Developer Josh Noh failed to read the room, we failed to respond appropriately

When Overwatch developer Josh Noh chimed into the thread about the concern of “power creep” in Overwatch, (a term used to describe the overall increase in average power of all the heroes in the game over time due to the addition of new overpowered characters and the buffing of old ones to compensate), he likely didn’t anticipate that his comment would cause a shitstorm. Starting his comment with a “fun fact”, Noh fielded the winrates of Genji and Soldier:76 in Grandmaster ranks to argue against those two characters being power creeped by other members of the hero cast. The developer continued to explain that perceptions of hero strength are often shaped by the discourse in the community and become entrenched, not necessarily reflecting the actual strength of a hero in the game.

From the point of view of pro players and those who follow the competitive scene, the argument felt like a tone deaf slap in the face, considering that neither hero had been considered a viable meta pick in over a year and a half, with Soldier:76’s last flash of viability arguably going back to the summer of 2016.  The argument of their winrate at a GM level was especially torn to shreds on social media, citing the lack of usefulness of uncontextualized stats and the large skill differences even at that rank from professional play.

Humbling ourselves

Let’s practice some humility and call ourselves out on our weaknesses. The community surrounding the esport of Overwatch is often not willing to collectively explore outside the dominant meta without incentives, as our own expectations for proper play, as well as those enforced by people on our team, discourage playing outside it. Evidently proven by increasing hero diversity the lower the ranks you go, the space between the bars of those mental and virtual prisons become wider and wider as the lower you go in the ladder, the relative hero strength changes and meta familiarity fades.

Overwatch’s game balance of course isn’t strictly about winrates. They are arguably secondary indicators that a game is doing well. More insightful questions to ask for a game in which character picks can be mirrored is: How fun and satisfying is a hero to be played, played against and played alongside with on your team? A 100% pickrate for a hero everyone feels good playing with is not a problem. While Orisa has a higher pick and winrate, would you not rather play with, against or as Reinhardt?

That said, even if a hero scores well in acceptance on all these three questions, it still doesn’t mean that this must be true forever. We’re novelty seeking creatures with an appetite for exploration and adventure. Many of us hope for a state of the game in which the meta is ever evolving, either around maps or new revelations in gameplay induced by the occasional balance patch. In our own interest, we should keep those aspects in mind when approaching Noh’s comment.

We do not really expect Overwatch developers to walk into office each morning, sitting down at their work stations, sipping on their blonde smoked butterscotch latte, checking Overbuff and seeing that no hero has peaked over a 52% winrate, so all is right in the world, do we?

An attempt at a charitable reading

During my A-levels exams in biology a decade and a half ago, we were tasked to analyze and write about an ecological system of nine different animals in a specific habitat and consider all their interdependencies with four data sheets to consider. Over six hours, I wrote 24 hand-written pages on how I expected those animals to balance out and which other possible animals to introduce to balance the system and all my classmates were around that number of pages too. There is no way anyone could’ve made an argument in 200 words that wouldn’t have felt reductive. Noh’s comment is 209 words long and Overwatch balance is vastly more complex than nine types of critters in Western Congolese swamp forests. 

When looking at the comment from a more charitable point of view, Noh could have merely tried to find the most easily digestible argument to make his point. It doesn’t necessarily mean that he considers it to be the strongest or even a strong argument, but laying out all the vastly complex factors to consider when looking at hero balance goes beyond the scope of a forum thread and quickly approaches the size of a full-blown dissertation. It also doesn’t necessarily mean that he did not sensibly contextualize those numbers in their actual applications in games such as filtering out winrate favorable scenarios in which players largely pick those two heroes in scenarios where the game is already won for them. This is an uncharitable inference based on our perception of the quality of the product that influences our judgement on the quality of the analyzer without really understanding the scope of the task.

Grounds for criticism

It is perfectly reasonable to disagree with the conclusions that Noh reached and I personally also believe that power creep is happening in Overwatch to a noticeable and perhaps detrimental degree. However, it is crucial to realize that it’s impossible and unfair to claim to be able to spot the difference in reasoning with which we depart from Noh’s opinions simply by reading this forum post. The furthest we can go in our criticism is to point out that the chosen argument is a terrible one, especially in its framing. Moreover, Noh clearly didn’t read the virtual room in which he chose to comment, a space full of consumers starved of developer attention both in communication, content additions and game play changes, with one foot in the Overwatch 2 and the other in the Project A waiting room.

After all, developers ought to not merely manage provable game play imbalances based on data but more importantly should cater to the social and psychological realities that those balance states cause in their player base. As we’ve already established with even the best balance states eventually becoming stale, statements are never to be divorced of their environment and the collective state or mood the community is in.

On the side of the community, our outrage is exaggerated and misguided as our feelings towards Overwatch as whole bubble to the surface, latching onto an unfortunate and clumsy comment which was likely not more than an offhand remark while chewing on a sandwich in a five-minute break to continue with the crunch, only to be called up by HR for causing a shitstorm. Strategically, we’re not doing ourselves any favors in disincentivizing developers from commenting in our situation. I equally sympathize with pro players stating that their feedback in private channels fell on deaf ears, but a divide between the parties won’t make Overwatch a better game more quickly. We certainly aren’t surprising anyone in Team Four that there is a dissatisfaction with the state the game is in and attacking a developer’s competence serves no practical purpose but to pop our vents off by getting a cheap jab in.

If we’re honest, the essence of the events that transpired can be synthesized down to the community asking “Can we please play Genji and Soldier:76 again without feeling bad about it?” and Noh answering “It’s complicated, but for many of you that’s already the case.” 

Image courtesy of Robert Paul for Blizzard Entertainment