AC 2v2 statistics (Wildcard, QT1, QT2, Finals)

UPDATE: Now also contains games for Qualification tournament 1 (12.11.2018)
UPDATE: Now also contains games for Qualification tournament 2 (26.11.2018)
UPDATE: Added games for the final round (12.12.2018)

Ok, cheap-ass job here, these are the stats for the ac 2v2 tournament. Brackets are here. Also, added the games for the first qualification tournament (brackets). And now also for the 2nd qualifier (brackets). Finally, brackets for the final round are here.

The total number of games I found are 161 (64 from the wildcard, 37 from qualifier one, 33 from qualifier two, 27 for the finals). According to the brackets there should be one more game in the wildcard round, but I wasn't able to find the second game of Barton/Frost vs. Bulat/LUCKBAR.

Wildcard
Show Spoiler
Qualification 1:
Show Spoiler
Qualification 2:
Show Spoiler
Finals
Show Spoiler

Total number of games: 161

Factions

	ostheer	soviets	okw	usf	ukf
picked	163	184	159	45	93
won	74	99	76	19	54
won(%)	45	54	48	42	58
Std. dev.(%)	5	4	4	11	4

Teams

	ostheer/ostheer	ostheer/okw	okw/okw	soviets/soviets	soviets/usf	soviets/ukf	usf/usf	usf/ukf	ukf/ukf
Picked	32	99	30	51	17	65	6	16	6
Won	12	50	13	28	6	37	2	9	4
Won(%)	38	51	43	55	35	57	33	56	67
Std. dev.(%)	9	7	4	5	12	5	24	18	17

Maps

games	avg. length (min)	max length (min)	map name
63	34	69	Elst Outskirts
34	35	65	Crossing in the Woods
32	33	54	Rails & Metal
20	42	76	Alliance of Defiance
12	41	64	Vaux Farmlands

Map stats

Elst Outskirts

Factions:

	ostheer	soviets	okw	usf	ukf
picked	57	67	69	21	38
won	23	36	33	11	23
won(%)	40	54	48	52	61

Teams:

	ostheer/ostheer	ostheer/okw	okw/okw	soviets/soviets	soviets/usf	soviets/ukf	usf/usf	usf/ukf	ukf/ukf
Picked	11	35	17	17	7	26	4	6	3
Won	2	19	7	10	2	14	2	5	2
Won(%)	18	54	41	59	29	54	50	83	67

Crossing in the Woods

Factions:

	ostheer	soviets	okw	usf	ukf
picked	29	36	39	7	25
won	13	20	19	1	15
won(%)	45	56	49	14	60

Teams:

	ostheer/ostheer	ostheer/okw	okw/okw	soviets/soviets	soviets/usf	soviets/ukf	usf/usf	usf/ukf	ukf/ukf
Picked	2	25	7	7	1	21	2	2	1
Won	0	13	3	4	0	12	0	1	1
Won(%)	0	52	43	57	0	57	0	50	100

Rails & Metal

Factions:

	ostheer	soviets	okw	usf	ukf
picked	39	41	25	6	17
won	19	22	13	2	8
won(%)	49	54	52	33	47

Teams:

	ostheer/ostheer	ostheer/okw	okw/okw	soviets/soviets	soviets/usf	soviets/ukf	usf/usf	usf/ukf	ukf/ukf
Picked	9	21	2	15	1	10	0	5	1
Won	5	9	2	8	0	6	0	2	0
Won(%)	56	43	100	53	0	60	-	40	0

Alliance of Defiance

Factions:

	ostheer	soviets	okw	usf	ukf
picked	22	28	18	6	6
won	12	13	8	3	4
won(%)	55	46	44	50	67

Teams:

	ostheer/ostheer	ostheer/okw	okw/okw	soviets/soviets	soviets/usf	soviets/ukf	usf/usf	usf/ukf	ukf/ukf
Picked	5	12	3	10	4	4	0	2	0
Won	3	6	1	4	2	3	0	1	0
Won(%)	60	50	33	40	50	75	-	50	-

Vaux Farmlands

Factions:

	ostheer	soviets	okw	usf	ukf
picked	16	12	8	5	7
won	7	8	3	2	4
won(%)	44	67	37	40	57

Teams:

	ostheer/ostheer	ostheer/okw	okw/okw	soviets/soviets	soviets/usf	soviets/ukf	usf/usf	usf/ukf	ukf/ukf
Picked	5	6	1	2	4	4	0	1	1
Won	2	3	0	2	2	2	0	0	1
Won(%)	40	50	0	100	50	50	-	0	100

Wildcard:
Most damage done by Zodiak! (101891)
Longest game was TwistedTootsy/Dr. Zaius vs. VonManteuffel/ElZocko on Crossing which took 65 minutes.

Qualification one:
Most damage done by Talisman (154180) who played (together with Tüme) vs. Price and Prodigly on AOD in the so far longest match overall (76 minutes).

Qualification two:
Most damage was done by Rutra (94529) playing with Dzaraa vs. Hans and DevM on AOD in what also was the longest match of that tournament (73 minutes).

Finals:I
The most damage was done by Isildur (110352) in what also was the longest game (68 Minutes) in the semi-finals playing with VonAsten against Kimbo and Stuve on Elst outskirts.

Thanks for your work.

I'd like to ask, since faction pick is a categorical choice, what are you using to calculate standard deviation?

I'd like to ask, since faction pick is a categorical choice, what are you using to calculate standard deviation?

Standard deviations are given for the win% and are estimated with a bootstrapping method.

It's a pretty rough estimate anyways, in particular given the low number of samples and that I ignore the granularity of the problem. Main reason for me to include anything like that is mostly to remind people that they should take these numbers with a grain of salt and not start threads like "Stats prove Axis wins 51% of time, nerf plz!"

Standard deviations are given for the win% and are estimated with a bootstrapping method.

It's a pretty rough estimate anyways, in particular given the low number of samples and that I ignore the granularity of the problem. Main reason for me to include anything like that is mostly to remind people that they should take these numbers with a grain of salt and not start threads like "Stats prove Axis wins 51% of time, nerf plz!"

I completely understand, I'm actually curious about the calculation outside of COH2 and am asking it from the mindset of statistics. I don't care to make grand assumptions about balance from these values, though I appreciate you sharing them if only to see what's popular and what's not.

I'm used to St. Dev = https://study.com/cimages/multimages/16/standard-deviation-formulas.jpg but those require ratio/interval data, and I was curious what you used to get an St.Dev for categorical variables.

Its always nice to have numbers, but these will be heavily skewed by very wide skill bracket of teams and individual players.

I'm used to St. Dev = https://study.com/cimages/multimages/16/standard-deviation-formulas.jpg but those require ratio/interval data, and I was curious what you used to get an St.Dev for categorical variables.

Yeah, well, there are different approaches of how you can tackle this (see e.g. Bootstrapping).

Details vary, but the idea is basically that instead of computing the win% from all games, you use different subsets of games and compute win% only for those. You end up with a different value for each subset. You can then estimate the devitation from the values for the different subsets.

lol allies op

Yeah, well, there are different approaches of how you can tackle this (see e.g. Bootstrapping).

Details vary, but the idea is basically that instead of computing the win% from all games, you use different subsets of games and compute win% only for those. You end up with a different value for each subset. You can then estimate the devitation from the values for the different subsets.

I am familiar with Bootstrapping (though haven't really used it in... well, forever I guess outside of Uni). Thanks for the explanation though, I'll stop derailing the thread now ;D

Who played USF/UKF?
That low ammount of games played but very high win % makes me think of a few players using that combo.

UKF 65% win rate Nerf em

Who played USF/UKF?
That low ammount of games played but very high win % makes me think of a few players using that combo.

Ciez & GiaA (3 games)
supajou & Haha hey (4 games)
ElZocko & VonManteuffel (2 games)
Stark & Sully (3 games)

Ciez & GiaA (3 games)
supajou & Haha hey (4 games)
ElZocko & VonManteuffel (2 games)
Stark & Sully (3 games)

Thank you.
USF and UKF do work together fairly well, will have to study these games.

As always well done

Looking forward for the 2 Quali cups and the final stats.

And yes USA can be very powerfull. There are 2-3 good tactics, which work

Thanks for info man.

Thank you Siphon

Updated to include the results from the Qualification tournament 1.

OMG!
The USF pick and win ratio is just sad.

well always interesting to see those numbers, thanks for the work done here and god god, poor USF Kappa.

Sorry, just noticed that instead of the total numbers for both tourneys, I posted the qualifier twice. Fixed now (individual numbers still available in the spoilers)...

OMG!
The USF pick and win ratio is just sad.

Well, USF bring nothing to early game that soviets can't bring and have inferior late game AND can't counter sniper at all.
Not much point to bring them to MP.