[KotD] Clash of the Civilizations (Stats)

[Rorarimbo]

The King of the Desert tournament will provide us with great fun and will determine who is the best Arabia player in the moment. But it also has another purpose. We are often wondering how the new changes affect the game? What is the new balance and the new meta? I think despite it is a bit early for established meta the way it is in standard AoC we will see some new trends as there were with the indians earlier this year. Collecting the results from all the games of KotD we will try to answer which civilization is the King of the Desert. This will be a race between the civs not between the players.

There are 4 categories in witch the civs will be ranked that will also be contributing to the final results.

I. Most wanted civilization. Each time a civ is selected by a player in his civilization pool it well get 1 point in this category and 1 point in the overall standings.
II.Most popular civilization. Each time a civ is played it will get 1 point in this category and 1 point in the overall standings.
III. Strongest civilization. Each time a civ wins a game it will get 1 point in this category and 1 point in the overall standings.
IV. Most feared civilization. Each time a civ is banned it will get 1 point in this category and 2 points in the overall standings.
The overall standings for The Best civilization in the tournament will be determined by the sum of the points as listed above.

RESULTS

All the results I am going to refer plus much more You can see in this link: https://docs.google.com/spreadsheets/d/ ... sp=sharing The end product as a collective work done by philipp94831 who created a tool to extract information from the recored games and gathered alot of interesting stats. He also prepared very interesting win probabolity rating and soccer table. misafeco added Civ ELO from the results of the games. And myself preparing the standings (overall and in different categories) as well as draft preference rating.

I. KotD Overall Standings
1 Aztecs 68
2 Indians 68
3 Mayans 67
4 Berbers 67
5 Incas 59
6 Malians 59
7 Burmese 58
8 Magyars 55
9 Ethiopians 55
10 Malay 52
11 Huns 46
12 Celts 28
13 Britons 23
14 Mongols 17
15 Vikings 11
16 Italians 10
17 Viatnamese 10
18 Franks 9
19 Byzantines 8
20 Spanish 8
21 Japanese 7
22 Chinese 6
22 Goths 6
24 Koreans 5
25 Portuguese 2
26 Slavs 1
26 Turks 1
Tie breakers: 1) number of wins; 2) number of bans; 3) games played

I.1 Most Wanted Civilization (Most times selected in civ draft)
1 Indians 28
2 Ethiopians 26
3 Magyars 26
4 Berbers 25
5 Mayans 23
6 Burmese 23
7 Aztecs 22
8 Malians 21
9 Malay 21
10 Huns 20
11 Incas 19
12 Celts 12
13 Britons 11
14 Mongols 7
15 Vikings 6
15 Italians 6
17 Viatnamese 5
18 Franks 3
18 Japanese 3
20 Spanish 3
20 Chinese 3
20 Goths 3
20 Koreans 3
24 Byzantines 2
25 Slavs 1
25 Turks 1
Tie breaker: number of bans

I.2 Most Popular Civilization (Most number of games played)
1 Berbers 24
2 Indians 23
3 Mayans 21
4 Aztecs 21
5 Malay 21
6 Magyars 20
7 Malians 18
8 Ethiopians 18
9 Incas 16
10 Burmese 15
11 Huns 14
12 Celts 8
13 Britons 7
14 Mongols 5
15 Viatnamese 5
16 Vikings 4
17 Italians 4
18 Spanish 3
19 Franks 2
20 Japanese 2
21 Chinese 2
21 Goths 2
23 Byzantines 2
24 Koreans 1
Tie breaker: draft selections

I.3 Strongest Civilization (Most number of wins)
1 Aztecs 15
2 Indians 13
3 Mayans 11
4 Huns 10
5 Burmese 10
6 Incas 10
7 Malians 10
8 Berbers 10
9 Magyars 7
10 Malay 6
11 Britons 5
12 Ethiopians 5
13 Celts 4
14 Mongols 3
15 Franks 2
15 Byzantines 2
17 Spanish 2
18 Koreans 1
19 Chinese 1
19 Goths 1
21 Vikings 1
Tie breaker: least number of games

I.4 Most Feared Civilization (Most number of bans)
1 Incas 7
2 Mayans 6
3 Aztecs 5
4 Burmese 5
4 Malians 5
6 Berbers 4
7 Ethiopians 3
8 Indians 2
9 Malay 2
10 Celts 2
11 Huns 1
12 Magyars 1
13 Mongols 1
14 Franks 1
14 Byzantines 1
16 Japanese 1
16 Portuguese 1
Tie breaker:most number of wins

II.Win Probability Rating
1 Aztecs 9.00%
2 Indians 8.18%
3 Mayans 7.50%
4 Berbers 7.14%
5 Huns 6.33%
6 Burmese 6.01%
7 Incas 5.74%
8 Malians 5.52%
9 Malay 5.08%
10 Magyars 4.73%
11 Britons 4.44%
12 Byzantines 3.98%
13 Ethiopians 3.74%
14 Mongols 3.03%
15 Celts 2.98%
16 Spanish 2.96%
17 Franks 2.78%
18 Koreans 2.35%
19 Chinese 1.78%
20 Vikings 1.67%
21 Goths 1.59%
22 Vietnamese 1.27%
23 Italians 1.16%
24 Japanese 1.05%

III. Civilization ELO
1 Aztecs 1702
2 Huns 1673
3 Burmese 1649
4 Incas 1641
5 Britons 1639
6 Indians 1636
7 Byzantines 1631
8 Franks 1629
9 Mongols 1624
10 Spanish 1614
11 Mayans 1610
12 Celts 1608
13 Chinese 1601
14 Goths 1599
15 Malians 1587
16 Japanese 1571
17 Vikings 1570
18 Berbers 1550
19 Magyars 1542
20 Italians 1540
21 Vietnamese 1532
22 Ethiopians 1518
23 Malay 1502
Only 1 game: Koreans 1618; Not played: Teutons, Persians, Turks, Saracens, Slavs, Khmer, Portuguese.

Both of the above models consider the civ strenght in relation the the strenght of thopposition they meet. I have to agree that the classic ELO would work better if we had all civilizations involved as of now fairly strong civs are towards the bottom of the ranking having played with only strong competition. For reference You cab see below the actual Win ration.

IV Win Ratio
1 Aztecs 71.43% 15/6
2 Huns 71.43% 10/4
3 Britons 71.43% 5/2
4 Burmese 66.67% 10/5
5 Spanish 66.67% 2/1
6 Incas 62.50% 10/6
7 Mongols 60.00% 3/2
8 Indians 56.52% 13/10
9 Malians 55.56% 10/8
10 Mayans 52.38% 11/10
11 Celts 50.00% 4/4
12 Berbers 41.67% 10/14
13 Magyars 35.00% 7/13
14 Malay 28.57% 6/15
15 Ethiopians 27.78% 5/13
16 Vikings 25.00% 1/3
17 Vietnamese0.00% 0/5
18 Italians 0.00% 0/4
Civilizations with less than 3 games not included.

V. Draft Preference Rating
1 Berbers 2.89
2 Indians 2.93
3 Mayans 4.64
4 Malians 4.69
5 Ethiopians 5.46
6 Burmese 6.08
7 Aztecs 6.27
8 Magyars 7.17
9 Incas 7.28
10 Malay 8.17
11 Huns 8.73
12 Celts 9.66
13 Britons 9.87
14 Viatnamese 10.35
15 Italians 10.39
16 Mongols 10.43
17 Goths 10.71
18 Franks 10.73
19 Vikings 10.74
20 Japanese 10.77
21 Spanish 10.81
22 Koreans 10.9
23 Byzantines 10.97
24 Chinese 11
24 Khmer 11
24 Persians 11
24 Portuguese 11
24 Saracens 11
24 Slavs 11
24 Teutons 11
24 Turks 11
Calculation includes the orders civilisatons were picked (1/10). All cases civs were not picked or were picked 11th/18th they are calculated as 11 -not wanted with 1 being first selection. The lower the number is the higher the prerence for that civization was during the drafts. Bans are not included.

VI. Soccer Table
1 Aztecs 36
2 Huns 36
3 Indians 34
4 Incas 31
5 Britons 31
6 Mayans 31
7 Malians 30
8 Berbers 30
9 Burmese 28
10 Byzantines 27
10 Franks 27
12 Mongols 27
13 Spanish 26
14 Celts 26
15 Koreans 25
16 Chinese 24
16 Goths 24
18 Vikings 22
19 Magyars 22
20 Japanese 21
21 Italians 20
22 Malay 20
23 Vietnamese 18
24 Ethiopians 18
Its done according the head to head results between the civs. Match ups that did not play together are concidered as a draw. Each win no matter the result is awarded 3 pts and each draw is awarded 1 pt.
Tie breaker is goal difference

As stated above You can check the link for much more interesting stats.

 

[buddy__1]

Re: KotD Clash of the Civilizations

whats the difference from 1 and 2

 

[Rorarimbo]

Re: KotD Clash of the Civilizations

In a bo5 series that finish 3-0 each player selects 5 civs in their civ pool but at the end play with only 3 of them. The first one shows their intention to use a civ the 2nd that they actually used it. For the overall standings a civ that is used will get 2 pts with a chance for a 3rd if it wins while if it was only selected in the pool but not used it gets only 1 pt.

 

[nimanoe]

Re: KotD Clash of the Civilizations

I think it would be better to either scrap rule 2 or give less points for it, since pro's might save their best civ(s) for last (especially for the person winning).

 

[swissboyy]

Re: KotD Clash of the Civilizations

Sounds really cool, looking forward to it!

 

[Rorarimbo]

Re: KotD Clash of the Civilizations

Thanks for the suggestion. I think the important part is collecting the data than everyone could reevaluate the formula as he wishes. Having said that I hope we have access to the drafting links as if we dont the whole idea is GG.

For your suggestion Im not too sure how players would decide their strategy. Loosing players will use their strogest civs for sure while winning players may choose differently but here it comes to anticipation and strategy we cannot evaluate really. Than it comes to the player actually trusting the civilisation giving it priority compared the others.

 

[DraCoNT]

Re: KotD Clash of the Civilizations

Nice idea, but great way to see this would be just the order of the picks. I'd say that civs like malay and berbers will both be likely picked in almost any set, but almost none of the players would pick malay over berber and none of the categories is really reflecting that and in the end you might end up with malay winning a bit more games than berbers (because berb will be used earlier vs stronger civs possibly) and found out that malay is better. The order of the picks is really important and relevant. Maybe like giving first 3 civs picked 2 points, remaining 1 point. Then you likely have to change the banned civ for 3 points overall, because the civ is very likely gonna miss between 1+1 (picked 4th+/played), 2+1 (picked 1st - 3rd/played), 2+1+1 (picked 1st - 3rd/played/won), 1+1+1 (picked 4th+, played, won).

And about the discussion - I doubt players won't play their civs in very similar order they picked those. You want to play strongest civ earlier, weaker civs later. If you're winning 2:0 in BO5, would you save indians for the last game? Only reason is like if you have magyars/huns as very high pick (I'd say players not that comfortable with expansions will have that) and you don't want to use them when its very likely enemy will pick indians, otherwise you want to play civs according to their "power". Especially top 5 - 10 players want to play stronger civ vs stronger civ, weaker vs weaker later on.

 

[nimanoe]

Re: KotD Clash of the Civilizations

Nice idea, but great way to see this would be just the order of the picks. I'd say that civs like malay and berbers will both be likely picked in almost any set, but almost none of the players would pick malay over berber and none of the categories is really reflecting that and in the end you might end up with malay winning a bit more games than berbers (because berb will be used earlier vs stronger civs possibly) and found out that malay is better. The order of the picks is really important and relevant. Maybe like giving first 3 civs picked 2 points, remaining 1 point. Then you likely have to change the banned civ for 3 points overall, because the civ is very likely gonna miss between 1+1 (picked 4th+/played), 2+1 (picked 1st - 3rd/played), 2+1+1 (picked 1st - 3rd/played/won), 1+1+1 (picked 4th+, played, won).

And about the discussion - I doubt players won't play their civs in very similar order they picked those. You want to play strongest civ earlier, weaker civs later. If you're winning 2:0 in BO5, would you save indians for the last game? Only reason is like if you have magyars/huns as very high pick (I'd say players not that comfortable with expansions will have that) and you don't want to use them when its very likely enemy will pick indians, otherwise you want to play civs according to their "power". Especially top 5 - 10 players want to play stronger civ vs stronger civ, weaker vs weaker later on.

You make some good points, but I'm still not convinced that a player leading 2-0 will pick his next best civ instead of a slightly weaker one, since most likely the enemy will pick his best civ, so he knows that for the next match-up he will have a civ advantage.

I think another solution might be to look at the order the players picked the civs in the civ selection, since I don't think they won't pick weaker civs first there. You could then apply some scoring system to differentiate between them, for example: first two civs banned get 3 points, first two civs picked get 2.5 points, next two civs picked get 2 points etc.

Also I'm not sure how much weight should be put on the banned civs, since it's global bans, so players might ban some bad civs because they feel they are better with the best civ because they practiced more with them, but that's why it's nice there's different divisions :D

 

[Rorarimbo]

Re: KotD Clash of the Civilizations

The order civs are picked is very important you both are right about that. I think we can see one or few civs that appear in every draft selected or banned. In that case even the most wanted ranking would benefit if we can see what civ was picked first and what was picked last. But at that stage my plan is to keep it as simple as possible. Also I want to see how things work with the draft links. The first week will be crucial if we get them all available. If we have them will be possible for us to make even more breakdowns.

Also I think that even if we make few changes the result will not be changed a lot. This we will see at the end of course but if you use the berbers and malay example and if berbbers are picked earlier and played earlier its likely they are played in more games as some series will finish before the malay get on the pitch . This would give more points to the berbers for being played and potentionally for winning as there will be more chances to win. Btw thats why I decided not to bother with the W/L ratio just add a point for each win.

As I said I want to keep it relatively simple and allow the results speak for themselves as nimanoe noted the different categories will also be giving information while the final result is matter of a formula between them.

 

[kw1k000000]

Re: KotD Clash of the Civilizations

Baring first round there are global civ bans as well. You may want to consider this as first priority civ considering the civs perceived as OP/Top Tier will get banned first.

 

[nimanoe]

Re: KotD Clash of the Civilizations

Baring first round there are global civ bans as well. You may want to consider this as first priority civ considering the civs perceived as OP/Top Tier will get banned first.

IV. Most feared civilization. Each time a civ is banned it will get 1 point in this category and 2 points in the overall standings.

 

[Cutht]

Re: KotD Clash of the Civilizations

It's going to be intresting to see how many AoK civs will be picked.

 

[jojkov]

Re: KotD Clash of the Civilizations

It's going to be intresting to see how many AoK civs will be picked.

Well, in first match there were 8 out of 10 new civs. (6 if you dont count FE) And zero AoK civs. It is pretty extreme, considering you cant repeat any civ.
Max vs Lyx

 

[patao]

Re: KotD Clash of the Civilizations

And the 2 old civs were both Meso of course, despite all the balance changes to old civs, Meso seems as good or maybe even better than ever :lol:

 

[Rorarimbo]

Re: KotD Clash of the Civilizations

And the 2 old civs were both Meso of course, despite all the balance changes to old civs, Meso seems as good or maybe even better than ever :lol:

well we had two strong american civs , now we have three. ( i think incas are not meso civ technicly). And all of them picking up a win today, too. But way to go. just 1 series played 30 more to go. After few days when Round 1 is over we will know alot more.

 

[Rorarimbo]

Re: KotD Clash of the Civilizations

Added link to a data file in the first post. SPOILERS. the link includes the games already played in the first three days of KotD.

Sofar the Mayans doing the best. Berbers, Indians, Aztecs and Magyars follow.

Have in mind that Tim and Yellow had one random Civ pick each ( indians and Slavs) wich Im not counting as they did not make the choice themselves but the tool made it for them.

 

[SiFly]

Re: KotD Clash of the Civilizations

Baring first round there are global civ bans as well. You may want to consider this as first priority civ considering the civs perceived as OP/Top Tier will get banned first.

IV. Most feared civilization. Each time a civ is banned it will get 1 point in this category and 2 points in the overall standings.

still.. 2 points is as much as any civ played while it was higher priority, and 1 point less if the civ wins that game.

 

[nimanoe]

Re: KotD Clash of the Civilizations

Baring first round there are global civ bans as well. You may want to consider this as first priority civ considering the civs perceived as OP/Top Tier will get banned first.

IV. Most feared civilization. Each time a civ is banned it will get 1 point in this category and 2 points in the overall standings.

still.. 2 points is as much as any civ played while it was higher priority, and 1 point less if the civ wins that game.

The problem is that people might ban Koreans against tatoh, not because it's the best civ, but because tatoh can be really annoying with this civ, but it doesn't mean Koreans is one of the best civs on arabia.

So it's pretty much impossible to have a perfect solution, but imo this one is pretty good.

 

[Rorarimbo]

Re: KotD Clash of the Civilizations

Added link to a data file in the first post. SPOILERS. the link includes the games already played in the first three days of KotD.

Sofar the Mayans doing the best. Berbers, Indians, Aztecs and Magyars follow.

Have in mind that Tim and Yellow had one random Civ pick each ( indians and Slavs) wich Im not counting as they did not make the choice themselves but the tool made it for them.

Updated the first post with the games from Saturday ( Be careful for spoilers). Mayans still leading 1 pt in front of Berbers. Indians, Aztecs, Malians follow.

 

[swissboyy]

Re: KotD Clash of the Civilizations

Very cool, thank you!
Would've done something similar myself but yours is already perfect. Pinned the topic to the top.

 

[DraCoNT]

I believe that maya is the best civ for 1v1 arabia both in AoC and WK + solid to great counter for many other top civs, while the best (and I believe only) counter civ is malians and that is still very close matchup. Another civ that I expect to get higher and higher is malay and I'd be surprised if they won't end in top 3 spots.
My top 10 prediction is:
1) Mayans
2) Malay
3) Malians
4) Aztecs
5) Ethiopians
6) Berbers
7) Incas
8) Indians
9) Magyars
10) Huns

 

[Tarsiz]

We'll see how it goes as the tournament progress.
Some civs get really popular at the moment but are not winning many games.

Ethiopians were played 10 times and only won 2 of these games (20%). Magyars are not faring much better, winning 33% of their games so far (3 out of 9).
On the other hand, Burmese have the highest winrate - of the civs that were played more than 4 times - with 83% (5 wins in 6 games), followed closely by Malians with 75% (6 out of 8) and Huns (3 out of 4).
Aztec and Incas are also doing pretty well.

Interesting fact, the three most played civs, Mayans, Berbers and Indians, all have an exact 50% winrate. It would make sense, since they are most likely to be playing against each other (most often played civs), if they are equally strong.

More detailed statistics, including detailed H2H match-ups between civs, would be interesting to study further in the tournament.

 

[philipp94831]

More detailed statistics, including detailed H2H match-ups between civs, would be interesting to study further in the tournament.

I wrote a tool to collect such statistics.

The most frequent match-ups were Eth 1-2 Mala, May 0-2 Mala, Azt 2-0 Ind, Inc 1-1 Mal, Ber 2-0 Eth, Mag 2-0 Ber, Eth 0-2 Mal, May 2-0 Mag, Ind 2-0 Ber, Inc 1-1 Ber, May 0-2 Mal, Ind 0-2 Bur, May 0-2 Ber

The win ratios by AoE Expansions: AOC 59%, AOK 50% AOFE 48%, AOAK 47%, AOR 45%

Most trained unit lines: EW 3082, Archer 2141, Skirm 1845, Spear 1457, Knight 1297, Karambit 1049, Militia 997, SC 838, Plume 759, Camel 758

34 games went imp, 18 went castle and 4 ended in feudal

TaToH's games took 27:22 (shortest) on avg while Viper's took 53:03 (longest)

I can collect more statistics from the recs if someone is interested

 

[_Melkor]

I assumed that aztecs was the most popular 1v1 civ in old aoc. What changed this to mayans in wololo?
I would have thought that the aztecs benefited more from the new bonuses.

 

[swissboyy]

I can collect more statistics from the recs if someone is interested

Yes, please!
The more the better!