Each month, we take a look back at the month that was, looking at gaming data and trends from the past month to identify how 40k is changing and how it affects the way armies play on the table. As we head into November, we’re looking back at October 2020 and how the first codexes of 9th edition have shaped things.
In our monthly meta analysis for Warhammer 40k, we look at tournament and game results from around the world, identify key trends and topics, and look at how the meta is changing as new factions and rules are released.
The dust never truly settled on the pre-codex 9th edition meta, but after two and a half months of games it’s pretty clear how things had shaken out with regard to faction tiers and some aspects of 9th edition play. Then three weeks ago new Codexes for Space Marines and Necrons dropped, shaking things up and starting us down the long cycle of new releases that is guaranteed to keep things fresh.
So this month we’re taking one last look at the meta that was, some broad 9th edition trends, and a look at early results for the two newest Codexes. We’ve also got some new data this time around, and we’re diving a bit deeper into some things people have asked us about after we’d published the prior articles. So get comfortable and let’s dig into the best Meta Analysis yet.
Also we’d like to extend a very special thanks to Peter “The Falcon” Colosimo from 40kstats.com – a wonderful site that we constantly refer to and reference for our own purposes – for helping with the analysis on this month’s Meta Analysis.
Methodology
This is it. Thanks to the wonderful efforts of tournament organizers and app developers around the world, we now have access to what is essentially every meaningful piece of data around competitive games of 40k, aside from notes scribbled on 6-10 lines of a notepad in someone’s house.
The data in this month’s study comes from three sources:
- The ITC Battles App, a wonderful app for tracking games both in and out of tournaments. In our latest data pull, this gives us data on more than 13 thousand games of 9th edition Warhammer 40,000, with 2,662 of those played since the new Codexes were released on October 10th. Additionally, the app now includes data on which player won the roll-off to go first, for deployment, and subfaction data. If you haven’t used the app, it’s the best way out there to track your practice games and it’s great for tracking games at events.
- Best Coast Pairings/Down Under Pairings – The wonderful folks at BCP/DUP make the app for collecting and reporting tournament results for ITC events. The BCP/DUP dataset includes data from more than 10,500 games played at events, includes subfaction data, and now tracks game stats through its game scorecard. If you haven’t used the app, it’s worth checking out, and subscribers can access events data and lists for any event run through the app. We couldn’t do our weekly competitive innovations article without this app.
- Tournament Organizers. We’ve had great support from tournament organizers around the world, who’ve helped us out by sending us results for events where the data isn’t being collected in other systems.
So That’s it, right?
Well, not quite. There are still a few areas where our datasets aren’t as robust as we’d like, and it’d be dishonest not to mention those. Here are three areas where we need more data, and what you need to know about them:
- The subfaction datasets are still small for some analyses. We’ve got some great preliminary data from ITC Battles App and BCP, but until we get a larger dataset – something we’ll likely have next month – the specifics of which factions perform better going second should be considered directional. We think they’re still interesting enough to look at, but without more data, there’s a chance that factors like individual player skill are impacting the results too much.
- Results by round are still sparse. We’ve got more data now on tournament results by round, but unfortunately that data set is still pretty small for round 4 through 6 as the number of events and players is still small due to precautions being taken as a result of COVID. We expect this to change in the coming months, but until it does, our understanding of how go first win rates change in later rounds or at top tables is going to be limited based on the smaller number of games we have data for.
- Terrain data. Quantifying terrain and its impact is difficult, and we don’t really have much data on this. We’ve done some work so far looking at table layouts for major events and while we’ve seen some variation in layouts, there doesn’t appear to be enough differentiating terrain layouts at many events to suggest there’s a major impact there. At the same time, we know not having adequate terrain can warp the game, so this is an area where we’d love to have more data.
The First Turn
We felt like we had put this particular problem to bed last time but even after our last analysis we still had people shouting about anecdotal results and some high-level players indicating a certain preference for going second. So, with even more data at our disposal, we thought we’d take another pass at this analysis and look at first turn win rates one more time. Before we dive into the stats, let’s talk about why we’re seeing these effects.
Why Does the First Turn Matter So Much?
By James “One_Wing” Grover
Something we’ve been asked a few times since we’ve started doing these analyses and looking at go-first win rates is why having the first turn results in these disparities.
Why Does the First Turn Generate an Advantage?
There are two main reasons that going first provides an advantage in 9th Edition’s missions – the advantage of getting to use your ranged threats to attack the opposing army first, and objective dynamics. Of these, the latter seems to have the substantially larger impact, and can be broken down into a few sub-factors as we’ll see in a moment.
First Strike Advantage
Let’s get first strike advantage out of the way first. Pure shooting lists are a lot less viable in 9th than 8th, but thus far there still seems to be a strong incentive to include at least one powerful mid-range shooting unit in most lists, the more “general purpose” the better. The best current example, as of the new Marine book, is Plasma Inceptors. These have high-volume, high strength firepower with good damage and AP that also benefit from Blast. This means that they can be pointed at almost anything and be expected to do substantial damage. Thanks to their mobility, with the first turn they can often be brought to bear against a premium target and do substantial damage to it. That gives the player going second the option of either hiding their units, which might reduce their early utility, or starting the game down one of their better units. This is especially true because the second player won’t have yet had the chance to activate any defensive abilities used in the Command Phase, such as the Chronomancer’s Chronometron.
The player going first also gets the first chance to bring in reinforcements, which if used to deploy powerful shooting units can again allow them to get the first bite at doing serious damage, in this case in a fashion almost impossible to hide from.
The good news in this section, and why it isn’t the main culprit when considering first turn advantage is that in many situations, player two gets significant compensation as long as there’s enough terrain for them to hide behind. Thanks to objective dynamics, 9th favours lists with far more melee and short ranged threats than 8th, and it can often be easier for player two to bring these to bear in volume on their first turn, as player one generally needs to move up to start taking objectives out the gate. In addition, while getting the first opportunity to bring on reinforcements is an advantage, player two gets the last chance, which can balance this out. If both players put units into reserves that happen to be fantastic at obliterating each other, player one has to put theirs down first, opening it up to a counterattack. If reserves are being used for harassment/objective stealing units (which they often are), player two can also wait so they know where all of player one’s have gone before putting their last units down.
With good terrain, therefore, these factors mostly balance out. There are still armies that swing heavily when they get the first turn consistently, notably AdMech and shooty Custodes. Since not all events run with great terrain, and those two armies are both quite good, this will still have some impact on the numbers we’ve seen, but is perhaps less significant than you’d think if you’re still picturing 8th Edition ITC armies.
Objective Dynamics
The much larger and more 9th Edition-specific factor that affects first turn balance is the dynamics around the scoring of primary objectives.
The overwhelming majority of GT2020 missions have a mixture of objectives in players’ deployment zones (“home objectives”) and objectives in the mid-board, with The Scouring being the only outlier, substituting home objectives for having two that are nearer to each player’s deployment zone (which doesn’t massively change the dynamics for many armies). Players are generally compelled to spend turn one ensuring they hold a competitive number of mid-table objectives, and can start to push for their opponents’ home objectives from turn two. Scoring happens at the start of players turns 2 through 5, and to maximise your primary score you can afford no more than one turn on which you score only 5VP, and as soon as you have scored 5VP for a turn you know you must score 15VP on at least two later turns.
This generates three broad dynamics that favour player one on the primary scoring:
- Player one can dictate what player two needs to achieve early game.
- Player one’s home objectives are much less likely to be threatened before their first chance to score them.
- Player one gets the “last touch” on primary objective scoring.
By moving on to multiple mid-board objectives on their first turn, player one puts player two under immediate pressure. Unless they counterattack sufficiently to ensure player one holds fewer than half the objectives, they accept:
- Player one gets 15VP for their first turn, giving them the breathing room to “absorb” a 5VP turn later on.
- They don’t get 15VP for their first turn, keeping them under scoring pressure into the mid game.
This is compounded by the second factor. Many armies cannot realistically threaten their opponent’s home objectives on their first turn. If you are going second, that means that however many home objectives your opponent has are definitely going into their scoring pool on their first turn, but the same isn’t true for yours. With multiple turns of movement and the opportunity to bring in reinforcements, most armies can attack their opponent’s home objectives on their second turn, meaning that player two doesn’t get any turns of safely racking up points from their home field. This combines with the above to mean that it is far from uncommon for the game to start with player one scoring 15VP to player two’s 5VP, putting player two under immense pressure to score perfectly from there on out.
Unfortunately for them that runs into the last factor – the last touch advantage. Because scoring is at the start of the turn, the last opportunity either player gets to shape how many primary points are scored is player one’s fifth turn – nothing that happens during player two’s fifth will affect the primary scores. That frees player one to go absolutely all-out on denying points to player two on their final turn, and they’ll often have a clear idea of what their own final score is going to be while doing so, giving them near-perfect information along with their freedom to make exceptionally aggressive plays.
Taken together, these mean that there is a tendency for player one to have an easier time building up a lead early on and an easier time defending a lead once they have it. The ease of building up an early lead, in my opinion, explains one of the starkest numbers from our dataset last time around. The number in question is the difference in average VP scored by the losing player depending on whether they went first or second. While the VP total for the winning player was roughly consistent whether they went first or second, players who went first then went on to lose scored on average 5VP more than players who lost after going second. That suggests that in games that go horribly wrong, the first turn gives you an easier ride at building up points early on, even if your opponent later overtakes you.
This is all consistent with what seems to be emerging as the best mission from a first turn balance perspective – Retrieval Mission. This is unusual in having six objectives, but only one home objective for each player. That means that player one doesn’t get as much of an advantage in how safe their objectives are from an early push, and the map also means that player two is quite likely to end their first turn with two “layers” of objectives held, making it harder for player one to fully knock them off. Combining these factors means player two probably has the best chance of holding level in the early game in this mission, and that’s likely why it comes out looking more balanced.
Does the New Data Back this Up?
This comes together to form what I think is a pretty convincing narrative of why this matters, but it would be even better if we could back this up with some data. Happily, we can! I mentioned above that this narrative was substantially shaped by the datapoint from our previous analysis showing that go-first losers scored higher on the primary than go-second losers. Thanks to the exceptional granularity that the ITC battles dataset provides us, we’ve been able to delve into the exact dynamics of this, and the results convincingly back this narrative.
What this chart shows is the average number of primary VP scored in each battle round by each player in each of the four possible combinations of which player went first and which player won. It demonstrates two key things:
- Players that go first and win tend to immediately build up just under a 5VP lead in an average game.
- Conversely, players that go second and win only build up a fractional lead on average in the first turn.
The main cause of the difference seems to be how many points the player who goes first scores in their first turn. That early disparity in the point patterns is then maintained all the way through the game – of the average 5.1 additional primary VP a go-first loser scores in a game, 3.8 of them are present at the end of that first scoring window in battle round 2. The player going first simply has an easier time scoring early, however the game plays out.
As for why that is, the data gives us insight into that as well. Once we’d noted that the disparity seemed to be due to player one scoring well early whoever won, but player two only scoring strongly in that turn in games they went on to win, we dug deeper into the scoring patterns of the second battle round.
There are your numbers folks! Players that go first and win are locking in 15VP turns over a third of the time, and managing to deny their opponents any victory points at all nearly a quarter of the time. Players that go second and win do nowhere near as well at achieving either of those goals, and only marginally edge out the eventual loser in any of the brackets.
For my money, this leads me to believe that of the mechanics that drive first turn advantage, the relative safety of player 1’s home objectives prior to the first scoring opportunity is by far the most probable culprit. In nearly a quarter of games that player 1 wins they’re able to do so by blitzing player 2 out of scoring any points at all out of the gate, while when the roles are reversed it happens under 10% of the time, and alongside that player 1 is much more frequently able to create a 15VP first turn. 9th games are short and brutal, and coming back from a situation where you start out 15-0 down with your opponent in your face is an incredible uphill struggle. Fundamentally, the reason going second seems to consistently result in losing more is that it massively increases the chances that you end up in that position.
What Doesn’t the Data Show?
The other thing worth pulling out from the same charts above is that player 1 doesn’t seem to be getting that much advantage from the “last touch” on scoring that going first gives you. Point changes for winners and losers from rounds 4 to 5 don’t seem to massively vary based on whether they went first or second. This is a very important takeaway if we’re considering any potential fixes or improvements to the missions, as a common trend in proposals in this area has been to give player 2 some kind of improved scoring opportunity in the final turn. Having done this deep dive, I’m vastly less convinced that this would actually help, and think that any interventions or mission changes need to be focused on giving player 2 more ways to attack player 1’s positions out of the gate, as the relative safety of player 1’s early scoring appears to be our driving factor.
What Can Players Do About It?
Going first appears to provide a clear advantage, but it isn’t an insurmountable one, and there’s lots of talk about building “9th Edition Lists” to counter the trends seen. To a degree, there’s something there. You can heavily mitigate the impact of going second on your chances of winning by packing lists that can do the following:
- Hide from or soak up any first turn firepower.
- Counterattack strongly into any early advance by your opponent.
- Threaten home objectives on the first turn even when you go second – perhaps the most important given what we’ve seen above!
The army that does all of this best right now is Harlequins, and it’s likely a major factor in why they’re performing so strongly. All of their units are fast and have small profiles, allowing them to hide behind Obscuring terrain (or use Shadowseer abilities to shut down ranged fire) and still strike the mid-board turn one, and Skyweavers, one of their best units, have the mobility to attack a home objective straight away. Slaanesh Daemons are another powerful option, as their large Seeker units can strike straight for the heart of the enemy with the rest of the force close behind. I do also feel that there’s some indication that 9th Edition codex design will provide more options for this – I’m definitely much more comfortable going second with new Necrons than I currently am with Craftworlds.
All together, you can definitely build “9th Edition Lists” and improve your chances of winning going second, but I don’t think it’s a complete solution by itself for two reasons:
- How well you can build to these goals varies wildly by faction, and the divide between the “haves” and “have-nots” of 9th seems to be starker than ever.
- Even the best go-second faction, Harlequins, still performs narrowly better going first.
As more 9th Codexes release, it will be interesting to see if all factions are being decked out with the tools they need to adapt to going second – but right now, I’m happy that some marginal boon for the second player is something that’s needed.
First Turn Win Rates
By Robert “TheChirurgeon” Jones
Now that we’ve established why the first turn matters and what the impact is on a per-game basis, let’s talk about what that means for 9th edition as a whole and give some additional context to what we’ve talked about here.
How Big a Deal is 5 Points, Anyways?
Well it turns out, that average swing of 5-6 Primary victory points for the player going first vs. second is a pretty big deal in aggregate. As we’ve seen before, the net result is that the player who takes the first turn is significantly more likely to win the game. In our ITC Battles data, the player taking the first turn won the game about 58% of the time. This is statistically significant at a better than 99% confidence level though as we’ll see, there are some other things that can affect this. We’ll also address some common questions and counterarguments we’ve seen to this data suggesting that it isn’t an issue.
How has this changed over time?
One thing we heard early on when we published our September analysis was that players at that time were unused to the mechanics and game structure of 9th edition, and over time would improve at going second. If we look at go first win rates by month from August through October, we can see that this is not the case – the rate at which players going first win their games has been remarkably stable, suggesting that even if players have gotten better at going second, they’ve also gotten better at going first too.
Are good players winning more going second?
Another rebuttal we’ve seen is that some top players – such as Nick Nanavati – often choose to go second, and are winning more games that way. This argument suggests that at high levels of play, going second isn’t a disadvantage, and may even be an advantage. So is this true?
Well, yes and no. The answer is much more complicated. As we mentioned before, some factions do better at going second than others. When we looked at our ITC Battles app data last month, we noted this was the case, but also that no faction seemed to do better going second than going first. That’s still true for our ITC Battles App dataset, but the gap narrowed substantially for the game’s top factions over the last month. In particular, Harlequins and Chaos Daemons both improved their go-second win rates substantially. We also have enough data to look at Slaanesh armies as a faction, and they likewise make a strong showing.
If we narrow our results to just look at tournament games from BCP and DUP, we can see that this edge is even more severe, and that here Chaos Daemons and Harlequins actually have higher go-second win rates, though they maintain their high go-first win rates as well. On the other hand, outside of Daemons, Harlequins, and big-tent Chaos (which includes Daemons), no other faction in BCP has a better-than 50% win rate going second, and no other faction has a higher go-second win rate than a go-first win rate.
It should be noted at this point that these are small samples and the differences in go-first and go-second win rates here should be considered directional, as they aren’t statistically significant at high confidence levels (with the exception of space marines, though there’s not nearly enough data in BCP or ITCBA to show chapter-specific results yet). It’s also worth noting that aside from these three factions, go-second win rates really aren’t much different at BCP tournaments than they are in our larger ITC Battles app dataset.
Instead, what we’re left with is a picture that shows specific factions that win a lot of games are also better at going second, and it may be the case that good players playing them are also doing well in games where they are forced to go second. But because these factions also just win a large majority of their games and are considered “top-tier” factions, they’re also likely to attract top players. The net result is one where it’s just as likely that top players are playing the factions that do better going second (and also first, mind you) than it is that top players are just enjoying some intangible benefit that going second offers.
And indeed, when playing Harlequins, it just appears not to matter.
What about the roll-off?
Another question we’ve been asked is how the roll-off for first turn affects this. If both players want to go first or both players want to go second, this can be a more meaningful factor.
Using data collected in the ITC Battles app, we can now look at the result of the roll-off for first turn, deployment, and the decision to deploy first, and look at the impact these have. Let’s start by looking at the potential scenarios and how often they come up:
So here we can see that a player winning the roll-off and choosing to go second happens in a little over 15% of games (those middle two conditions), and in nearly half of games (49.7%), the winner wins the roll-off and chooses to take the first turn. These results are in-line with what we’d expect from our go-first win rates analysis and, true to form, win rates for players who win the roll-off are essentially the same as win rates for going first.
There is a bit more nuance in here, however and that’s where that 15.7% of games where a player chooses to go second come in. While a player who wins the roll-off chooses to go first 84.3% of the time, suggesting that in 71% of games both players want to go first, scenarios where both players want to go second (roughly 2.5% of games) will also affect this. So let’s break out the conditional probability formulas and look at win rates in the four scenarios that can result from this die roll and decision:
Players who win the roll-off and choose to go first win 59% of the time, followed by players who lose the roll-off and are “forced” to go first. We can certainly see there’s a large bump here for players who choose to go second – they win 48.3% of their games after winning the roll-off, but players who lose the roll-off and are forced to go second average out to only a 41% win rate. This also suggests that if two players both want to go first, then there’s something like an 18-point swing in win probability if they lose the roll-off, while if two players want to go second, the player that loses that roll-off may actually be in a better position.
The lower win rates for players choosing to go second suggests that even if going second is advantageous for some armies or players in some situations, most players are probably bad at evaluating those situations, and mis-judging their armies, the situation when they arrive at the table, or both.
How does the mission affect this?
We’ve previously done some work looking at first turn win rates by mission to identify what factors, if any, impacted first turn win rates. Our analysis last time turned up some interesting results, showing Battle Lines as the mission with the lowest first turn win rates and Vital Intelligence as the mission with the highest. This was an area where we expected additional data to change things and now with another month of data and a new tournament-heavy dataset, it’s time to revisit those results.
These results aren’t that dissimilar to last time, though there are two things to note:
- None of these are statistically significantly different from the average go-first win rate.
- Vital Intelligence and Battle Lines have both moved, with the former becoming less go-first heavy and the latter become more biased.
Seeing some movement is interesting, and caused me to wonder if we should look at month-by-month results to see if some missions were trending up or down, and if that movement was significant. Turns out, there’s something to this; over the last month several key missions have shifted:
Specifically, in October since our last analysis, Vital Intelligence shifted significantly downward along with Retrieval Mission, and Battle Line shifted significantly upward. This falls somewhat in line with the results we see from higher level play collected in October from our smaller BCP/DUP dataset, where Vital Intelligence is one of the lower go-first win rate missions and Retrieval Mission is the best in both.
This makes it a little bit tougher to figure out how missions are really impacting go-first win rates; in our prior analysis, we did a regression analysis that suggested that the number of objective markers was the primary factor on go-first win rates. Instead we’re left potentially looking at missions that have a clear deployment zone objective marker looking like those with a lower go-first win rate. There may be something to mission secondaries being a culprit – as we’ve seen in our prior analysis, mission secondaries – particularly Priority Targets, Vital Intelligence’s Data Intercept, Sweep and Clear’s Direct Assault, and Battle Lines’ Vital Ground all generate greater VP returns than many other secondary options, though both players are able to choose them.
This is an area where we’re just going to need more data and there may be other factors at play. It’s also likely to impact the results we see in our events data.
Is this a bigger problem for higher level play?
By James “Boon” Kelling
Given that our data spans the spectrum of play from casual to competitive, the question that many want to know is how these numbers change when you focus solely on “top-level” play. And while there’s no true definition to what constitutes top-level play, generally speaking we expect that any given GT or Major will attract a wider range of talent that includes top talent – and in the few GTs and Majors to date we have indeed seen a number of top players turn out. So have the previously identified trends held up in these events?
The short answer is “yes.” In 12 events across the world, including Gibraltar, New Zealand, Australia, the United Kingdom, and the United States – the go-first win rate in over 1,300 games of top competitive play has exceeded 57%, which is well in-line with all of our other datasets.
What’s more interesting is the idea we had previously theorized – that Go-First Win rates (GFWR) increase as Swiss-paired events go on. The logic goes that as early round player mismatches occur, more skilled players may overcome a go-second disadvantage and drive a more narrow GFW percentage in the early going, but we’d revert to the mean as like-skilled players play. While we have over 1,300 games to review, it’s simply not enough data at a 5% confidence level to draw any significant conclusions on any given round and its deviation from our observed go-first advantage. However, the non-significant trend does support such a hypothesis (we note that we would conclude statistical significance at a lesser 10% confidence) and believe that with another month of data we’ll be able to draw more firm conclusions. It is unlikely that we’ll be able to draw conclusions on the later rounds until we have either many more months of data, or a greater velocity of results. What we can say is that round 1 results are statistically significant against all remaining rounds at the 95% confidence level.
Below you can see these effects, charted in the BCP/DUP data, and below with data we’ve collected from TOs directly where first turn was not tracked in BCP.
Assuming the continued data collection bears out our hypothesis – the implication is even stronger evidence in favor of a structural disadvantage to going second within 9th edition. Essentially, players who understand the advantages inherent, understand how to counter it, and can outplay their opponent may buck the trend – but that disappears immediately once you begin to match the ensuing rounds as the skill gap closes and both players understand the advantages or disadvantages inherent.
Using This Information
So now that we’ve presented all these findings, what can you, the player, do with this information? Well, the first thing is you can make smarter decisions about when to go first and when to go second when it’s your choice. Most of the time you’ll still want to go first, and that’s OK. More importantly…
The Game Is Not Actually Over If You Lose the Roll-Off, You Big Baby
Rob: While a 58% win rate for players taking the first turn is bad and skewed and is definitely something we’d like to see fixed in future missions, it’s also not the end of the world, either. If you find yourself being forced to go second, the game is not over; it’s just going to be harder.
Boon: And that’s the key – it’s going to be just one thing that makes the game harder. But ultimately this is just one possible advantage or disadvantage in a wider system of advantages and disadvantages. A player mismatch, an army mismatch, a bad deployment, bad terrain – it all provides advantages and disadvantages and sometimes they’ll stack in your favor, balance, or stack against you. Sometimes they’ll stack so heavily against you that it’s nearly impossible to win and sometimes you’ll be able to make mistake after mistake and it won’t matter because the variables stacked in your favor.
The point is that, while the roll-off is important, and generally factions do not win more going second than they do going first, it’s not game over. You can still prepare for second, know your army, know your opponents army, and out general them on the tabletop. And that matters – because ultimately you cannot control the roll off, the terrain, or your dice rolls but you can control whether you’re in the best possible position to suffer those disadvantages and more often than not still come away with a win when you shouldn’t. And frankly, that’s what makes a top player.
Understanding the Meta
OK so let’s go back to the current meta. Given that the weekend of the release of Codex: Space Marines and Codex: Necrons was pretty sparse, that really gives us about 14 days – two full weeks – to look at the impacts of these new books and how the field has changed. Let’s start with the overall win rates by faction, using games of 9th from August through October 26:
Mostly, things haven’t shifted since last month. We have Slaanesh armies making a strong showing now, thanks to another month worth of games data, and we expect that it’s legit owing to the fact that many successful daemons armies rely on Slaaneshi daemons. Otherwise, the top armies remain Daemons, Harlequins, Sisters, Space Wolves, Orks, and uh, Chaos Knights. More on them in a moment. Let’s talk about the major trends.
Necrons are back, baby
In news that shouldn’t be a shock, Necron usage is way up since the new book dropped. 524 games with Necrons were recorded in the ITC Battles app after the new books went on sale; this was more than any other faction save Space Marines (786), for whom a new book was also released. And in other news, the new book is a significant improvement over the last one, since there was really no way to go but up: Over this time, Necrons have won 54.4% of their games, and that’s a massive jump from the 37.7% win percentage they were posting before the new Codex. In our sample, new Necrons sport a 67% go-first win rate and a 45% go-second win rate, though these numbers will likely even out over time.
The new marine codex shakes things up
Rob: On the other hand, there were few ways to go but down for the new Codex: Space Marines and it appears like they’ve dropped a little across the board, thanks to major nerfs to Salamanders, one of the key factions that were absolutely tearing things up prior to the new book. If we combine BCP/DUP data from before the new books released with ITC Battles app data from games played afterward, we can build the following picture. Note that these results are based on a smaller number of games played and should not be considered stable.
That’s a lot of movement! Of these, only the changes to win rates for Space Wolves, White Scars, and Salamanders are statistically significant at a 95% confidence level, but even those may change over the next few weeks as we see hundreds of games being played. So it’s time to ask: What’s real here and what’s an illusion?
Well, here are my thoughts:
Real
- The Dark Angels bump is absolutely real. The faction is significantly stronger than it was, even with the loss of warlord traits and stratagems. Several Dark Angels units got huge boosts and there are many intriguing ways to build the faction now.
- The Space Wolves drop is also real (and more stable), but with a caveat: Space Wolves legitimately got worse with the loss of their Sagas, discipline, and stratagems and they should rebound when their new codex releases. Given that’s in two weeks, this drop likely doesn’t matter anyways.
- Imperial Fists dropping like a stone seems real as well. There’s almost no upside to the army over other chapters.
- Blood Angels staying about the same seems relatively stable as well given the number of games. They’ll likely improve in December with their new Codex as well.
- Raven Guard were already struggling and the new book is even less kind to them. A 4-point drop in win rate seems accurate, even with the small number of games. I won’t be shocked if this sticks.
Illusory
- Salamanders and White Scars dropping by an average 15 points each seems drastic. While we may not see them return to their pre-codex 60% win rates, they’ve got some nasty tricks in the new book and I expect them to end up over 50%.
- Deathwatch are absolutely going to rise with the new Index and their new Codex releasing in two weeks. This week’s preview showed a number of really strong rules for them and it’s unlikely those won’t combine with their access to new units to make for a stronger faction.
- Ultramarines being a top marine faction seems like it’s not real, and their bump isn’t statistically significant at the 95% confidence level to begin with. They have some things going for them, but don’t seem to do anything that other chapters can’t do better.
Chaos Knights?
Rob: Chaos Knights are surprising here – they have high win rates both here and in BCP/DUP data. But this is where the Falcon’s/40kstats’ Tournaments in Winning Position (TiWP) stat comes in handy – while Chaos Knights have a high win rate, they tend to lose hard in round 2 of events, meaning they’re rarely in a position to win an event. And so we have yet to see a Chaos Knights list reach the podium at a GT or Major in 9th.
Harlequins
Rob: Harlequins continue to make strong overall showings as one of the game’s best factions, with one of the highest win rates of any faction in our combined BCP/ITC Battles app sample (56%) and second only to Salamanders in 40kstats’ Tournaments in Winning Position (TiWP) metric and TiWP ratio. We and others have discussed Harlequins and why they’re so good in this edition previously so we’re not going to cover that again here.
Daemons
Daemons are the other faction suddenly tearing things up, despite a quiet start to 9th edition. Despite what look like huge vulnerabilities to the Assassinate and Abhor the Witch secondaries, the faction is actually thriving, and posting some of the game’s highest win rates, both in the ITC Battles app dataset and the BCP/DUP data.
Boon: The first thing to note about Daemons is that even with another month past, there hasn’t been a lot of players bringing forward Daemon lists, which makes it hard to say anything conclusively. The second thing I’d note about daemons is that there are a lot of ways that they can be brought – pure god lists, mixed god lists, and primarily daemon lists which masquerade as general Chaos (such as with the addition of Death Guard – seriously have you seen Chaos Spawn? They’re absurd!). It’s the latter that I think is perhaps the most potent, but let’s set that aside for a moment.
Daemons are capable of doing a number of things that top-performing factions do – rapidly grabbing objectives early, forcing opponents off contested objectives, shifting weight effectively, and in some cases, bringing extreme resilience. I think the most effective daemon lists generally choose a few aspects to focus on and then go all-in, and Slaanesh really exemplifies this. They (Slaanesh) play very similar to Harlequins but replace fusion pistols with potent big-bads that can deliver smites as well as a beating in close combat, surrounded by extremely fast beast and infantry units. The combination daemon armies may feature some nearly unkillable exalted greater daemons with a mixed Lord of Change and Great Unclean One with the right gifts and relics – as a core they form one extremely intimidating and hard to shift objective sitters and psychic defense.
But perhaps something that is most benefiting both Daemons and Harlequins is the fact that the meta itself is simply not geared to tackle them as opponents. They are minority factions in a marines-dominated world and the weapons you’re bringing to kill marines are largely wasted on these guys. That’s a big advantage to both factions.
Second turn scoring is the key to winning when you’re second (and also first).
Rob: As Wings mentioned earlier on, something that Harlequins and Daemons share, along with several other top-tier factions, is the ability to score primary VP early, by threatening objectives in the opponent’s deployment zone. While Orks tend to do significantly better going first in our data, they’re also capable of scoring early and threatening opposing objectives with Da Jump. The net result is that we can see several key armies – Daemons chief among them – consistently average nearly as many Round 2 VP going second as they do going first, and this is something that top factions share, though it’s worth noting that no faction averages more round 2 primary VP going second than going first.
On the flip side, we can see that while second-round scoring for factions going first is pretty competitive, with the lowest average being about 7.5 points and most factions averaging 8+, going second introduces a much, much larger spread – 8 factions currently average fewer than 7 round 2 primary victory points, showing us just who the haves and have-nots are when it comes to making up the difference when they’re going second.
The Meta is pretty healthy, actually
Rob: We’re in an interesting place. While marines are the dominant faction in terms of the number of armies in the field, they’re not nearly the powerhouses they once were, and the preponderance of marine targets and armies opens things up for minority factions that do well in the 9th edition mission structure. The end result is a meta where there’s a decent variety of winning lists and competitive factions, even if some are hamstrung badly.
If you want a deeper look at the lists that are winning at events each week. I strongly recommend you check out One_Wing’s Competitive Innovations in 9th column, which explores winning lists from events each week and identifies why they work and what makes them interesting. Check out last week’s article here, looking at new marine lists.
Next Month: Space Wolves, Deathwatch, and More Data
That wraps up our look at the meta this month. Stay tuned next month when we’ll have even more data to really flesh out and put some of these concepts to bed, plus we’ll have another dataset that just recently became available.
Finally, a huge thanks again to Joshua at ITC Battles app, the wonderful folks at Best Coast Pairings/Down Under Paritings, and the Falcon for helping put this together and really make this the most thorough analysis yet. Y’all rock and we always have a blast jamming these out.
In the meantime, keep making those lists and if you have any questions or comments, drop us a note in the comments below or email us at contact@goonhammer.com.