Disclaimer: This is a personal and obviously biased point of view. I am addicted to GW products and Lore. But I am convinced that there are game design issues, a recurring pattern of “failing by design” approach in 40k rules, almost independently from the editions concerned. When did it start? For my when 7^th Ed hit the ground… Feel free to comment, to add arguments in order to make me change my mind (you will see that this is a starting point of this Blog entry – how I (partially) changed my mind. If you feel this contribution is not constructive, feel free to ignore it. If you think you can launch a constructive discussion and chat, whatever you agree or not with my point, you are more than welcomed sharing it. But always remember: these are my views, influenced by my own background and my amount of time invested in the hobby which is a very individual trajectory that may have nothing to do with your own experience.

Last week the WTC 2023 championship has been celebrated in Belgium. By its size it is probably one of the major events on this side of the pond, being comparable, for example, with the number of attendees of the tourney within Adepticon.

While not being very hooked to tournaments, neither in terms of participating to tourneys (I almost did not play (any serious 40k game) over the past decade), nor in terms of following live streams, I could not avoid having, this time, a look at the published results of the WTC Warmaster 2023, i.e. the individual tournament. I wanted to know if I would be comforted in my preconceptions, largely fed by others statements red on various discussion forums.

WTC Warmaster Tournament was won by Aeldari. As expected, one would say. 3oo4 top positions were Aeldari. An interesting quote from a live event commentator was (translation by me): “Aeldari victory looks logical considering the current state of the meta….”. It made me fall from my chair. My conception was that the meta resulting from the results of events and not a founding state that could be used explain the distribution of the results. The egg and the hem issue, or kinda. Then I gave it a though (or too), and I realized my misconception: of course, the meta is predating the results of the tournaments! I am able repeating in chats, one and again, that 40k is fore anything else, driven by statistics. And as such it is in essence the interactions between sets of rules access that defines the meta as the distribution of armies and units that gives access a priori to the highest umbral of probabilities to get successful dice rolls.  How could I have missed that?

As Tournament players are competitors, aligning themselves on purpose of competing, and winning, most should have led to this conclusion already. Earlier than I did (by “earlier” it should be understood “years ago”). And as their decisions are driven by this analysis, they should individually reach the same conclusions, and this will lead to some overrepresentation of the factions that should allow leading to better results. In WTC Warmaster 2023, 29 factions were enlisted, but 80% of the lists (thank you Mr Pareto) were centered over 10 factions. Aeldari dominated with 41 lists (42 if the oddball Asuryani is added), but the combined Imperial+Chaos Knights would not be too far with a cumulated number of lists of 38 (25+13). Custodes, 1000 Sons and GSC are quite present which is however not always so obvious in the final results. With exception of Adeptus Astartes/Vanilla Marines, all the players in the top 10 had lists taken from these 10 factions. But 50% of the top ten were Aeldari which is still a big bias. Interestingly, no Knights of any flavour in the top 10...

So, instead of shouting out about how Aeldari are biased (I will not, I have an Craftworld army too), I had a look at the statistics of the winner. And it goes out that 80% of his games were against lists from the top factions that cover 80% of the lists. Not surprising you will tell. But we shall remember that it means he faced essentially Aeldari, Custodes, 1000 Sons and Knigths. Let´s say overwise, and in a VERY simplified way:  Aeldari won the game against other Aeldari (or equivalent). So, if benefits and advantages of faction selection compensates, the winner won because he has been the best at analyzing the most efficient faction but even more because he played better against people that shared its analysis. As would have said Maverick (or Rooster), “it is not the plane; it is the pilot”. So, congratulations to the winners, as they really deserve it. No matter what the meta is.

But is it prone to 10^th Ed.? It is of course easy to claim that the new rules and the indexes are biased, unequilibrated and ill designed (statement that I, at my limited scale, contributed banalizing by sharing it)… By doing the same shortcut analysis, but this time with the previous WTC 2002, I obtained the same results. So no, it is not only 10^th Ed. Of course, one year ago, the factions were different, the balance dataslate too, but the same kind of Pareto was obviously found. The WTC 2022 winner fielded Necrons, the second most represented faction at the time. Again, 80% of its opponents were in the 80 percent of most represented lists. Top ten was distributed among the 80% of the most represented lists, … So the logic behind list selection and distribution looks like being respeted. But, this time, none was showing an overweight such as the one of Aeldari in 2023.

So, where does of all of this heading?

Well, I feel like it is not an issue of 10^th Ed., neither an issue of Aeldari in 10^th Ed.. Or at least it is not more of an issue than it used to be with Sisters and Necrons under 9^th Ed. at the time of WTC 2022. Of course, I propose a very narrow viewed analysis based on 2 similar tournaments only over a 1 year gap. In between, factions have been redistributed because rules have changed. But a striking effect for me is that, eventually with exception of 1000 Sons, almost everything has been redistributed, in depth. Today´s champion factions have few things in common with yesterday´s, leading to a complete reshuffling of the cards. The reboot has a powerful impact, and this is an issue. It is an issue because it means that, after 10 iterations of 40k rule revisions, no (or very few) lessons learnt have been retained. The newest version of the game is probably as mature as if it was a fresh new crowd funded game hitting the shelves for first time (Am I too harsh?).

But back to the WTC comparative, as some common points emerge between the 2 events: some factions were underrepresented in 2022 and keep on being poorly represented in 2023. Space marines. AdMech. Astra Militarum. Orks. And this is not good for the game.

Why is it bad? Why did/do they behave (too) poorly? First, they do not win because they are not fielded (close to the french national lottery slogan: “100% of winners have played”). And second, if they are not fielded it is because, based on my initial assumption that competitors are clever and analytical, these factions do not present a high enough probability to win. It could be because of the unit types/rules they have, which makes them not apt for facing the main/most represented factions (AdMech?). Or because their rules are too much of an archetype and too “niche” (Astra Militarum ?).

But it is also affecting at Space Marines: the faction that has the largest choice of units. And this is an issue for the game.

Indeed, if a tool exists in any other faction, then SM probably have it too, or at least something quite close. So what´s wrong in that case?  Good question. I do not have any definitive answer, but the risk that rules have been designed out of the "main frame" comes to my mind. By “main frame” I mean for example rules designed to balance factions sorted out by volume of sales and not by their distribution of probability to win against an uncommon reference/standard. Imagine for example that SM have been designed to allow a great balance at playing against others SM (extreme case is HH) but that the rest of factions has not been adequately taken into consideration and/or removed from the equation... Under such an hypothesis (eventualy caricatural), if a faction´s sales is representing a limited percentage, you do not design game balance on the oddity (pointy ears), but on the majority (SM). It may in that case lead to rule sets that are contextualized for casual games (50% players get SM) and not adapted to the competitive scene. In that case Mr Pareto 80% for casual games would be made of SM, CSM and other most recent poster boyz co-staring factions (Necrons) probably... And we are on the other side of the mirror.

Which leads to a conundrum: trying to adjust a type of game based on data coming from a different way of using the rule sets. Data sourec can be argued (and often/always is) because of their supposed relevancy (or lack of) in a distinct gaming context (“socializing” vs. “competing” contexts for example). Of course, rules can always be adjusted based on Tourneys´ results, but one thing coming out from here makes me think that it is still not so good for the game.

Not so good, because the way the metawatch is communicated makes me feel that the Win ratio window is not an appropriate indicator to make decisions. Of course, it is important but, is it representative of the whole? If in the end, my point being that the winner wins against factions that are as much overrepresented as its own faction, the victory is fore anything else the result of a gamer that played better, within a restricted operational window (size of the window being driven by the meta). The statistical approach to the whole list of factions may lead to serious secondary bias too:  how can you “fix” poor or limited yield of armies that are underrepresented, or even not played, while pretending acting with data (backward regulation) that do not exist or are lowly representative because of insufficient volume (for example: "Is the only player that fielded AdMech in WTC 2023 not well ranked because of its faction´s rules or because of how he played that day?") ?

The only purpose of dataslate should end up being limited to try to find a way to equalize the distribution of factions enlisted. A way to go, step by step, to a new state where the % of each faction is close or egal to the over ones in registered participants (but we stop focussing on the victory rate which becomes a matter of gamer skill only). But it is not really satisfying as it is a long work that will leave the bad impression of levering from the bottom while playing on margin adjustments. And margin adjustment works only if everybody is walking close to others. Now some Factions looks like they looks like being very far away from that close gap.

Dataslate as main adjustment method still indirectly points out to the key factor, the root cause of the results we see: most of the laddering of results from tournaments of some factions vs. others is due to a genetic flaw of the game. It is a defect ab inicio. It comes directly from the factions and unit rules´ bodies. So the root cause is in my opinion to be looked for into rules and special rules that are not adapted to the needs and requirements of the competitive scene. It is not that the rules are intrinsically bad. It is that they are not done for all the purposes of their uses. Trying to fix a balance dataslate by patching points and some aspects of units rules looks therefore like being more of a Sisyphean works in that case. It may allow some adjustments at margin but at expenses of potencially unbalancing a fragile equilibrium if it goes too far. And it does not guarantee that the final result will really be better and more stable than the starting point. It becomes an iterative process by necessity. At the same moment, permanent changes and adjustments that eventually invalidate units may only lead to upsetting people especially if they feel like being affected by external events (the competitive scene) they just do not care about or that do not correspond to the game the want to play and for which they invested their money into. So it is a difficult exercise that should be always be warmly welcomed but that will ultimatly always frustrate and leaves people angry or dissatisfied. Including sometimes myself.

But is there any alternative solution? (except disconnecting from the internet and stopping downloading dataslate)

• On the short term, there is one: the whole rules should stay free of charge. A kind of compensation. Better than nothing. Grade of realism? I let you judge.
• On the medium term, it would be easy to claim for  a revision of Codex design methodology with much more play testing. But considering the development times, it is probably already too late for the codices to be released in the next 6 to 12 months. And anyway it is probably an illusion as it might be physical impossible due to the size of the factions. Yet should it be feasible, it might lead to a distortion into the edition life cycle that will need to be addressed in the mid of the road. Without saying that it might lead to release of codices radically different from the index too. A difference that will need to be managed as it will add another layer of frustration. So, probably that there is no mid term option.
• On the long term, I personally see huge room for digitalization of the game design and numerical testing, but it is already too late for this edition. While listening to some episode of The Painting Phase Vlog, stating that GW hired chemist to talk to chemists while designing its paint ranges, maybe it is time to repeat this winning strategy and hire a couple of Mathematicians and Probability experts to solve it out before launching on the market the next big product. A way to avoid the pitfall of a product bearing inside its own causes for weak/lowly resilient design? Calculation potential of modern computing is not a barrier. Number of datasheet and rules neither, as in the end it deals with statistics issued from profiles + special rules. And special rules affect at profiles through statistical enhancments. It remains a multivariable problem, but number of variables remains somehow limited – think about it: a desktop computer can solve combustion reaction kinetics with more than 800 independent variables while our 40k datasheet barely have 3 (probability to hit, wound and save). The real key is how to define what cost is worth for each associated probability… But how GW runs its business is not my area of expertise… And I should be advised not pretending giving lessons on the topic.

Yet, the balance dataslate based on tourneys´ results remains something that is a MUST HAVE, like it or not, and whatever its results might be. And it has nothing to do with lack, or poor, or weak design. Think about it: can one (a game editor) pretend not doing nothing while the hype (and its future sales) is so hardly affected by influencers, fast communication, and its impacts on the success/failure of a product? From there comes, I guess, the absolute need of the dataslate. It should allow managing the impact of the tournament scene, as a communication vector. Results published must be adequately dealt with and answered, answered though what looks like essentially to a damage mitigation strategy. Because if issuing a dataslate based on tourneys´ result is not IMHO so good for the game, not doing it would be worse, potentialy leading to potentially disastrous issues, such as player base reduction and/or Factions commercial axing...

So, in the end, it leads me to revise my preconceived statements about balance dataslate: it is not bad to try to fix the game using results from tournaments, not at all. But it is ultimately useless unless it buys you time to treat the root causes. Here is the main hue of mine. I´d wish that by setting up a more radical design process revision in parallel to the “fire fighter strategy” of “fixing stuff on the march”, GW could make the real next difference: giving access to a great and stable core game in addition to the already great miniatures and good lore of them. The 2 latter are the reasons I am still into the hobby since the early 90´s. The former is the main reason why I did not play a serious 40k game over the past 10 years… But being after 10th Ed at the almost very same point as at each cycle makes me reasonably sceptic about the chances that it might happen.