NA LCS caster David "Phreak" Turley recently faced strong public backlash after releasing his marksmen tier list, one that didn't sit well across fans, players, and even team support staff. Discussions about the ranking immediately sprung up across a number of social platforms, and in the interest of transparency, Phreak himself showed up later in a Reddit thread to share the reasoning behind his choice of order.
As it turned out, the ranking from top to bottom was mainly determined using comparisons of player stats. This led to even more fervent criticism by analysts across the globe, experts and Reddit alike.
There's no getting around it: Phreak's methodology has too many holes. Here's why, and how we can improve it.
Let's take a look at Doublelift's placement. With fantastic laning stats but very low damage share relative to the gold he obtains, he’s placed 6th in this tier list. Phreak states as the main issue, "He's getting resources, and doing comparatively nothing with it."
However, key information around Doublelift's performance casts a different light on his numbers. As we've discussed in a previous video, TSM has been experimenting a lot with engage-heavy bottom lanes. For them, the goal is to start the snowball in the bot lane. The extra pressure and resources that the team invests into Doublelift explain why he has fantastic laning numbers.
However, engage-heavy ADCs like Ashe and Jhin, which Doublelift has played in nearly half of TSM's games so far, lie on the lower end of damage output for all ADCs. It’s only natural that his damage relative to gold share is lower than other players’ who have been playing more Tristana or Xayah.
Based on stats alone, Phreak's explanation seems very reasonable. But with context on top of the numbers, the reasoning becomes tougher to justify.
The Spring Split MVP takes 7th place on the list because his laning phase is poor, despite his fantastic teamfight contributions.
But that doesn't quite add up. How can a man hailed as last split’s MVP be a monster in teamfights but also an atrocious laner? The story that the stats tell is that he's both horrible and great at his job as an ADC. Let's look beyond the numbers to see why this may be the case.
Phoenix1 has gone through multiple Supports and Junglers just this season. Both of these roles are critical to Arrow's early laning performance, aside from his own skill. Weaker synergy, due to acclimating to new players, could also contribute to lowering laning numbers. He's facing enemy bot lanes that don’t have that disadvantage. The struggle seems to be more of a team issue than anything else.
Does he Arrow deserve the 7th-place rank he got? Perhaps--after all, part of his job is to lead his team to victory. But his bad laning numbers shouldn't be only reason for his low rank.
This brings us to another key point: most individual player stats are ultimately a collaborative effort by the team. Playstyle, composition, and the team's performance all influence a player's numbers. Accurately judging individual player performance in a vacuum becomes very hard.
NUMBERS CARRY BAGGAGE
As anyone who has ever played solo queue can tell you, your performance (and the associated numbers) can be heavily influenced by factors completely out of your control.
Take damage share, for example. Damage share measures a single player's damage output, relative to the total amount dealt by his team. It's often used as a way to measure the contributions (and, implicitly, value) of a player to his team.
But this stat cannot stand alone.
For AD carries, the longer a game goes, the larger their damage share will be. This means that damage share will naturally make marksmen look better on late-game-oriented teams. Using this stat alone as a measure of individual quality, ADCs on slower teams look way better than their counterparts on teams that try to close games quickly.
ACCOUNT FOR CONTEXT
While the reasoning behind Phreak's ADC tier list was sound from a pure numbers perspective, it unfortunately ignored many of the less tangible factors that determine players' value to their teams. This is why there was such a strong response from the community.
Stats are a huge part of performance analysis that can't be ignored—but they must be taken with a grain of salt. While they are useful for basic head-to-head comparisons and as fun bragging rights amongst friends, in pro play, a deeper look at context is necessary before drawing conclusions. Each team's composition, playstyle, and external circumstance affect their players' numbers too much for stats to be the sole decider of a player's worth.