Stats are in... Some obvious factions to address!

Faction	1v1	2v2	3v3	4v4	Totals	Overall Win Streak Trend
Brit	97	36	56	55	244	21.35%
Amer	99	28	27	57	211	18.46%
Sov	123	61	24	67	275	24.06%
Totals (Weighted)	212.6667	83.33333	71.33333	119.3333	486.6667
Ost	48	40	25	45	158	13.82%
Okw	31	78	83	63	255	22.31%
Totals	79	118	108	108	413
					1143
Allied Win Ratio	2.691983	0.706215	0.660494	1.104938	1.17837

I know a lot of people like react and say axis bias to my observations, but I think using win streak is a good barometer of how the factions are doing among the top 10. Sure you could get crazy into the numbers but with all the talk from the DBP team that the Ost was "more or less" in a good spot. I gotta say the numbers don't back that up. When one of the goals of the DBP was reignite underutilized units, I'd say they had a massive fail by not addressing the Ost Tier 4 in any positive direction towards increasing their game play.

Here are my takeaways from some of the data.

1. OKW and Ost in 1v1 are at a serious disadvantage now having a losing streak disadvantage of 2.7 for every allied win. DBP has effectively killed 1v1 matches. Equal skill, Equal rank favors heavily towards an allied win. Soviets being the most dominant in 1v1.
2. Surprisingly the likely combination of Ost/Okw or Okw/Okw in 2v2, 3v3, and 4v4 tend to fair better. Okw has the higher win streaks in the category offsetting the lackluster Ost. So for all the complaints about the Jackson being OP, in 2v2+ games it doesn't seem to be affecting the win streak stats when up against an axis combo. If you're in a team game with all Ost then you're more than likely to lose.
3. 4v4 games seem to be the most balanced with allied having a +10% win streak. Who woulda thought 4v4 would be the most balanced?

So it seems 1v1 in competitive play is completely broken now. You're more than likely to lose playing axis than allies(of any faction). Ost by virtue of its overall win streak of 13.82% is by far the lowest of the factions. If there is any more gas in the tank to roll out a fix soon, fix 1v1 and buff select Ost units to make them more competitive. There have been a ton of good suggestions in other posts for Ost so I won't suggest here. I know I've heard/seen interviews where balance team errs on decreasing OP units than doing buffs but when its clear all other units are OP to Ost then Ost needs a buff. Just do something and do it quick before Relic goes belly up and 1v1 is left in a broken state.

Thanks.

Imo Ostheer are used as benchmark it the other factions that are OP and combination of nefs to other factions and buff to Ostheer should be the aim.

Care to elaborate what the numbers in the table represent? Like, how did you gather them and what are they?


Ah, and

Care to elaborate what the numbers in the table represent? Like, how did you gather them and what are they?


Ah, and

I totaled the streak count as of today for top 10 players for each faction across 1v1, 2v2, 3v3, and 4v4. I figured the streak numbers is a better indicator of recent games played vs overall stats since I can't discern exactly when those games were played. I know its not precise but I do think it provides a good indicator in absence of having access to the time element of when games were actually played.

1 more thing, the totals weighted is a way of putting 3 allied factions on the same basis as 2 axis factions in computing a win streak ratio.

Thanks.

No Problema

I totaled the streak count as of today for top 10 players for each faction across 1v1, 2v2, 3v3, and 4v4. I figured the streak numbers is a better indicator of recent games played vs overall stats since I can't discern exactly when those games were played. I know its not precise but I do think it provides a good indicator in absence of having access to the time element of when games were actually played.

1 more thing, the totals weighted is a way of putting 3 allied factions on the same basis as 2 axis factions in computing a win streak ratio.

I honestly doubt that is even a decent indicator of faction balance. I (or any of the players you took stats from) could severely inflate those numbers (think 20+ streaks) by intentionally queuing during off-hours. Streaks in the top 10 are generally just an indicator of which players are noob-stomping the hardest.

Furthermore, look at the current UKF 1v1 top 10. UKF has a total streak of 99......prodi currently makes up 44/99 of that. A single fluke loss by him, and that value youre using gets lowered to 54 (around 45%). It's similar for most other factions, though not as severe.

Another word that I just thought to add, but if youre going to be doing these kinds of statistics, youre either going to need the randomness and sample sizes to ensure the results are relevant, or enough context to properly interpret and explain them (think tournament/streamed matches where the wins and losses can be analyzed and the players are known). I just don't think that measuring top 10 streaks has either of those.

I honestly doubt that is even a decent indicator of faction balance. I (or any of the players you took stats from) could severely inflate those numbers (think 20+ streaks) by intentionally queuing during off-hours. Streaks in the top 10 are generally just an indicator of which players are noob-stomping the hardest.

Furthermore, look at the current UKF 1v1 top 10. UKF has a total streak of 99......prodi currently makes up 44/99 of that. A single fluke loss by him, and that value youre using gets lowered to 54 (around 45%). It's similar for most other factions, though not as severe.

Another word that I just thought to add, but if youre going to be doing these kinds of statistics, youre either going to need the randomness and sample sizes to ensure the results are relevant, or enough context to properly interpret and explain them (think tournament/streamed matches where the wins and losses can be analyzed and the players are known). I just don't think that measuring top 10 streaks has either of those.

I think you missed the part where I said barometer ie it's an indicator maybe not perfect and yes could be influenced heavily if someone chose to do so, but I do think it does highlight the Ost regardless as being weaker than the rest.

One question for you (being a top 10 player), when the tournaments go to a rubber match where the faction is chosen by the vp count holder, does axis get picked the majority of times now?

Btw, I didn't post to pick holes in the flaws of the methodology. I wanted to see if this study aligned with the sentiment i am reading that says Ost is not any good and Jackson is OP.

I think you missed the part where I said barometer ie it's an indicator maybe not perfect and yes could be influenced heavily if someone chose to do so, but I do think it does highlight the Ost regardless as being weaker than the rest.

One question for you (being a top 10 player), when the tournaments go to a rubber match where the faction is chosen by the vp count holder, does axis get picked the majority of times now?

Btw, I didn't post to pick holes in the flaws of the methodology. I wanted to see if this study aligned with the sentiment i am reading that says Ost is not any good and Jackson is OP.

My issue is that its use as even a general indicator is probably being vastly overstated. If it highlights anything, its likely by happenstance. I guess the part of me that took one semester of stats (lol) is crying at how faulty all of the stats are. If people are to say that ostheer is struggling, I'd prefer it to be because of sound arguments and reliable statistics, not what I honestly consider to be the summation of completely arbitrary values.

Another issue is that your original post doesnt align with what you just posted. "Obvious factions to address," suggests there is a high degree of confidence in your results. "Good barometer" means you think its a good indicator and that the results are reliable and fairly accurate. "Completely broken" isnt the kind of statement you make based on unreliable statistics either. I just don't think the study itself supports these kinds of statements you made.

As to your question, I don't think its possible to get definite values from scc5 specifically since not all of the replays were posted. Also, in such tournament play, there are other factors to consider like what the map is. Spawn number 2 (axis spawn) is the favored spawn on most of the tournament maps.

The sentiment that ostheer is lacking goes back any number of patches. One of the original intents of the community patches were to address that sentiment. Although it was decided that this would be achieved indirectly by changing anything except how ostheer was lacking.

Its funny:

On the one side people say only wehrmacht is good now in 1vs1. OkW sucks.
And now you say the opposite.

In 2vs2-4vs4 you almost never see brits anymore, because you die vs indirect fire and mg play.

Numbers can be great and fine, but also deceptive!
Another point is, which elo is this? Is this only random vs random? premade vs premade etc.
Which maps? There are axis and allies maps etc

I honestly doubt that is even a decent indicator of faction balance. I (or any of the players you took stats from) could severely inflate those numbers (think 20+ streaks) by intentionally queuing during off-hours. Streaks in the top 10 are generally just an indicator of which players are noob-stomping the hardest.

Furthermore, look at the current UKF 1v1 top 10. UKF has a total streak of 99......prodi currently makes up 44/99 of that. A single fluke loss by him, and that value youre using gets lowered to 54 (around 45%). It's similar for most other factions, though not as severe.

Another word that I just thought to add, but if youre going to be doing these kinds of statistics, youre either going to need the randomness and sample sizes to ensure the results are relevant, or enough context to properly interpret and explain them (think tournament/streamed matches where the wins and losses can be analyzed and the players are known). I just don't think that measuring top 10 streaks has either of those.

Care to elaborate what the numbers in the table represent? Like, how did you gather them and what are they?

Ah, and

Since we won't have coh2charts anymore since Relic never ended fixing the API, is it possible to have the bot you were using to extract information about the current W/L ?

If anyone is really interested on getting any MEANINGFUL information, this guy above is the indicated to talk with.

Win streaks?

Huh, say what? Are you on drugs?

Could you look at each players past games and see the last 10 games they played and infer any data from those matchups? It will be bad data since some play 2v2 to warm up before automatching 1v1 etc.

Yes the tank destroyer overperforms yes ostheer faction is less forgiving then other factions of mistakes,

RIght now usf is hands down the weakest faction in 1v1 overall balance is pretty damn good in 1v1 in team games allies spam is very hard to beat

Stats system are same for all sports, football (dont try call it soccer), NBA, NHL, NFL, baseball.
In generally stats of wins/lose can say all and nothing, coz you must look at detail, matchup, map, side, buildorders, units, moves, RNG.
I like football, play it, watch it and here are full of players taht are good in one team and system, but are bad in another. In logic, if player score 20 goals he must be good ? yes, but if look in detail statics its can be mean nope.
Same matchup in coh2, ostheer vs brit, with and without sniper, will be 2 different games. Based on moves, out covers, mortar and maybe light vehicles. Only things that can unify them, its lmgs/gev43 for both side.

My 2 cents to the topic

In 2v2 perspective you don't see anymore USF/UKF combo. In all games there is atleast 1 soviet player on ally side. The recent balance patch really messed up the 2v2 scene. The situation switch 180 degree and allies are the "easier" side. Katiushas, guards, snipers, cons and t34s spam that's what you most likely to see.

Plus there is more OKW players on 2v2 than ostheerm which is wierd if we look that okw was accually nerfed the most in a DBP.

One says OKW is completely crap and need re-balance ASAP, the other says Ost is the weakest according to stats, the other states actually USF is by far the weakest in 1v1 now, then someone says you don't see Brits in 2v2-4v4 anymore.

All in all, I would come to the conclusion the game is in a pretty damn good spot right now, if the cry for weak/unbalanced factions and units is showing so much diversity these days.

Well to be fair, so far no one really said Soviets are under-performing at the moment lol

I guess, there is always gonna be someone complaining even if the game were to be (theoretically) perfectly balanced. Yet it is important to not fully ignore them and look into it, and I think the community is currently doing a fantastic job on that.

Win streaks?

Huh, say what? Are you on drugs?

+1

hope he wont edit original post. Most insane post of the year nominee. Nothing even remotely connected to any sense

Since we won't have coh2charts anymore since Relic never ended fixing the API, is it possible to have the bot you were using to extract information about the current W/L ?

Yes, but I'm actually not the biggest fan of these kind of stats because they don't actually measure what everybody intuitively thinks they measure, so they are bound to be misunderstood...

I think this is a very very poor way of measuring balance. I also find it strange you are using the word barometer. That word is specifically for atmospheric pressure measurements and doesn't fit this context at all. You can use the shorter word "meter" to represent your custom measurement better. Meter can be either the standard measurement everyone but the USA uses or it can simply represent a measurement of any kind. Finally, this win streak stat is highly flawed. It can be compromised entirely by luck. You are reaching in and trying to pull a constant out of a variable. The only time you can do this is when you take multiple samples over time. I really don't agree with using this particular measurement to help with a balance discussion. I think we can do better.