrate

Spoiler Victory Type / Civ Performance History :

Civ	4.5 Rank	4.5 Winrate	3.10 Rank	3.10 Winrate	3.6 Rank	3.6 Winrate	3.2 Rank	3.2 Winrate	2.7 Rank	2.7 Winrate
Austria	1	0.33	2	0.26	1	0.27	1	0.35	4	0.31
The Netherlands	2	0.31	13	0.15	32	0.08	21	0.11	12	0.19
China	3	0.28	19	0.14	8	0.18	2	0.27	2	0.4
The Inca	4	0.28	24	0.11	14	0.15	30	0.07	30	0
The Maya	5	0.25	1	0.27	34	0.07	28	0.08	11	0.19
Germany	6	0.24	6	0.2	5	0.25	4	0.25	14	0.17
Korea	7	0.23	3	0.24	11	0.17	22	0.1	8	0.22
Morocco	8	0.22	26	0.11	15	0.14	16	0.15	9	0.21
The Ottomans	9	0.22	31	0.07	4	0.26	14	0.16	6	0.26
Songhai	10	0.21	20	0.13	26	0.09	12	0.17	26	0.07
Greece	11	0.21	10	0.16	6	0.24	13	0.17	7	0.22
Siam	12	0.2	14	0.15	21	0.11	17	0.14	3	0.35
Venice	13	0.19	21	0.13	41	0.03	19	0.12	22	0.12
Portugal	14	0.18	12	0.15	28	0.09	25	0.09	27	0.07
Poland	15	0.17	28	0.09	7	0.19	26	0.08	5	0.29
Sweden	16	0.16	32	0.07	24	0.1	31	0.07	41	0
Russia	17	0.15	9	0.18	30	0.08	37	0.04	37	0
The Celts	18	0.15	15	0.15	31	0.08	32	0.07	25	0.09
Byzantium	19	0.13	11	0.16	22	0.11	24	0.1	38	0
Babylon	20	0.13	8	0.19	13	0.15	34	0.07	15	0.16
Brazil	21	0.12	5	0.2	12	0.17	3	0.27	10	0.19
The Iroquois	23	0.11	22	0.13	23	0.11	11	0.17	18	0.13
Assyria	22	0.11	18	0.14	17	0.14	33	0.07	17	0.14
Arabia	24	0.1	4	0.23	2	0.26	7	0.23	1	0.44
India	25	0.1	17	0.15	10	0.17	20	0.12	19	0.13
Egypt	26	0.08	23	0.12	18	0.13	10	0.18	13	0.19
Ethiopia	27	0.07	37	0.05	19	0.13	15	0.16	16	0.14
America	28	0.06	42	0	27	0.09	8	0.21	39	0
Carthage	29	0.05	16	0.15	38	0.05	35	0.06	23	0.1
Denmark	30	0.05	36	0.05	36	0.06	42	0	35	0
Persia	31	0.05	25	0.11	33	0.07	23	0.1	34	0
Indonesia	32	0.05	39	0.03	16	0.14	5	0.23	36	0
Rome	33	0.05	29	0.09	9	0.17	38	0.03	20	0.12
The Zulus	34	0.04	35	0.05	20	0.12	18	0.14	40	0
England	35	0.04	30	0.07	29	0.09	29	0.07	28	0.06
France	36	0.03	27	0.09	43	0.01	36	0.04	43	0
The Aztecs	37	0.02	33	0.07	25	0.09	9	0.21	33	0
Japan	38	0.02	7	0.2	3	0.26	6	0.23	24	0.09
The Shoshone	39	0.02	43	0	35	0.06	40	0	31	0
Polynesia	40	0.02	34	0.06	39	0.05	39	0	21	0.12
Mongolia	41	0.02	41	0.03	37	0.05	43	0	32	0
The Huns	42	0	40	0.03	42	0.03	41	0	42	0
Spain	43	0	38	0.05	40	0.05	27	0.08	29	0.06

Victory share by version:

Victory Type	4.5	3.10	3.6	3.2	2.7
Cultural	28.8	26.2	36.3	50.7	58.7
Diplomatic	30.6	25.2	24	22	29.3
Science	31.3	37.9	19.7	14	9.8
Domination	8.3	5.8	13.7	6.7	0
Time	1.1	4.9	6.3	6.7	2.2

Kind of strange to me that Japan got absolutely gutted for being a top 10 civ while Austria remains arguably the best performer; I always found Austria way more annoying to play against personally

Heinz_Guderian · Feb 24, 2024

Interesting, many of the civs that the AI performs well as are the ones I consider among the weakest (China and Germany especially), but I guess I am not really up to date.

Btw, your first bar graph and the table for victory types doesnt match - for example it says in the table that china won 4 games with dom and 0 with science but in the chart it shows wins via science but not domination, unless I am misunderstanding of course.

L. Vern · Feb 24, 2024

Heinz_Guderian said:
Interesting, many of the civs that the AI performs well as are the ones I consider among the weakest (China and Germany especially), but I guess I am not really up to date.

Btw, your first bar graph and the table for victory types doesnt match - for example it says in the table that china won 4 games with dom and 0 with science but in the chart it shows wins via science but not domination, unless I am misunderstanding of course.

This time around the processing was done in a re-implementation of my original processing code in a new language/framework - somehow the science and domination victory column headings got swapped; I have fixed the affected tables.

Good catch, thanks for bringing it to my attention!

statusquo · Feb 25, 2024

L. Vern said:
I've already heard rumblings of a Songhai nerf

There was a time i didnt know why is it so strong, untill i saw amphibious Knight on a river

Im learning to play on diety now, and im at the stage im trying to judge any tile basing on its relative value,
and value of a production is a lot, especially early game. You get like 8 tiles of river near your capitol, and you are set up, to dominate
your neighbours. So yea, those Tabya-s seems like almost OP, not sure why there is this additional 10% towards buildings there. Its already strong

It also gives early culture from being UB, which at this stage of the game for many civs could be 50% more culture in the expansion cities.

L. Vern said:
Also, here is a video of the first 225 games

Wow, is those strange triangles are offensive roads build by new automated workers

? Impressed

And they clean after xD
Hmm, just wonder, do workers know, which roads were built by automated workers, so that whey dont remove roads that wer build by human?
I mean, i wonder what would happend to offensive roads build by human, is it a risk, that automated workers would remove it prematurely?
I have problem with current workers, that sometimes i build road not optimally, so that i could squize more villages between close cities, and i see sometimes, that
there are parralel roads build next to mine

Just learned to accept , that i cant squize 3 villages, when on optimal road could be only 2

pineappledan · Feb 25, 2024

We like war and we don’t care who knows.
Fighting my enemies and fighting my bros.

It’s my opinion that these win % graphs are basically the end goal that we had set out for ourselves more than a year ago. This looks like we have achieved terrific progress in balancing victory conditions relative to version where cultural victory was winning >70% of all games.

It is not desirable that domination victories take up any larger of a slice of the pie. In real game terms, winning via domination means eliminating players. More dead civs earlier in the game means less fun games. Having a more equitable distribution of win types is not a higher priority than fun games, so the fact that civs are winning at least some of the time via domination is more than adequate.

epicivfreak · Feb 25, 2024

Is it just me, or is it kind of disturbing that Time only has a 1.1% chance of being a factor? Certain Civs are so powerful at certain things, that 90% of games are either Cultural, Diplomatic or Scientific victories. Domination and Time are virtually pointless.

pineappledan · Feb 25, 2024

epicivfreak said:
Is it just me, or is it kind of disturbing that Time only has a 1.1% chance of being a factor?

It's just you.

Science victory should be achievable in the vast majority of games before turn 500. SV mainly requires that a civ research the entire tech tree, and is therefore a more dynamic version of the hard cutoff presented by a Time victory. Ideally, Time victory should not exist at all, because SV accomplishes the same time limit function, but in a more interactive way.

From multiple iterations of these tests, it has become clear that time victories are functionally domination victories, where a single militaristic civ has managed to cripple every other player on the map, but then decides to simply run the clock out instead of constructing spaceship parts, or conquering those last few cities.

Heinz_Guderian · Feb 25, 2024

Yea I agree, Time victory being that low is good to see.

epicivfreak · Feb 25, 2024

pineappledan said:
It's just you.

Ok, gotta admit that's a concise reply and it kinda makes sense. I'm still getting back into Vox Populi after years of playing the same 3.something version, which absolutely ruined my gameplay, in hindsight. It's not good to stick with one version for too long. :lol:

pineappledan · Feb 25, 2024

These games are run on emperor difficulty where the AI has some considerable handicap bonuses. When run at lower difficulties, time victories make up a larger share of games.

epicivfreak · Feb 25, 2024

pineappledan said:
These games are run on emperor difficulty where the AI has some considerable handicap bonuses. When run at lower difficulties, time victories make up a larger share of games.

Ok. I just thought it would make more sense for the proportions to be more equal, since I thought this was an auto-run of all-AI players.

L. Vern · Feb 25, 2024

Stalker0 said:
The pseudo don stat is interesting but still seems a minor portion of the winning civ strats.

I think the more interesting (and likely harder to get) stat is, if you assume your pseudo don as a win condition…how often would a civ actually have “won” over the civ that actually won the game?

I’m other words you are saying “this civ won, but they kind of won by a different method” but I want to know “this civ won, but if pseudo dom was a thing this other civ would have won”

Preface: to avoid confusion with the previous Pseudo-Domination victory I am defining "Authority Victory" as a win on the first turn that a civ achieves ownership of at least 2/3 of territory and cities

Here's a list of all Authority Victories along with information of the actual win for that game

Spoiler Authority Victories :

game_id	Authority Victor	AV Turn	Real Victor	Real Victory Turn	Real Victory Type
2024-02-11T00.16.40.734983	Songhai	236	Songhai	251	Domination
2024-02-13T08.17.49.416029	Songhai	247	Songhai	288	Domination
2024-02-23T11.08.36.523679	The Zulus	255	The Zulus	296	Domination
2024-02-11T07.50.41.420653	Songhai	261	Songhai	284	Domination
2024-02-14T20.18.44.086194	Songhai	263	Songhai	313	Domination
2024-02-13T22.10.53.420113	Mongolia	276	The Inca	412	Science
2024-02-17T09.42.12.528199	Denmark	285	Denmark	327	Domination
2024-02-21T08.58.43.676421	The Iroquois	295	The Iroquois	356	Domination
2024-02-15T06.21.52.499974	Rome	298	Rome	327	Domination
2024-02-21T11.29.21.372145	Songhai	300	Songhai	309	Domination
2024-02-21T16.08.39.007682	Denmark	310	The Maya	418	Science
2024-02-18T18.02.58.818224	Songhai	317	Songhai	324	Domination
2024-02-15T07.48.32.299210	Denmark	321	Denmark	338	Domination
2024-02-19T09.20.51.789619	Persia	325	Persia	353	Domination
2024-02-15T15.48.43.378448	Songhai	328	Songhai	478	Diplomatic
2024-02-16T13.53.24.491920	The Iroquois	330	The Iroquois	436	Science
2024-02-14T11.06.00.843948	Rome	339	Rome	392	Diplomatic
2024-02-10T20.52.41.547925	Songhai	345	Venice	363	Diplomatic
2024-02-18T19.27.47.727314	Persia	351	Persia	376	Domination
2024-02-22T11.50.54.546380	Sweden	352	Sweden	379	Domination
2024-02-18T08.47.34.489964	France	361	France	490	Science
2024-02-15T21.59.32.589359	The Iroquois	362	The Iroquois	429	Domination
2024-02-22T09.33.07.499233	Russia	362	The Celts	408	Cultural
2024-02-14T13.14.37.118004	The Netherlands	362	The Netherlands	392	Diplomatic
2024-02-22T04.08.52.449756	China	363	China	440	Science
2024-02-21T01.42.41.505082	Sweden	369	Sweden	403	Domination
2024-02-16T00.32.14.502454	Songhai	372	Songhai	422	Domination
2024-02-22T14.08.00.477711	Assyria	372	Assyria	372	Diplomatic
2024-02-07T17.08.46.649795	Sweden	373	Sweden	400	Domination
2024-02-13T10.54.39.989786	Denmark	375	Ethiopia	466	Science
2024-02-13T02.04.10.079379	Sweden	381	Sweden	481	Cultural
2024-02-19T22.45.23.423739	The Aztecs	382	The Aztecs	404	Diplomatic
2024-02-13T06.08.21.893802	Songhai	390	Songhai	401	Diplomatic
2024-02-22T01.14.24.233770	The Zulus	393	The Zulus	489	Science
2024-02-13T22.47.42.214257	Carthage	396	Carthage	402	Diplomatic
2024-02-19T10.58.54.629149	Songhai	400	Carthage	447	Diplomatic
2024-02-12T00.55.06.623287	America	401	America	444	Domination
2024-02-21T04.31.51.021538	Denmark	402	Persia	499	Science
2024-02-15T09.25.01.958476	The Iroquois	408	The Iroquois	429	Domination
2024-02-17T00.59.48.672601	Siam	411	Siam	411	Diplomatic
2024-02-11T22.07.32.813727	Denmark	412	Austria	433	Diplomatic
2024-02-11T02.24.38.626345	Songhai	442	Songhai	490	Science
2024-02-17T05.06.08.745181	Sweden	445	Sweden	461	Domination
2024-02-20T15.29.42.968048	The Aztecs	449	The Inca	449	Science
2024-02-18T01.43.55.417040	Sweden	472	Sweden	479	Science

Here's a list of failed domination games - the above list filtered for games where an Authority victory was achieved but someone else actually won the game

Spoiler failed domination games :

KungCheops · Feb 26, 2024

statusquo said:
There was a time i didnt know why is it so strong, untill i saw amphibious Knight on a river

Im learning to play on diety now, and im at the stage im trying to judge any tile basing on its relative value,
and value of a production is a lot, especially early game. You get like 8 tiles of river near your capitol, and you are set up, to dominate
your neighbours. So yea, those Tabya-s seems like almost OP, not sure why there is this additional 10% towards buildings there. Its already strong
It also gives early culture from being UB, which at this stage of the game for many civs could be 50% more culture in the expansion cities.

Wow, is those strange triangles are offensive roads build by new automated workers ? Impressed
And they clean after xD
Hmm, just wonder, do workers know, which roads were built by automated workers, so that whey dont remove roads that wer build by human?
I mean, i wonder what would happend to offensive roads build by human, is it a risk, that automated workers would remove it prematurely?
I have problem with current workers, that sometimes i build road not optimally, so that i could squize more villages between close cities, and i see sometimes, that
there are parralel roads build next to mine Just learned to accept , that i cant squize 3 villages, when on optimal road could be only 2

I have considered adding a flag for "human created roads", making it so automated workers don't remove them, and perhaps also highly prioritize using them for city connections.

CppMaster · Feb 26, 2024

epicivfreak said:
Is it just me, or is it kind of disturbing that Time only has a 1.1% chance of being a factor? Certain Civs are so powerful at certain things, that 90% of games are either Cultural, Diplomatic or Scientific victories. Domination and Time are virtually pointless.

It's not just you. Time victory is the most "natural" victory. You win, because you've created the greatest empire.
The other victories: launching a rocket, winning a vote, being the most popular feel kinda little gamey, yet the game is being balanced around them anyway. :rolleyes:

Rustonreddit · Feb 26, 2024

CppMaster said:
It's not just you. Time victory is the most "natural" victory. You win, because you've created the greatest empire.
The other victories: launching a rocket, winning a vote, being the most popular feel kinda little gamey, yet the game is being balanced around them anyway.

Unlike the victory points that are very "natural" and not gamey at all. Also, the world just stops at 2050 ad. very natural

pineappledan · Feb 26, 2024

Here is a graph showing win rates for all civs over time. I know it's a little overwhelming to parse at a glance, but I find this easier to read than the table.

I have also attached the excel sheet so people can play around with it, delete some lines and remove clutter, etc.

Spoiler :

If you use variance in the win rates as a heuristic for civ balance, then it has gotten worse from 3.1 to 4.5. We can see the lines converging closer together, and then splitting apart:

version	2.7	3.2	3.6	3.1	4.5
winrate var	0.013651	0.007013	0.004736	0.004541	0.008441

epicivfreak · Feb 26, 2024

CppMaster said:
It's not just you. Time victory is the most "natural" victory. You win, because you've created the greatest empire.
The other victories: launching a rocket, winning a vote, being the most popular feel kinda little gamey, yet the game is being balanced around them anyway.

Go support Saga of Man in any way possible for you

game_id	Turn	Victory Civ	Owned Territory Percent	Owned Cities Percent	Victory Type
2024-02-18T01.43.55.417040	479	Sweden	0.686901	0.695652	Science
2024-02-13T22.47.42.214257	402	Carthage	0.809038	0.808824	Diplomatic
2024-02-15T15.48.43.378448	478	Songhai	0.783427	0.797101	Diplomatic
2024-02-19T22.45.23.423739	404	The Aztecs	0.835094	0.836066	Diplomatic
2024-02-22T04.08.52.449756	440	China	0.715509	0.689655	Science
2024-02-13T06.08.21.893802	401	Songhai	0.772095	0.772727	Diplomatic
2024-02-17T00.59.48.672601	411	Siam	0.687927	0.666667	Diplomatic
2024-02-18T08.47.34.489964	490	France	0.76601	0.723077	Science
2024-02-14T13.14.37.118004	392	The Netherlands	0.742896	0.71875	Diplomatic
2024-02-22T01.14.24.233770	489	The Zulus	0.895785	0.875	Science
2024-02-11T02.24.38.626345	490	Songhai	0.814072	0.828571	Science
2024-02-22T14.08.00.477711	372	Assyria	0.688261	0.679245	Diplomatic
2024-02-16T13.53.24.491920	436	The Iroquois	0.745914	0.75	Science
2024-02-14T11.06.00.843948	392	Rome	0.807518	0.811594	Diplomatic
2024-02-13T02.04.10.079379	481	Sweden	0.865801	0.87931	Culture

4.5 AI-Only Game Statistics (we love war)

Warlord

Emperor

Baller Magnus

Warlord

Prince

Warlord

Warlord

Deity

Prince

Deity

Prince

Prince

Deity

Prince

Warlord

Warlord

Deity

Chieftain

Deity

Attachments

Prince

Similar threads