Events like the Mid-Season Invitational (MSI) are exciting because they provide opportunities to measure regions against one another, watch the best go up against the best, and see clashes of play style that often produce unpredictable results.
In the leadup to MSI, there’s a lot of speculation about how different players match up one-on-one, and no phase of a League of Legends embodies 1v1 and head-to-head like the early laning phase, where pure mechanics reign (alongside, to be fair, champion selections and tactical choices like lane swaps). To give some context for how the lane phase might shape up between different players, we’re going to review Average Creep Score Differential at 10 Minutes and Average Gold Differential at 10 Minutes, supported by some additional numbers around experience (XP) and early-game combat. These numbers reflect each player’s ability to not only secure his own farm and other income sources, but also to deny his opponent.
We’ll look first at each role individually, then summarize what we can expect from each team at the end of the article.
All of the numbers in this article are drawn from each team’s regional playoffs. (For Be?ikta?, we’re using the Semifinals and Finals of the International Wildcard Invitational playoffs, with the caveat that statistics were not available for Game 1 of their Semifinal.)
Note: Edward Gaming’s players are not included in the charts because the LPL unfortunately does not release its match histories. The Chinese team will remain somewhat of a statistical enigma, but for some insight you can check out Kelsey Moser’s excellent article MSI Preview: the 5 Stages of Edward Gaming.
Top lane players are usually the sacrificial lambs of the lane-swap meta, being forced to accompany their Junglers on some or all of the first jungle clear, then trudging reluctantly into their lane to soak up experience and pick up a stray CS here and there, whenever the opposing AD Carry and Support allow them to, or perhaps wandering into the jungle again to grab a Gromp or some Krugs (especially if they’ve taken Smite).
Some teams are more willing to initiate the lane swap, and some champions perform better in a 1v2 lane. While CS and gold differential numbers don’t show us how often a team lane swapped, they do show how well the Top laner was able to pick up farm compared to his opponent, whether facing off 1v1 or scraping by in the 1v2.
Among MSI Top laners, Dyrus had the best early-game showings. On average, he was up 7.6 CS on Impact and Balls at the 10-minute marks, which accounts for most of his 175 gold lead. Dyrus also had an average lead of 440 XP at 10 minutes.
Huni is a strange beast, statistically, with a small CS deficit at 10 minutes on average, but a massive 10-minute gold lead. That gold came from multiple sources. One source was the jungle: Huni may have been behind in overall CS, but he was usually ahead in jungle CS,and jungle monsters are generally worth more gold than lane minions. Another source was early tower kills: his team took three pre-10 minute tower kills in 10 games. Huni also gained gold leads from early kills: he averaged 0.7 kills at 10 minutes in the playoffs.
MaRin‘s numbers are weak, but he was up against LCK competition and should have an easier time against some of his MSI opponents. Thaldrin‘s numbers are reasonably strong, but he will be facing a big step up in opponent quality at MSI.
Ziv had average early games in CS and gold, but was the only one of these five Top laners to be behind in XP at 10 minutes, on average (-46.5 XP @ 10).
Based on these numbers, Dyrus should be positioned well to win his lane: he came through strong in CS, gold, and XP while facing fairly strong competition in Impact and Balls, and should be up to the international challenge. Ziv, on the other hand, may struggle.
Junglers have two main pathways to success in the early game: one is hard farming, and the other is aggression via ganks and invades. A Jungler with strong CS differentials probably farms harder, and can gain gold leads that way, but it’s also possible to gain gold leads from ganks while being even or behind in CS. The numbers below give an indication of not only each Jungler’s early-game play style, but also their ability to carry that style to success.
Like his teammate in the Top lane, Reignover boasted large gold leads at 10 minutes, partly because of a CS advantage and early tower kills, but also because of his average 0.3 kills and 0.9 assists at 10 minutes.
Santorin impressed in the first 10 minutes with CS leads alongside an average of 0.4 kills and 0.8 assists. He was usually ahead in XP, as well.
Bengi was out-CSed, but had slight gold leads because of a ton of early-game aggression: he averaged 1.0 kills and 1.7 assists at 10 minutes! This suggests that he focused more on early ganks and playmaking than on farming, and since he didn’t give up any pre-10 deaths, his playmaking was certainly successful.
Mountain was out-jungled quite heavily early on, with large CS and gold deficits, along with large XP deficits and an average of 0.4 deaths at 10 minutes.
Theokeles performed well in the first 10 minutes, with CS and gold leads, as well as XP advantages, but as with the rest of his team, it may be difficult to maintain that pace against stronger opponents.
Reignover and Bengi will both be looking to flex their early-game muscles through aggression, but expect Santorin to take a more balanced approach. Mountain may have a difficult time remaining relevant prior to the mid game.
The Mid lane is the purest 1v1 matchup in the first 10 minutes, since it isn’t characterized by lane swaps or big decisions between farming and ganking. The biggest piece of context for these Mid lane statistics is champion choice: some Mid champions just want to survive until level 6, while others love to trade and bully to secure early advantages, or may do some early roaming to gank the side lanes. The numbers below highlight these play style choices and each player’s relative effectiveness at that style.
In the NA LCS playoffs, NA’s favourite adopted son, Bjergsen, carried on his dominant 1-on-1 performances in the Mid lane, accumulating big CS and gold leads over XiaoWeiXiao and Hai. Bjergsen also earned 327 more XP than his opponents at 10 minutes, on average. These advantages came partly from Bjergsen’s average 0.4 kills and 0.5 assists at 10 minutes, with no pre-10 deaths.
Fnatic’s Febiven had a similar pattern to his teammate, Huni, with CS deficits but gold advantages. Febiven, however, wasn’t involved in much early-game fighting, and earned his gold leads more through denying his opponent and soaking up early tower kill gold.
Faker‘s lower averages, which include a small average XP deficit, need to be put in context of his high-quality competition. We also have to bear in mind that Faker only played in four playoff games. One of those games was a very poor showing on Xerath, which dragged his numbers down heavily, but his two Lulu games and his game 5 on LeBlanc were superb showings.
Easyhoon‘s early games were a bit weaker than Faker’s, but that is partly explained by his use of Azir, Cassiopeia, and Vladimir, who have a harder time early on than Lulu or LeBlanc do.
Westdoor was typically very far behind in CS at 10 minutes in the LMS playoffs, but managed to keep the gold relatively close due to having an average of 0.6 kills and 0.7 assists by the 10 minute mark. The AHQ Mid’s preference for combat over farming could serve him well at MSI, but only if his opponents let their vision control lapse and don’t see his roams coming.
Bjergsen and Faker will be the two Mids to watch in early laning, if they continue to pick stronger laning champions. Westdoor and Energy are likely to be pretty heavily outclassed. The question for Westdoor is whether he’ll able to overcome early deficits to make his presence felt in the mid game.
When a lane swap happens, the AD Carry is put in a position of strength, able to CS at will against most Top champions. The ADC’s decision becomes whether to fast push and try to take the tower down early, or whether to freeze the lane and try to deny the enemy Top laner from CSing himself. ADC CS and gold differentials will reflect that decision to some extent: an ADC who fast pushes more often will build larger CS leads, while freezing will slow down his CSing pace.
Nardeus had dominant laning performances during the International WildCard Invitational playoffs, though his CS advantages didn’t translate into very large gold leads. He built up pretty big XP advantages, as well. Facing much stiffer competition at MSI, will Nardeus be able to replicate his early-game success?
Fnatic’s Steeelback tended to come out behind in the first 10 minutes, despite the early tower gold he received from Fnatic’s three pre-10 minute towers. He didn’t give up any pre-10 deaths; his deficits can only be attributed to either heavy lane freezing, or simply poor CSing.
WildTurtle averaged 0.4 deaths at 10 minutes, showing how much his opponents worked to suppress him early on, and that is evident in his CS and gold differentials. Despite this, WildTurtle averaged a slight XP advantage, on average. WT’s performance in the laning phase will depend a lot on how much opposing teams prioritize him as an early target.
SKT’s Bang had only small leads, despite averaging 0.3 kills and 0.6 assists at 10 minutes, with only 0.1 deaths.
Based on these numbers, Nardeus has a chance to represent his team well in the early game, potentially standing up to his international competition. WildTurtle and Steeelback have some work to do to improve their early-game showings at MSI. Bang’s small advantages should not be overlooked, especially given the level of competition he faced.
I say it all the time: Support is probably the hardest position in League of Legends to measure statistically. Creep Score and earned gold are not meaningful metrics for Supports.
As an alternative, we’re going to look more closely at Supports’ average kills, assists, and deaths at 10 minutes, and also review their overall vision contributions.This will give us a sense of how active the different players are, and what their activity is directed towards.
Look for Dumbledoge and Lustboy to be especially active early on: Dumbledoge averaged 0.7 kills and 0.7 assists at 10 minutes, adding an average of 0.5 deaths, while Lustboy averaged 0.1 kills and 0.8 assists to go with 0.3 deaths.
Wolf accumulated 0.3 kills, 0.5 assists, and 0.4 deaths at 10 minutes, on average.
Albis was plenty active early on, putting together an average of 0.3 kills and 0.4 assists, and didn’t give up any pre-10 deaths in AHQ’s playoff run.
YellowStar kept the early combat to a minimum, averaging 0.1 kills, 0.2 assists, and 0.1 deaths at 10 minutes.
Based on these numbers, we can expect aggression and playmaking from Dumbledoge and Lustboy and more cautious, reactive play from Albis and YellowStar, with Wolf somewhere in the middle.
Wards Placed and Cleared
Among MSI Supports, Albis of AHQ was the most active in the vision game during his region’s playoffs. His ward clears per minute were especially high, clearing more than 25% more wards than his closest competitor.
Dumbledoge of Be?ikta? placed fewer wards than his counterparts, but kept up fairly well in ward clears.
Lustboy, YellowStar, and Wolf had middle-of-the-road vision numbers.
Not a lot separates the MSI Supports in vision control, outside of Albis’s impressive ward clearing pace.
Team Summaries and Conclusion
Team Solo Mid
In the NA LCS playoffs, TSM was usually able to secure leads in Top, Jungle, and Mid, giving up deficits on their ADC. Throughout the Spring Split, TSM was more likely to give gank priority to their ADC than their Top lane, but that doesn’t appear to be true of their tactics in the playoffs, based on these numbers. Which approach will TSM favour at MSI?
Judging by Lustboy’s stats, we can expect TSM’s Korean Support to wander the map early on and try to get involved in creating action. But don’t expect TSM as a whole to go all-out on early tower dives or ganks; their overall numbers are fairly balanced.
TSM can be expected to win more early games than they lose.
Outside of their Jungler, Fnatic was usually behind in CS at 10 minutes, but still secured gold leads in Top and Mid because of tower pressure and early kills. If Fnatic continues to focus on early playmaking, they will have to execute without any mistakes against high-end competition like SKT, EDG, and TSM.
Expect Fnatic to seek ganks and make big play attempts early. If their opponents execute successful defensive play with good lane freezing and vision control, Fnatic may put themselves behind by forcing plays with low probabilities of success.
SK Telecom T1
SKT played very close early games, on average, so the big question for the Koreans will be whether their first 10 minutes are actually middle-of-the-road, or whether those numbers are just a symptom of how stiff their competition was. Will their early game look better against international competition than it did in their domestic playoffs?
Depending on whether Faker or Easyhoon is playing in any given game, SKT’s early tactics could vary significantly. Faker will be more likely to be aggressive and make plays, joining up with Bengi’s aggressive ganking style, but Easyhoon is more likely to sit back and farm in preparation for the mid game. With Faker in, SKT will probably win early game most of the time, but with Easyhoon playing they are more likely to just keep things close.
AHQ e-Sports Club
On the whole, AHQ had poor 10-minute numbers, and their path won’t be getting any easier. If AHQ takes the same approach to MSI as they did to the LMS playoffs, they will be hard pressed to survive into the mid game with enough resources to mount a comeback through team fights.
AHQ’s early game weakness will likely be punished quite heavily by their MSI opponents.
Be?ikta? e-Sports Club
Be?ikta? had excellent 10-minute numbers in the IWCI playoffs, outside of their Mid lane, but they can’t expect to gain the same early leads in a tournament of MSI’s calibre.
Be?ikta? should be able to stay relevant and perhaps keep the early games close, maybe even gaining small advantages, but will it be enough?
Since LPL stats are not available, EDG is a bit of an unknown, but according to Kelsey Moser of theScore eSports, EDG shouldn’t be expected to dominate the early game with aggression. She writes, “Since the patch change [to 5.5/5.6], Edward Gaming have gone back to where their primary shotcaller and captain are most comfortable — the late game. Clearlove, Koro1, and Deft have grown in terms of their ability to impact the early game all split, but with the Cinderhulk meta taking over, there’s even more benefit to stalling and playing EDG’s 2014 team fight style.” That early game play style may translate into strong CS numbers at 10 minutes, but might not lead to big differential numbers if they aren’t aggressively working to shut down their opponents with ganks and dives.
EDG may look to keep the early game close, enabling them to win their games in the mid and late game.