It’s the Ping, Stupid
Much has been made of North America’s consistent regional failure in League of Legends. Pundits, analysts, casters, pros, and fans wring their hands and wonder what the endemic issue is. Why do North American teams get consistently blown out of the water by statistically weaker teams at international competitions?
Arguments have been made, scapegoats announced and crucified, but the entire question simply boils down to a harsh truth: North America will never win worlds as long as professional players practice on 70 ping.
To a layman, the difference between the average ping for pros in North America (67 ping) vs the average ping for pros in Korea (8), Europe (20) and China (6) might seem ridiculous. You’re talking about tenths of a second. However, when compared to other sports where hundredths of a second matter, these differences become more pronounced. A difference of just a few milliseconds can have a pronounced negative effect on player performance and experience. People have famously talked about Michael Phelps undergoing rigorous training to shave off milliseconds from his overall time.
Let’s look at average ping times for soloqueue practice of pros globally. In Berlin, where the LEC headquarters is and many European pro teams stay, average ping to EUW servers from Berlin is 22ms. Average pro ping in Korea is 11ms. Average pro ping in NA is 67ms. Average pro ping in China is 10ms. Setting aside cultural differences that may contribute, the teams with the lowest ping have won the most trophies over the past 10 years. Korea has won 6, China 2, Taiwan 1 (average ping in Taipei: 9) Europe 1. Everyone else: zero. In fact, the best NA has ever done at worlds has been two third place showings: once in 2011 and once in 2018.
To understand this, let’s look at raw data.
The average human reaction time is 250 milliseconds. With training or good genetics, that can be lowered to between 140–170ms. Players like Faker and Dopa consistently exhibit reaction time of between 110–130ms, which should be considered a genetic gift. Ping works with reaction time additively. A player sees stimulus on his screen, that stimulus is cycled through his brain and reaction time travels to his input (mouse, keyboard), then the mouse and keyboard latency is processed (another issue) and then server latency is processed for the action to hit the server. Since mouse and keyboard latency is fairly consistent through most brands of keyboard, we’ll ignore it in the scope of this article and focus on server latency.
Let’s take a generic pro and call him “Pro X.” If Pro X has a reaction time of 180ms, by playing in Korea he has a total reaction time (TRT) of 188ms, an increase of 4%. If he plays at a gaming house in Los Angeles, he would have a TRT of 247ms, an eye-watering increase of 37%.
The way this argument is traditionally refuted is that pros practice in the tournament realm, which as a LAN system has little to no latency. Let’s call it 1ms. Pros playing in the tournament realm, the argument goes, are essentially on the same footing mechanically as pros from other countries. That practice is given a greater value than the practice that pros receive in soloqueue, so they should be roughly equivalent mechanically to other professionals. However, there are two problems with this argument. The first is that pros spend far longer playing on the public ranked ladder than they do in practice. The second argument is that even proven veterans from other leagues eventually fall to playing at the NA level. That is to say, sub-optimally.
But they aren’t.
Let’s take two pros, one from Korea and one from North America. The Korean pro is slow, and has a base reaction time (BRT) of 180ms, giving him a TRT of 188ms. The NA pro is remarkably fast, with a BRT of 150ms, with 67 ping that gives him a TRT of 217ms. Let’s say the 3 week bootcamp in Korea was enough to lower his overall reaction time by 15% to better match the server. That means his TRT at the time of the tournament is now 184ms, while the Korean pro is still at 188. This shows that even players that possess physically faster reactions to visual stimuli will be on a barely even playing field with a much slower Korean pro.
Correlation is not causation. However, it is not far-fetched to see players reacting sub-optimally in championship settings due to the increased average latency that they perform under. This can be seen indirectly in the deterioration of professional players’ skills upon moving to North American teams to play there. The human brain is especially adept at understanding rhythm, and building latency into activities. Take, for example, a popular mall game called Stacker. In order to reduce payouts, they build latency into the button presses. However, understanding that people adapt this way, they randomize the latency of button presses in order to throw off players. League of Legends pros practicing from LA have latency that can be as high as 71ms and as low as 64ms, depending on a variety of factors. Why should this not be considered an insurmountable handicap?
All physiological arguments aside, let’s translate this to the sports world. Professional athletes practice in ideal conditions. Asking a professional athlete to practice in less than ideal conditions would be ridiculous. In fact, if you asked a professional football team to practice on a college field, they would be outraged. Asking Lebron James to spend half his day shooting baskets in a neighborhood park’s concrete basketball court would be laughable. Making golf pros play on poorly maintained links with used clubs would not be expected to produce championship quality players. To step even closer, why do esports players need powerful computers? Why use good chairs? Why use mechanical keyboards? Why not practice on MacBook Airs, sitting on a bed with their mousepad on a textbook?
It’s not just negligent to ignore competitive advantages of this scale and prominence, it’s tangibly destructive.
How can this problem be fixed? There are three possible solutions, each of which presents unique challenges. In no particular order, here are three possible solutions.
First, teams could base themselves in Chicago instead of LA, allowing players to practice on 10–12 ping since the servers are located in central Illinois. This presents the challenge of broadcasting LCS, for which players would have to fly to LA every week for games. This is not ideal, but also not different from the current life of a professional athlete who might travel on a weekly basis. A flight from O’Hare to LAX is only three hours and thirty minutes, meaning the travel would not be overly problematic, particularly if the players were in first or business class, or on a chartered turboprop or corporate jet. Teams could also purchase a team jet for around a million USD, which could then serve the company for 5–10 years and be cheaper and more convenient than chartering every weekend or purchasing first class tickets every weekend.
Second, Riot could split the NA population with a server split, splitting into NA west and NA east, with new servers in Nevada or California. This would drastically cut overall server population for ranked, but this difference could be made up with heavy advertisement and promotion to build up both populations. If NA East servers were moved to New York, some EU West players would also play on this server since ping between the UK and NYC is, on average, 70. That would prevent catastrophic server depopulation. Having additional promotions, such as offering new players that register with a phone number and reach level 10 a one-time gift of $20 of RP would both ensure unique players are joining and provide a reason for players to stay.
Third, players could move to Chicago, and Riot could move the LCS studio to Chicago. This would be inconvenient for Riot employees that also work as casters, or casters that live in the LA area, but this could be solved through either weekly travel or with the same telecasting setup that was used for Worlds 2020. The LCS studio in LA would hold casters, while an esports arena or LoL Park analogue in Chicago would hold the actual game. This is a modern approach that could seem too different or aggressive, but is still feasible.
Occam’s razor is a demanding foe, and any discussion about player mentality, North American esports culture, or server population merely distracts from the core issue.
NA’s consistent failure to perform on the international stage have been blamed on a number of baroque and unquantifiable reasons. However, of this menagerie of excuses, one continues to rise to the top. A simple, yet expensive, problem to solve. Professionals have talked about it, streamers have mentioned it, pros have raged against it, but the problem remains. Occam’s razor is a demanding foe, and any discussion about player mentality, North American esports culture, or server population merely distracts from the core issue. It’s the ping.