Blog: EyeSports: Using Eye Tracking Metrics in Competitive Gaming

eSports as a domain of competition is heating up and growing fast in terms of dollars invested, fans courted, and teams established. And as the core of eSports continues to expand, so too do the peripheral verticals around it. One of the most interesting of these secondary areas is the world of eSport player health; a critical feature that has largely been ignored. Today, many of the most successful competitive teams have begun to bring healthcare in-house, with some of the larger teams having just completed the process of setting their houses up in the first place. Team SoloMid (TSM) is in the process of building a 20,000 sq/ft facility in Los Angeles, while the Chinese team OMG has already taken up residence in their “Starship” HQ. Building state of the art facilities is one thing, but they need to be jam packed with eminently qualified staff and state of the art technology to be truly impactful. Many teams raising capital in the past two years have begun to include earmarks for full time trainers and performance staff members. Additionally, third party providers such as fitness coaches, nutritionists, and sleep specialists are being brought in to ensure that the players maintain a good physical standing, so as to boost their digital standing.

The marriage of eye tracking to eSports is a natural one. Already, many players elect to display their eye movements during live streams on Twitch TV and YouTube by using consumer grade attachments to their computers. While there aren’t metrics around these traces, audiences enjoy them as they add an additional layer of clarity in each player’s unconscious micro decision-making workflow. A glance at a minimap or a saccade to a newly revealed enemy speaks volumes.

The applicability of eye tracking metrics to eSports are manifold. While assessing for oculomotor impairment stemming from an in-game injury is certainly an unlikely happenstance (save for the rogue flying mouse or frustrated keyboard head-slam), there are still plentiful use cases that make adoption of the technology worthwhile. Just like their physical counterparts, eSports teams confront many challenges in assessing new recruits. While in-game metrics like kill to death ratio (KDR), assists, and objectives offer a degree of insight into overarching player performance, there are no functional measures of player performance to date. Additionally, all of these in-game metrics are retrospective in that they define past performance. While this may be indicative of possible future performance, the prospective value of eye tracking is unparalleled in its predictive ability of near-future performance by a player. Further, by screening new recruits for oculomotor performance, teams are able to glean a much deeper understanding of the player’s dynamic visual attention, a central component to a cast of sports where the eyes are the main conduit of engagement. In the same way that eye tracking can be utilized in initial screening before on-boarding a new player, it can also be utilized in the internal pipeline many teams have established for matriculation of academy players to the core team. This is a very similar pathway to baseball with AAA and the majors, or basketball with G-League and the NBA. Finally, eye tracking performance can also be factored into game-day decision making processes, wherein performance degradation due to fatigue can be quantified, and players can be substituted upon their teammate exceeding a given threshold. While this is certainly not applicable across all competitive formats, many events such as the League of Legends Championships do have built in mechanisms for player substitution.

In the exploratory research conducted so far, the eye tracking outcomes of eSports players, specifically those playing League of Legends, has been unanticipated. While elite soldiers and elite athletes normally have stellar dynamic visual orientation abilities due to their top tier human performance in extreme circumstances, and though hypothetically this should carry over to elite performers in the digital realm, the reality of the situation is markedly different. Rather than all being God-tier, eSports players tend to fall into three sub-groups. The first sub-group, the high performers, are on par with elite performers in the physical realm, due to their quick and precise eye movements made during gameplay. On the other end of the spectrum, are players who have dazzlingly poor eye tracking, the origin of which is unclear. This cohort warrants further investigation and is most likely to benefit the most dramatically from dynamic visual orientation training. Finally, the third cluster are those players with clear and dramatic sleep deprivation signals. This group would benefit the most from active monitoring of their impaired ability to attend and orient to the world, but should also use eye tracking to evaluate the benefits gleaned from external correction efforts, whether that be new mattresses from a sponsorship, new monitoring tools such as Whoop bands, or general sleep hygiene intervention.

One of the most outstanding findings of the exploratory research conducted so far has been the revelation of Position Specific Eye Movements in League of Legends players. Across the tested teams, it is apparent that the relative performance across the positions (Top, Mid, Bot, ADC, Jungler) remain consistent; that is to say that junglers have consistently the worst performance while mid and bot laners have consistently the highest performance. While at face value this relative degradation in performance may seem like a burden, it most likely stems from compensatory adaptation from the requisite actions required by each position. More work needs to go into understanding this and the data set needs to be expanded, but there is tremendous potential here for teams to characterize players and set their role with objective information, not just subjective personal preference or opinion.

Looking to the future, eye tracking will also impact eSports outside the vantage of direct player performance. The industry is currently confronted with the imposing obstacle of player drug use and abuse, in both recreational and performance enhancing modes. Stimulants, depressants, and cannabis all clearly impact eye tracking and screening against these will be an important evolution of the eSports regulatory landscape.

While so far there has been clear evidence of Position Specific Eye Movement in LoL players, as the data set continues to grow and technology continues to produce more granular outputs, it is probable that there will also be strayation in eye tracking performance across the different eSports. This is to say that twitch shooters like CS-GO will probably produce overall better performers than their LOL counterparts, who in turn will be higher performers than their Hearthstone playing brethren. This performant hierarchy would be mimetic to the world of physical sports where baseball players dominate in dynamic visual performance due to the tiny size and ridiculous velocity of their object of import. Following baseball, the other quick small-ball athletes from sports like lacrosse and tennis follow suit, with the athletes playing with larger and slower objects like basketball and soccer trailing behind. Additionally, as VR eSports grow in popularity and VR native eSports finish gestating, the importance of vestibular screening will come to the fore. Here, there will certainly be a vestibular performance rift between the static screen gamers and dynamic VR players. Accelerating adoption and supplying metrics specific to each eSports community will forever alter the way these games are played, and will help push eSports strategy to be more rigorous and scientific in its off-court approaches.