For all who have been asking...AoS Stats are BACK

[LLV]

 

enjoy!

 

[LLV]

https://thehonestwargamer.com/aos-stats-21st-october-2019/

 

 

[AaronWilson]

I am so glad this is back!

[prochuvi]

Slanesh broken and umbalanced......nothing new since those months

[Forrix]

Slaanesh isn't as crazy as I expected. I think FEC and DoK both got to almost 80% if I'm remembering right, balance is actually look pretty good if you disregard the less than 1% of the meta faction due to sample size.

[mmimzie]

23 minutes ago, Forrix said:

Slaanesh isn't as crazy as I expected. I think FEC and DoK both got to almost 80% if I'm remembering right, balance is actually look pretty good if you disregard the less than 1% of the meta faction due to sample size.

It's because one of thier major counters is skaven, with high numbers and shooting/magic. To the natural counter is already well represented. Similarly, with new cities we might be slannesh  having more issues. That said slannesh stop most multi wound and  melee armies from the word go. 

[WatcherintheWater]

@LLVThis is so awesome! Thanks!

I think there might be a sort error in the Match Win % column in the Meta% table. Looks like the Slaanesh win rate is showing for Stormcast for example.

[HollowHills]

Thanks for these, glad to have some more stats. 

[zilberfrid]

So, of all the CoS options, Freeguild performs best (after devoted, but devoted is now disintegrated). Since Devoted is in the list, that's pre-CoS Freeguild with great companies.

Getting 47% is actually nice.

I don't think they received a boost in CoS, with loss of great companies, indomitable trait and new musicians.

EDIT, also PT at 50/50

Edited October 23, 2019 by zilberfrid

[Inquisitorsz]

5 hours ago, Forrix said:

Slaanesh isn't as crazy as I expected. I think FEC and DoK both got to almost 80% if I'm remembering right, balance is actually look pretty good if you disregard the less than 1% of the meta faction due to sample size.

really? Statistically they are WAY above the curve. 67% win rate is massive. Ideal balance for a game should probably be 50%+/-5... 

I don't recall if those other ones hit 80 or not... but 67 is till very bad.  

Some other interesting bits....
Skaven and Stormcast are the most used armies.... but Skaven have a 10% higher win rate. 

Gotta be careful with stats because something being strong, makes it popular, and then that skews the results. The link above has some more stats and different lists. 
5 Rather large main armies are sitting below 45% win rate (Seraphon, Nighthaunt, Nurgle, Sylvaneth and BOC).
If everyone is at 50+/-5 that's not too bad.... but when your top few armies are sitting above 60% then disparity is huge. 

apart from a few outliers with a small sample size (LoN, Order Draconis, Devoted of Sigmar), the top end actually isn't that bad. DOK and FEC are still a bit high and Slaanesh obviously... 
But I think the bigger problem is the lower end of the scale. 
There will always be a few strong armies at the top, but I think GW should do more to pull up all those mid 40% armies closer to 50%. 

[l1censetochill]

47 minutes ago, Inquisitorsz said:

really? Statistically they are WAY above the curve. 67% win rate is massive. Ideal balance for a game should probably be 50%+/-5... 

I don't recall if those other ones hit 80 or not... but 67 is till very bad.  

Some other interesting bits....
Skaven and Stormcast are the most used armies.... but Skaven have a 10% higher win rate. 

Gotta be careful with stats because something being strong, makes it popular, and then that skews the results. The link above has some more stats and different lists. 
5 Rather large main armies are sitting below 45% win rate (Seraphon, Nighthaunt, Nurgle, Sylvaneth and BOC).
If everyone is at 50+/-5 that's not too bad.... but when your top few armies are sitting above 60% then disparity is huge. 

apart from a few outliers with a small sample size (LoN, Order Draconis, Devoted of Sigmar), the top end actually isn't that bad. DOK and FEC are still a bit high and Slaanesh obviously... 
But I think the bigger problem is the lower end of the scale. 
There will always be a few strong armies at the top, but I think GW should do more to pull up all those mid 40% armies closer to 50%. 

If you watch the video, LLV (the guy who did the collecting/compiling/analysis of the data) clarifies that if you remove the mirror matchups (which are extremely common at GTs in rounds 3-5 due to Swiss pairings) from the Slaanesh winrate data, the win % is actually between 75 and 80%. So the winrate against other factions is significantly higher. It's weird, but the fact that Slaanesh so rarely loses to anything in rounds 1 & 2 of a GT actually lowers its winrate, because you get an unusually disproportionate number of mirror matches later in the tournament.

Edited October 24, 2019 by l1censetochill

[Inquisitorsz]

2 minutes ago, l1censetochill said:

If you watch the video, LLV (the guy who did the collecting/compiling/analysis of the data) clarifies that if you remove the mirror matchups (which are extremely common at GTs in rounds 3-5 due to Swiss pairings) from the Slaanesh winrate data, the win % is actually between 75 and 80%. So the winrate against other factions is significantly higher.

Ah... that's interesting too.... I'd like to see the stat list with mirror matches removed. 
I haven't watched the video yet (at work). 

[l1censetochill]

7 minutes ago, Inquisitorsz said:

Ah... that's interesting too.... I'd like to see the stat list with mirror matches removed. 
I haven't watched the video yet (at work). 

I'd like to see that too! Because knowing that mirror matches are included in the data raises other questions, as well.

For example, Stormcast - 10% of the meta, 45% winrate. But in theory, if you remove the mirror matches (which would be somewhat common, logically, given their representation and the nature of Swiss pairings), the win rate would actually be lower, because every mirror match brings the faction win rate closer to 50%. Same thing would apply to Skaven, but in the other direction.

Edited October 24, 2019 by l1censetochill

[Mayple]

1 hour ago, Inquisitorsz said:

really? Statistically they are WAY above the curve. 67% win rate is massive. Ideal balance for a game should probably be 50%+/-5... 

I don't recall if those other ones hit 80 or not... but 67 is till very bad.  

Important to note that while saying that 50% winrate for general game balance is usually correct , it does not carry over to stuff like Age of Sigmar and similar types of games, simply on account of the nature of the game, and how (competitively) -good- players (and by extension, factions statistics) minimize their losses to rarely leave with a worse than 3/5 winrate, thus they'll -never- have a 50% since losing every second game you play is a sign that you're doing something very wrong  
A flat 50% winrate would therefore be an indicator of a faction in need of love, better players, or both.

60% (3/5) is the magic number to strive for, with room for variables; 55/65.

And that's ignoring the rabbit hole of connecting it to meta %, as more people playing a faction equals less skilled players dragging the percentage down. (That's how you get statistics like Skaven, who has a pretty top notch faction at the moment, sitting at a measly 55,9%) 
The opposite doesn't neccesarily apply (you can have a small meta % that consists entirely of sub-par players, but you can't have a large meta % that consists entirely of master-tier players) --- tl;dr: If the meta % is large, and the winratio is above 60%, you're likely looking at a faction that is crushing the competition simply by having better stats//mechanics. 

On a more speculative note; a lot of super solid players jumped ship to DoK back when they dropped, which pushed the winrates up to around 80/85%, and unless they decided to remain beyond the Slaanesh release, I'd expect a fair deal of them to be playing Slaanesh nowadays due to the similar playstyles/mindsets of the two armies. Could be worth checking. 

Edited October 24, 2019 by Mayple

[l1censetochill]

9 minutes ago, Mayple said:

Important to note that while saying that 50% winrate for general game balance is usually correct , it does not carry over to stuff like Age of Sigmar and similar types of games, simply on account of the nature of the game, and how (competitively) -good- players (and by extension, factions statistics) minimize their losses to rarely leave with a worse than 3/5 winrate, thus they'll -never- have a 50% since losing every second game you play is a sign that you're doing something very wrong  
A flat 50% winrate would therefore be an indicator of a faction in need of love, better players, or both.

60% (3/5) is the magic number to strive for, with room for variables; 55/65.

And that's ignoring the rabbit hole of connecting it to meta %, as more people playing a faction equals less skilled players dragging the percentage down. (That's how you get statistics like Skaven, who has a pretty top notch faction at the moment, sitting at a measly 55,9%) 
The opposite doesn't neccesarily apply (you can have a small meta % that consists entirely of sub-par players, but you can't have a large meta % that consists entirely of master-tier players) --- tl;dr: If the meta % is large, and the winratio is above 60%, you're likely looking at a faction that is crushing the competition simply by having better stats//mechanics. 

On a more speculative note; a lot of super solid players jumped ship to DoK back when they dropped, which pushed the winrates up to around 80/85%, and unless they decided to remain beyond the Slaanesh release, I'd expect a fair deal of them to be playing Slaanesh nowadays due to the similar playstyles/mindsets of the two armies. Could be worth checking. 

So, given Slaanesh's 67% winrate (stated in the video to be 75-80% when mirror matches are removed) at 6% representation, would you agree the stats are indicative of issues with the faction's mechanics? I'd actually be quite interested to see what the winrates would look like if Slaanesh dropped to 55% or so.

Your last point is actually brought up in the video as well, and I found it interesting - the meta % lost by DoK and LoN since late 2018 is almost exactly the % that was gained by Skaven and Slaanesh. So yes, a solid proportion of those players most interested in chasing the meta are now on Slaanesh.

[Mayple]

1 minute ago, l1censetochill said:

So, given Slaanesh's 67% winrate (stated in the video to be 75-80% when mirror matches are removed) at 6% representation, would you agree the stats are indicative of issues with the faction's mechanics? I'd actually be quite interested to see what the winrates would look like if Slaanesh dropped to 55% or so.

Your last point is actually brought up in the video as well, and I found it interesting - the meta % lost by DoK and LoN since late 2018 is almost exactly the % that was gained by Skaven and Slaanesh. So yes, a solid proportion of those players most interested in chasing the meta are now on Slaanesh.

I think they're definitely on the strong side, but not the end-of-all-things-good kind of broken OP stuff. I'd expect them to naturally mellow out as new factions and mechanics are introduced. To digress a bit, my money is on a blindsided City of Sigmar winrate surge, to be honest, as it received tools that deal very well with the current state of the game, and I'm sure people have caught on to that. I don't think it'll have the meta chasers running for it though (not their usual cup of tea, even if it brings cheese bigger than a skaven's wildest dream)

Slaanesh is just a daughters of khaine 2.0 as far as I'm concerned, and they (DoK) mostly got their wins through a combination of expert players, and the general jist that players in general have a tendency to get absolutely demolished by any fast moving army that can dish out glass-cannon types of damage (I think it comes down to poor positioning, since that doesn't get spoken about enough for a lot of people to pick it up) -- It helps that Slaanesh is less of a glass cannon than Daughters of Khaine, since they can do more or less the same stuff, but then also get to summon their daemons from doing the very thing their army is built to do. It is pretty elegant design, even if they might have overtuned it a bit (which I'm sure they'll get around to adjusting whenever the next opportunity arises) 

In less words; Yeah, they're strong. I think their mechanic is alright, but could do with a slight adjustment. If I understand correctly, they're currently farming those points by throwing Keepers of Secrets into things, right? If that's the case, that seems more of a Keeper of Secrets issue than an issue with their mechanic. Heard their new battalion was pretty far out though. 

Out of curiosity; What's the logic behind the mirror matchup percentage drop? Why would that matter for Slaanesh's statistics when it didn't matter for Daughters of Khaine (which had a larger meta % if I recall correctly, making it more prone to fight itself) -- Is it somehow a hard counter to its own army? 

[l1censetochill]

5 minutes ago, Mayple said:

Out of curiosity; What's the logic behind the mirror matchup percentage drop? Why would that matter for Slaanesh's statistics when it didn't matter for Daughters of Khaine (which had a larger meta % if I recall correctly, making it more prone to fight itself) -- Is it somehow a hard counter to its own army? 

They didn't get into it too much on the video itself, but the gist was simply that because Slaanesh has an inflated winrate against most every other faction, and because GTs typically match up armies of similar records, you get a lot of Slaanesh mirrors in rounds 3, 4, and 5 - and naturally, whenever you have a mirror match, one Slaanesh wins and one loses, thus lowering the faction winrate toward 50%.

As for why DoK didn't have the same issue, they didn't go into it and I'm not certain - they did mention that there were far fewer GTs to collect stats from last year (the number this year is double, over the same time period), so sample size and a low percentage of DoK mirrors in those particular tournaments might have something to do with it. It'd be a good question to ask if one were watching live, though!

[Inquisitorsz]

9 minutes ago, Mayple said:

Out of curiosity; What's the logic behind the mirror matchup percentage drop? Why would that matter for Slaanesh's statistics when it didn't matter for Daughters of Khaine (which had a larger meta % if I recall correctly, making it more prone to fight itself) -- Is it somehow a hard counter to its own army? 

It matters statistically. 
If an army is beating other factions at 80% and ends up at the top tables more often than not... it's going to face a lot of mirror matches. At least statistically more than other factions would. 
Therefore, you could say for example that if half of the Slaanesh games are mirror matches, then those wins also count as losses (for the other player), since there's always one winner and one loser. So half the Slaanesh game results wouldn't affect the win/loss percentage as both columns would get 1 result. 
So all that does is increase the games played by that faction, which dilutes the results pool and makes the overall winrate lower.   

If there was say a 3rd army that had win rates and popularity numbers similar to Slaanesh and Skaven, you would see more variety at the top table, thus less mirror matches, thus a different overall winrate.  

Of course, the overall win rate is the important one, mirror matches included... but you just have to bare in mind what assumptions and conditions account for those percentages. 
When we talk about balance and the strength of various armies, we almost always compare different factions, thus naturally ignoring mirror matches. 
If Slaanesh beats Slaanesh... that's a net effect of zero. 
If Slaanesh beats all other factions at a rate of 80%... that's a balance problem.  

In term of good players moving around, that certainly does have an impact. But you'll also so not so good players jumping on the latest "strong army" bandwagon and they can drag down the average as much as good players boost it. 
I think across 7000+ games and almost 100 Slaanesh tournament entries, that's enough data to statistically show how strong they are right now. 

[Mayple]

@Inquisitorsz

I understand that, but the crust of the question was why it would matter for Slaanesh if it didn't matter for Daughters of Khaine. 

 

If statistics tell us that Slaanesh sits at 67% -- and that same statistical source also told us that daughters of khaine sat at 80/85%, then we can't apply variables to Slaanesh to push it to 80% if those variables weren't also being used for DoK, unless the variable is unique to Slaanesh specifically. 

Am I wrong?

[Inquisitorsz]

@Mayple
You're right.... you'd have to ask the statistician. 
I'd say it's like because they changed the way they gathered/displayed the data since last time.  

We'd have to look at the previous data. It's likely what @l1censetochill said that the sample sizes are much bigger now which normalizes the results more.  
Or that Daughters was generally used less, which can inflate that win rate by reducing the chance of mirror matches (that's why there are 3 outlier factions above 60%). They had good results but so few games that it's not really statistically relevant.  That's why they sort the list by number of wins because that kind of also includes the meta% or popularity to an extent.  

If you look at the June 13 data on the website, there were only 3 factions above 60% - FEC (64%), DoK (64%) and Slaanesh (62%). 
FEC obviously above everyone else at the time. 
But DOK was above Slaanesh too... except both DOK and FEC had over double the number of matches that Slaanesh did. So you can potentially argue that those two factions had double the chance of hitting mirror matches at the top tables, which bring their winrate down. 

Also at that point Slaanesh had only been out for roughly  a month. I think they released start of May.

I can't see the data further back than that so check the the 80% win rates. 

It's possible simply, that when people talked about DOK win rates in the past, they were already removing the mirror matches. I dunno. We'd have to find that data and check. All I could find was some data talking about 70.9% which is worse but not that much above the 67.7% we see now with slaanesh. 

There's also other outside factors too.... If the GHB2019 strengthened some other armies compared to last year, that can bring the win rate of other stuff down across the board. Especially if the popularity and spread of armies has increased. 
For example Legion of Nagash fell from 10% popularity to 2%
The faction split in general was quite different with the previous data. 

If you look at this table comparing Dec 2018 with June 2019, you can see that prior to this latest data, DOK dropped 6.7% when FEC, Slaanesh and Skaven started winning things. Basically the newer power creep armies showed up. 
Ignore Phoenix table and greenskins.... they are outliers with very low data points. 
Legions of Nagash dropped 3%  in popularity and in win rates. It's dropped heaps more in the latest data too. 
(I think there might be an error in the new data too because Legion of Nagash and Grand Host of Nagash are listed separately... I think that's the same thing isn't it?)

In any case it might also be something like that.... Say for example Nagash was a favorable match up for DoK... So if less people are playing much less Nagash now, then DoK have a slightly harder time. Especially because LoN was the most popular army back then (13.1%) even more so than stormcast. 

If you look at just those 2.1 win rates, DOK were simply well above anyone else... the next highest being deepkin at 61%. Where as the current top 3 are much closer (with 2% of each other). 
But DOK popularity relative to other factions was lower.... like LoN was 13.1%.
Back then DOK were the 6th most popular... while LON and SCE made up for 25% combined. 
Now, Slaanesh is 5th (so quite similar)... but the gap is smaller because  the most popular SCE and Skaven are 10% and 9%. 

So I'd argue that it's more likely for Slaanesh to hit mirror matches today than it was for DOK to hit mirror matches back then. 
Of course... that's probably not the case in round 5 or 6 at the top table... but it will have an impact on round 3 and 4 match ups. 
that might be enough to explain that difference between the previous DOK 70.9% win rate and Slaanesh's current 67.7%. 




 

[LLV]

1 hour ago, Mayple said:

@Inquisitorsz

I understand that, but the crust of the question was why it would matter for Slaanesh if it didn't matter for Daughters of Khaine. 

 

If statistics tell us that Slaanesh sits at 67% -- and that same statistical source also told us that daughters of khaine sat at 80/85%, then we can't apply variables to Slaanesh to push it to 80% if those variables weren't also being used for DoK, unless the variable is unique to Slaanesh specifically. 

Am I wrong?

No not wrong.

few reasons-

DoK had a more apparent win %. Close to 80%. It was less of the meta though and whilst it likely did cannibalize itself somewhat it wouldn’t  have been as much. 

I have access to more detailed matchup data now than I did during DoK’s Reign.

it doesn’t matter more or less-DoK was just as powerful then as Slaanesh now. I was just responding to Rob being surprised it wasn’t higher. I was also doing it off the top my head which is why I said don’t quote me. But most Slaanesh players stay undefeated until G4 or 5 and a higher percentage than I’ve noticed in the past lose to themselves.

when I have the time I’ll run the matchup data, but it’s a hard task given the form I receive it in.

 

Edited October 24, 2019 by LLV

[Mayple]

@LLV

Fair enough That solves the confusion.

Edited October 24, 2019 by Mayple

[schwabbele]

So in the AOS 2.5 - GHB 2019 - Meta% Image with the table in it, SCE is on top with 67,7% Winrate... this is an error right? Or what that table tell me besides SCE most played army?

[zilberfrid]

6 hours ago, Mayple said:

Important to note that while saying that 50% winrate for general game balance is usually correct , it does not carry over to stuff like Age of Sigmar and similar types of games, simply on account of the nature of the game, and how (competitively) -good- players (and by extension, factions statistics) minimize their losses to rarely leave with a worse than 3/5 winrate, thus they'll -never- have a 50% since losing every second game you play is a sign that you're doing something very wrong  
A flat 50% winrate would therefore be an indicator of a faction in need of love, better players, or both.

60% (3/5) is the magic number to strive for, with room for variables; 55/65.

And that's ignoring the rabbit hole of connecting it to meta %, as more people playing a faction equals less skilled players dragging the percentage down. (That's how you get statistics like Skaven, who has a pretty top notch faction at the moment, sitting at a measly 55,9%) 
The opposite doesn't neccesarily apply (you can have a small meta % that consists entirely of sub-par players, but you can't have a large meta % that consists entirely of master-tier players) --- tl;dr: If the meta % is large, and the winratio is above 60%, you're likely looking at a faction that is crushing the competition simply by having better stats//mechanics. 

On a more speculative note; a lot of super solid players jumped ship to DoK back when they dropped, which pushed the winrates up to around 80/85%, and unless they decided to remain beyond the Slaanesh release, I'd expect a fair deal of them to be playing Slaanesh nowadays due to the similar playstyles/mindsets of the two armies. Could be worth checking. 

I don't get anything of what you mean here.

If you lose every other game (and win the other half of the games), you're not doing anything really wrong, you're just being as good as your opponents.

If 50% is indicative of something being really wrong, there is something wrong with the entire hobby.

[LLV]

 

2 hours ago, schwabbele said:

So in the AOS 2.5 - GHB 2019 - Meta% Image with the table in it, SCE is on top with 67,7% Winrate... this is an error right? Or what that table tell me besides SCE most played army?

It’s an error, that table meant to only show meta.