Warcom Metawatch

[techsoldaten]

The latest Warcom metawatch is out.

 

https://www.warhammer-community.com/2022/12/01/warhammer-40000-metawatch-2022-in-review-12-codexes-4-dataslates-2-mission-packs-and-one-new-army/

 

This one left me scratching my head. The following chart presents the win rate at tournaments since the last dataslate update. 

 

Spoiler

 

What precisely are players supposed to infer from this chart?

 

If I understand the data correctly, this is a visualization of raw win rates for individual factions in tournaments. The majority of factions have a win rate between 45% and 55%. People thirsting for balance in the game could take this as a sign of progress.

 

That perspective might be misleading. The underlying data does not reflect actual representation of each faction, nor does it reflect placement of the wins in a tournament. This could meaningfully skew any conclusions about relative balance in a few ways.

 

- Sample Size: DG may have appeared in 100 competitive games. LoV may have appeared in 7. The larger sample size of DG may be a more accurate reflection of their relative strength, while the smaller sample size of LoV might be a less accurate indicator simply because there haven't been enough competitive games against them. This could apply to a lot of other situations. 

 

- Placement: Not all wins are the same. As army lists reflect the underlying strength of a faction, an army that goes 5-0 at tournaments can be considered stronger than one that goes 4-1. Wins from a list that places highly in a tournament should matter more than wins from a list that mostly loses.

 

GW deserves some credit for getting 10th edition to the point where no faction is dominating all others. But charts like this are indicators, they only answer a narrow question. We should be very clear about what we are supposed to draw from it.

 

Anyone know if the underlying data is accessible? I'd really like to look at it myself.

[Cactus]

From previous articles I understand the purpose of that chart to be showing which factions have a win rate over 55% and need reining in, and which are below 45% and in need of a boost. The goal is everybody's win rate being between 45% and 55%.

[Rik Lightstar]

1 hour ago, techsoldaten said:

The latest Warcom metawatch is out.

 

https://www.warhammer-community.com/2022/12/01/warhammer-40000-metawatch-2022-in-review-12-codexes-4-dataslates-2-mission-packs-and-one-new-army/

 

This one left me scratching my head. The following chart presents the win rate at tournaments since the last dataslate update. 

 

  Hide contents

 

What precisely are players supposed to infer from this chart?

 

If I understand the data correctly, this is a visualization of raw win rates for individual factions in tournaments. The majority of factions have a win rate between 45% and 55%. People thirsting for balance in the game could take this as a sign of progress.

 

That perspective might be misleading. The underlying data does not reflect actual representation of each faction, nor does it reflect placement of the wins in a tournament. This could meaningfully skew any conclusions about relative balance in a few ways.

 

- Sample Size: DG may have appeared in 100 competitive games. LoV may have appeared in 7. The larger sample size of DG may be a more accurate reflection of their relative strength, while the smaller sample size of LoV might be a less accurate indicator simply because there haven't been enough competitive games against them. This could apply to a lot of other situations. 

 

- Placement: Not all wins are the same. As army lists reflect the underlying strength of a faction, an army that goes 5-0 at tournaments can be considered stronger than one that goes 4-1. Wins from a list that places highly in a tournament should matter more than wins from a list that mostly loses.

 

GW deserves some credit for getting 10th edition to the point where no faction is dominating all others. But charts like this are indicators, they only answer a narrow question. We should be very clear about what we are supposed to draw from it.

 

Anyone know if the underlying data is accessible? I'd really like to look at it myself.

 

They explicitly call out your concerns about the data set sample sizes within the first few minutes of the video.

 

The charts are purely a "top level" view of the current state of the meta. The video goes into some depth on those deeper questions.

 

For example there's some discussion about the "Command Point Economy" and "Secondary Missions" in the Nephilim Mission Pack Season and how these things have impacted the choices players make. Not unsurprisingly as there will be a new Mission Pack soon.

 

Rik

[Karhedron]

Will mission packs survive the (assumed) arrival of 10th edition? Will 10th even have missions or will those live in the supplements now?

[techsoldaten]

2 hours ago, Rik Lightstar said:

 

They explicitly call out your concerns about the data set sample sizes within the first few minutes of the video.

 

The charts are purely a "top level" view of the current state of the meta. The video goes into some depth on those deeper questions.

 

For example there's some discussion about the "Command Point Economy" and "Secondary Missions" in the Nephilim Mission Pack Season and how these things have impacted the choices players make. Not unsurprisingly as there will be a new Mission Pack soon.

 

Rik

 

Thanks for the answer. I didn't watch the video, that's nice to know they are conscious of what the chart represents.

 

I still wonder if there are better metrics are possible. Like weighted averages based on placement and frequency.

 

Just listing the total frequency to the right on the existing graph might be a good indicator.

[Rik Lightstar]

29 minutes ago, techsoldaten said:

 

Thanks for the answer. I didn't watch the video, that's nice to know they are conscious of what the chart represents.

 

I still wonder if there are better metrics are possible. Like weighted averages based on placement and frequency.

 

Just listing the total frequency to the right on the existing graph might be a good indicator.

 

The pretty strong inference from the video is that they're aware of the other metrics and use them for the balance work they're doing.

 

Stu mentions for example that while Harlequins have had very good win rates they're 19th out of 23 factions as far as representation goes and acknowledges that unless you're playing on the Top Table(s) then you're unlikely to play against them at most tournaments. You're far more likely to be playing against Space Marines.

 

They also discuss in general terms what to expect from the next balance update, in short BIG improvements for Marines and AdMech and Flamers of Tzeentch should expect to be nerfed right into the warp and out the other side.

 

Rik

[Emperor Ming]

9 hours ago, Karhedron said:

Will mission packs survive the (assumed) arrival of 10th edition? Will 10th even have missions or will those live in the supplements now?

Thats why 10th will prob be June, so the warzone book cycle doesn't not get interrupted

 

 

[BLACK BLΠFLY]

I have lost all confidence in the geedub development team for 40K.

[Jorin Helm-splitter]

On 12/2/2022 at 7:27 AM, techsoldaten said:

The latest Warcom metawatch is out.

 

.....Snip....

 

Anyone know if the underlying data is accessible? I'd really like to look at it myself.

 

I'm pretty sure they're using the 40stats data on Goonhammers site. Not really aware of another data set that is out there, and they don't have their own app to collect it.

 

On 12/2/2022 at 10:09 AM, Karhedron said:

Will mission packs survive the (assumed) arrival of 10th edition? Will 10th even have missions or will those live in the supplements now?

 

I'm pretty sure they will continue to have them. Hopefully the missions change a bit more in the next edition because the minor changes they made made the product hard to justify for me. The worst part is if they made a good app for keeping score of a mission's primaries and secondaries that was updated on a regular basis I'd subscribe fairly quickly, and I don't play as much as I used too (lot happier with BattleTech)

 

On 12/2/2022 at 8:03 PM, BLACK BLΠFLY said:

I have lost all confidence in the geedub development team for 40K.

 

I actually don't mind the development team, what I hate is that they are also the balance team. There aren't enough of them to really make a large enough sample to actually know how things fit in. Then you factor in that realistically they are working on books for 10th edition at this point. There should be a balance team that takes a deeper dive into what units aren't being used, what units are used too much, how armies are doing, and evaluating if unique codex specific secondaries are a good thing. Then they just adjust the books, to make things work.

Edited December 4, 2022 by Jorin Helm-splitter
left out a word

[Lord_Ikka]

6 hours ago, Jorin Helm-splitter said:

 The worst part is if they made a good app for keeping score of a mission's primaries and secondaries that was updated on a regular basis I'd subscribe fairly quickly, and I don't play as much as I used too (lot happier with BattleTech)

Use Tabletop Battles- it's done by Goonhammer and they have done an excellent job with it. Great at keeping track of both primaries and secondaries.

 

[jarms48]

I really doubt Imperial Guard have a 40% yearly winrate and a 45% current winrate. 

 

Just look at Meta Monday guy on Reddit. Imperial Guard currently have a 38% winrate over the past 6 weeks. Before the dataslate that added AoC to Imperial Guard vehicles the winrate was bouncing around the high 20's to low 30's. 

 

So they absolutely couldn't be in the 40% range. That implies there were periods they were well above these values bring up the average. There's some creative accounting going on here. 

Edited December 5, 2022 by jarms48

[Jorin Helm-splitter]

7 hours ago, jarms48 said:

I really doubt Imperial Guard have a 40% yearly winrate and a 45% current winrate. 

 

Just look at Meta Monday guy on Reddit. Imperial Guard currently have a 38% winrate over the past 6 weeks. Before the dataslate that added AoC to Imperial Guard vehicles the winrate was bouncing around the high 20's to low 30's. 

 

So they absolutely couldn't be in the 40% range. That implies there were periods they were well above these values bring up the average. There's some creative accounting going on here. 

 

The current win rate table was just the data from the US opens.

 

For the yearly data I could see a big spike in games with guard after the dataslate. That could affect the yearly data quite a bit. They also could be counting non event games there as well.

[Rik Lightstar]

I don't think it's just based on the US Opens.

 

I guess the table they presented might be, but they say in the video that they have over 330,000 game results logged that they're drawing the data from. So that's way more than the US Opens.

 

Rik

[WrathOfTheLion]

Either way, the win rate for Imperial Guard is wholly meaningless before the general release of the new Codex.

[Jorin Helm-splitter]

5 hours ago, Rik Lightstar said:

I don't think it's just based on the US Opens.

 

I guess the table they presented might be, but they say in the video that they have over 330,000 game results logged that they're drawing the data from. So that's way more than the US Opens.

 

Rik

For the first table the warhammer community article said it was the results of the 2022 US open series. I don't think they mentioned it in the video. 

 

The US open tournaments use pods (which I like) but it can skew win rates. If you go 1-3 the first day. Then win your 3 games in that pod it's a bit different than getting paired up as you win. Full disclosure I didn't attend a US open so those numbers may be off.

 

The second table was a yearly one. I'm not sure where they got the data for it it's a bigger set than 40k stats. The Point I was trying to make about that data is that if a bunch of players gave a Guard shot right after tge balance update and decided it wasn't enough that might skew the data because the number of people playing Guard isn't consistent.

[tychobi]

I think that us players are supposed to infer that the balance team is trying.  They have numbers, they have feelings, they have rules that will change.  Getting the arcane art of game design right using blunt math seems ok for those of us who want balance but it seems more likely to me that these numbers are to display that there is someone paying enough attention to put some fuzzy numbers from somewhere on a graph in an effort to explain future changes.  

[Orange Knight]

My biggest issue with "Metawatch" and the entire facade that GW are trying to present, (that warhammer 40k is somehow a legitimate competitive game) is that their entire sales model works contrary to this message.

 

They release rules piecemeal throughout an edition. Factions and sub factions can be left in the dirt for years before a new codex suddenly turns their fortunes around. The fact that this happens often is indicative of the fact that GW recognise the lack of balance, and are also fully aware of rules that can shift the balance of play, and that they hold these rules back until the optimum time to sell models according to their schedule.

 

I completely understand that models have to be released apart from each other, and that you can't overwhelm the customer base with an avalanche of releases, never mind the logistical issues with trying to shift 60-100 new kits in one go. It simply isn't feasible.

 

But the rules? The codex books?

The current format isn't conducive to a competitive system. I see no reason why all factions can't exist in a living encyclopedia that updates the entire roster simultaneously, and if need be it even includes the datasheets for models that are yet to release. This would even allow players to mentally analyse and prepare for releases ahead of schedule, and to test them with theorycraft ahead of time. The book could exist digitally, with the option for players to purchase the sections that are relevant to their own factions, updated annually.

 

Of course, this could potentially lead to the loss of luster in all things shiny and new. The rules create as much hype as the models themselves in certain circles. 

 

And this brings me back around fully. GW shouldn't present this as a competitive game whilst their model exists to bolster excitement and FOMO purchases in the short term ahead of everything else.