Archive:Leaguepedia Articles/2014 All-Stars Champion Pairing Analysis

From Leaguepedia | League of Legends Esports Wiki
Jump to: navigation, search

2014 All-Stars Champion Pairing Analysis

All Stars Champion Pairing Analysis

The 2014 League of Legends All-Star Invitational took place this past weekend in Paris, France. North America's Cloud 9, Europe's Fnatic, China's OMG, Korea's SK Telecom T1 K, and Southeast Asia's Taipei Assassins all met for a rare glimpse into international competition.

The event showcased the similarities and differences between regional metas and showed how each team values different champion picks. A lot of information can be learned by looking at champion pick and win rates, but in this article the focus will be an analysis of pairs of champions across the 17 games of the tournament. What sets of two champions were picked together the most, and how did they fare?

While reading this article, keep in mind that our sample size is relatively small, and two of the teams that played---SKT and TPA---must be, to some extent, considered outliers: SKT went undefeated over 9 games, while TPA lost all 4 of their group-stage games and were then eliminated. That said, we can still make some observations, and the results here, being inter-regional, are a good foundation for further inquiry.

To understand how well each champion pair did, it’s important to know how each champion performed individually.

AllstarsPairs1.png


ShyvanaSquare.png Shyvana, TwitchSquare.png Twitch, LuluSquare.png Lulu, Kha'ZixSquare.png Kha'Zix, and LucianSquare.png Lucian were the five most-picked champions over the course of the All Star weekend. Shyvana had a strong showing with 9 wins and 4 losses, uncontested in the other 4 games. TPA, without a single win across the weekend, was the only team to never pick her; TPA instead opted for other choices, including LissandraSquare.png Lissandra and RenektonSquare.png Renekton, which only they played.

Twitch also did well, with 7 wins and 3 losses across three separate teams: OMG, SKT, and C9. The other top picks did not do as well; each of Lulu, Kha'Zix, and Lucian won only 4 times out of their 10 games. Throughout the weekend, no champion picked at least three times had a 100% win rate. With that knowledge, let’s now take a look at pairings: win rates of sets of two champions that were often picked together. The chart is ordered primarily by number of games that the champions played together, then by number of wins, then alphabetically, grouping champions together were possible.

Article2.png


The pairings shown here are lane-indifferent: for example, games with Lee SinSquare.png Lee Sin top and LeonaSquare.png Leona support are lumped together with games with Lee Sin jungle and Leona support. Each game contained 20 different champion pairings (10 from each team), so there were 340 pairings in total. Of these, 69 were played more than once, 30 more than twice, and there were 228 unique entries.

Of these sets, the pairings of Shyvana-Thresh and EliseSquare.png Elise-Lulu were picked the most, 5 times each. The first won 4 of 5 games, played by OMG and SKT; the lone loss coming when OMG fell to SKT in game 2 of the finals. Elise-Lulu did not fare as well, with a 2-3 record across the tournament, played by Cloud 9 and TPA.

Looking at botlane pairings, Lucian-Thresh was played the most at 4 times. VayneSquare.png Vayne-ZyraSquare.png Zyra, Twitch-Morgana, and Lucian-KarmaSquare.png Karma each saw 3 games together. Among those top pairings, there isn’t a single repeat support, showing both the larger variety in support picks over AD Carry picks and also potentially that teams were reluctant to pair certain supports with different carries.

As mentioned above, no single champion picked more than twice had a 100% winrate---however, of sets of champions paired together at least three times, 4 had perfect winrates. The most successful was Twitch-Shyvana, played 3 times by 3 different teams (OMG, C9, and SKT), with 3 wins. What about the other perfect-winrate pairings? Shyvana-Zyra, Vayne-Zyra, and Lee Sin-Zyra were also 3 for 3, but those all came from a set of 4 games from SK Telecom, so it’s difficult to draw any conclusions here.

On the flip side, Karma and Kha'Zix were picked together three times, by Cloud 9 and Fnatic, and lost all three times. This is unsurprising as both champions had poor individual win rates, at 4-6 for Kha'Zix and 1-4 for Karma.

Let’s now look at Lee Sin. He had a decent tournament, played by every team at least once and winning 5 of his 8 matches. His 3 losses came when Leona was the support on his team, and each of those losses were by different teams: OMG, Fnatic, and TPA. His only success with Leona came at the hands of SKT. Leona herself had a pretty bad weekend overall, winning only 2 of 7 games.

The next pairing to look at is Trundle-Twitch. TrundleSquare.png Trundle lost both games in which he was picked without Twitch, but together they won 3 of 4 games, 2 wins from SKT and a win and a loss from C9. The first Trundle-only loss was OMG to C9 in the group stages, and the second was OMG's revenge in the semifinals.

The fact that Kha’Zix-OriannaSquare.png Orianna fared so poorly is somewhat surprising; in theory it’s a combination of champions that works well due to Kha’Zix’s ability to leap in with Orianna’s ball to prepare a Shockwave.

Finally, Lulu had a decent 4-6 record overall. With regards to pairings, Lulu performed well with Shyvana, going 3-1, while without Shyvana her overall record was an abysmal 1-5. The Chinese team favored the Dragon's Descent + Wild Growth combo---whenever OMG picked Lulu for their mid lane, they picked Shyvana for their top lane. Fnatic and TPA, on the other hand, never picked Shyvana and Lulu together. Maybe they could have had greater success if they followed OMG’s example.

Finally, let’s look at team differences. Note that the total pairs will always be 10 * (number of games played), and a large number of unique pairs relative to total pairs corresponds to fewer similar team compositions picked.

AllstarsPairs3.png


TPA only played one pairing multiple times over the course of four games, Elise-Lulu, for a total of 39 unique pairings. They also picking a large variety of champions, far more than any other team relative to the number of games that they played. Note, however, that because the set of tournament-viable champions is somewhat small relative to total number of picks, this is a correlation we expect to see; it simply stops being possible to maintain a large ratio of distinct champs played to number of games played for larger numbers of games played.

Cloud 9 made the most effort to keep the same pairings/team composition out of all five teams; they have the lowest number of unique pairs per game, despite the fact that their total number of champions per game is more average. On the opposite end, SKT chose a low total number of champions relative to the number of games they played, but with a fairly high percentage of unique pairs; this indicates that they cared more about picking strong individual picks rather than prioritizing team synergy. Fnatic and OMG fell in between.

While 17 games is far too small a sample size from which to draw sweeping conclusions about synergies, especially considering that one team went through the tournament with an 9-0 record, these statistics seem relevant to look at and can be built upon with additional data.


Published on 16. May 2014
Written by Andy Dietzel (Guest Writer) - @Bold_Strategy and "RheingoldRiver" - @RheingoldRiver
Edited by David "TheDynasty" Spitler - @TheDynastyLoL