Interstellar collision: The LPL group format and two isolated metas

by theScore Staff Feb 15 2016
Thumbnail image courtesy of LPL / LPL Screengrab

In the League of Legends Pro League, teams interact within the confines of a unique group stage format. Multiple teams have also come forward to express the inefficiency of the scrim culture, placing more priority on professional games to develop strategically. This combination has seemed to create two unique approaches to the meta game within the LPL, including different champion priorities and varying importance on different phases of the game.

More of a focus on lane swapping has evolved in Group A, while Group B's games tend to revolve around closing games out quickly after skirmishes and obtaining the first Baron. Understanding the differences between the kind of game Group A teams and Group B teams have been playing sets up the possibility for interesting interactions in Round 2, when Group A teams play against Group B teams for the first and only time before the playoffs.

The objective time indicator

Average time teams in each group first take an objective

Objective Group A Group B A-B
Kill 7:56 7:56 0:00
Dragon 16:46 16:16 0:30
Tower 11:21 13:14 -1:53
Herald 16:43 14:55 1:48 
Baron 30:14 27:07 3:07
Inhib 33:18 31:51 1:27
Game 36:50 33:20 3:30

Teams in both groups have similar average times at which they take their first kills or dragons, but beyond that, numbers diverge by at least a minute and a half. On average, Group A teams take their first turret two minutes earlier than Group B teams, but Group B teams have shorter game times by three-and-a-half minutes and take Barons three minutes earlier as well.

The first step is to look at teams that seem to be driving these averages within their groups and see if it's indicative of the entire group or just one or two extreme outliers.

Group A Objective Timers

Team Kill Dragon Tower Herald Baron Inhib Game
EPA 6:36 14:47 12:32 16:09 30:55 29:48 36:56
LGD 8:44 14:40 14:6. 18:37 29:34 37:02 36:43
M3 6:28 11:42 8:25 17:11 28:42 33:45 37:22
QG 8:59 15:27 14:34 19:42 30:35 31:55 34:34
SS 10:23 20:20 8:03 15:16 30:48 35:46 39:53
WE 6:19 20:03 10:03 16:34 29:51 32:13 34:56

In observing the low turret average time, it appears to be primarily driven by three teams: Snake, Masters3, and Team WE, which are below the stated average of 11 minutes and 21 seconds for the group. Average Baron time is very similar across the group with Masters3 taking the earliest average first Baron at 28 minutes and 42 seconds. Late inhibitor times seem to be driven by LGD Gaming, Masters3, and Snake Esports. The later game times don't seem to be driven by any team in particular, with the exception of the QG Reapers and Team WE standing out as shorter time outliers, but their games still end later than the Group B average.

Group B Objective Timers

Team Kill Dragon Tower Herald Baron Inhib Game
EDG 11:05 14:45 16:19 13:40 31:33 33:08 34:14
HYG 6:10 12:08 12:25 11:01 27:37 31:51 32:24
OMG 7:45 17:16 11:06 17:07 29:07 33:29 36:54
RNG 6:26 19:07 12:17 17:02 25:01 32:07 33:10
VG 8:04 19:51 13:17 14:32 25:39 31:19 31:22
iG 8:46 13:30 14:31 14:12 27:28 29:43 32:13

OMG tend to be the outliers with the shortest time until they take their first turret at a value closer to the Group A average. Early Barons are driven by Royal Never Give Up and Vici Gaming primarily, but Hyper Youth Gaming and Invictus Gaming hover near the average. EDward Gaming and OMG appear to be the outliers with average first Baron, first Inhibitor and game times more similar to Group A's average.

The lane swap phenomenon

Investigating the driving factor that seems to create this difference in objective timers, in particular the turret, Snake and WE have participated in more games involving lane swaps than any team in Group B. This has increased the overall rate of lane swap games in Group A relative to Group B.

Group A lane swap rates

Team Lane swap games %Total Duo top %Total
EPA 4 33.33  3 25
LGD 4 36.36  3  27.27
M3 4 40  1 10
QG 5 50  3 30
SS 11 84.62  9 69.23
WE 8 66.67 5 41.67

We define a lane swap as a case where, following the first clear of a jungle quadrant (or the equivalent of two small camps and a buff) by a jungler in the game, there are a different amount of laners from opposing teams in the top and bottom lanes. In the current meta, that usually means that one lane contains the duo lane and the top laner of one team pushing against an empty lane and the other lane contains the opposing team's duo lane and top laner.

The data indicates that Snake in particular have played a lot of lane swap games, having swapped in 11 of their 13 games. WE have also lane swapped in more than 50 percent of their games, and QG Reapers have played lane swaps in five of 10 of their LPL games played.

In tracking the number of instances in which a team shows willingness to send their duo lane to the top lane at game start, we can also approximate the instances where a team is open to initiating a lane swap. In the LPL, scouting wards that anticipate lane swaps near the enemy team's base are rare, meaning many swaps are blind. As a result, teams sending their duo lane top at the start of a game is a rough indicator for willingness to initiate a lane swap.

This statistic is not perfect, as teams could anticipate their opponents sending the duo lane top in a blind swap and only send their duo lane top to counter swaps. Yet we see WE and Snake continue to drive this statistic, with Snake sending their duo lane top in nine of 13 games and WE sending their duo lane top in five total games. By contrast, Group B has very few teams driving lane swaps.

Group B lane swap rates

Team Lane swap games %Total Duo top %Total
EDG 1 11.11 0 0
HYG 2 18.18 2 18.18
OMG 4 36.36 5 45.45
RNG 3 25 3 25
VG 2 16.67 3 25
iG 0 0 1 9.09

Shockingly, given the popularity of the fast-push lane swap, EDward Gaming have played only one lane swap game, and Invictus Gaming have yet to play a match that utilized a lane swap. OMG has driven the majority of lane swap games in Group B, but still have only sent their duo lane top in 45.45 percent of their games.

Group B's incredibly low lane swap rate has driven up the time until each team takes their average first turret. It has also seemed to create a more swinging game style where games are decided almost entirely by the first Baron taken. Group front-runner, Royal Never Give Up, has an average gold deficit at 10 minutes, but has the earliest average first Baron time in the league at 25 minutes and a 100 percent win rate on securing the first Baron of the game. The team gains momentum around mid game to turn a 10 minute gold disadvantage into the highest gold lead at 20 minutes of any team in the LPL on average.

By contrast, in Group A, Snake Esports have the highest average gold lead in 10 minutes (362 gold) and the second highest average gold lead at 20 minutes (1,061 gold), but an individual game win rate of below 50 percent. They're able to manipulate lane swaps well for strong leads, but poor team fighting has led Snake Esports to lose more games than they've won. QG have manipulated lane swaps in a different way, freezing for AD carry Yu "Peco" Rui while losing turrets to make comebacks later on in the game. Similar to Royal, once they secure a mid game lead, they can close, but it usually takes them longer to get there.

Champion drivers

Other teams in Group A have adopted QG's freezing approach in dealing with teams like Snake. More scaling picks have permeated Group A, such as Ezreal, which has been picked in 20 games in Group A, but only 11 in Group B. Viktor has been picked in 10 Group A games, but only four Group B games. Kassadin has been picked in seven Group A games, but zero Group B games. Braum, known for stopping wave clear and sieges, has also enjoyed more popularity in Group A than Group B, lengthening games. Team WE, LGD, and QG have all favored giving free side lanes to their AD carry for farming and freezing.

By contrast, Group B has favored picks like Kalista and Corki more heavily than Group A. Corki's early power spike allows the pick to dominate lanes, and Kalista can take objectives (like Baron) earlier and enjoys an oppressive 2v2 lane in conjunction with picks like Poppy (also more popular in Group B than Group A).

These different power picks create different power spikes that put more emphasis on mid game fights and skirmishes in Group B, and more emphasis on late game fights in Group A. This is evidenced by gold distribution variance from the LPL average. In general, Group A teams, with their method of freezing side waves for primary carries, distribute slightly more gold to top, mid, and AD carry roles, while Group B teams, with a more mid game skirmish based approach and more Graves jungle picks, tend to put more of the team's gold share on junglers and supports relative to the league average.

In the following radar charts, the blue line represents the difference from the league average of the respective group's average percentage of team gold for each role. The orange line represents the league average.

Implications for Round 2

In Round 1, the data show that Group A and Group B developed different approaches to the meta game on Patch 5.24. Round 2 will introduce teams to Patch 6.2 following the end of the New Year break in the LPL. Group A teams with longer game times and more focus on freezing for scaling carries following a safe lane swap will collide with the more skirmish-based approach of Group B.

A change in patch will see a slight alteration in champion priority, but in other regions, most of the same priority picks have remained with patch shifts. The 3v0 fast push lane swap style still dominates. Teams from Group A could have a workable advantage in dictating the pace of the game with more lane swap experience. Teams like EDward Gaming showed disorientation in rare instances where they were forced into a lane swap scenario. OMG and Vici Gaming could have an advantage in this regard, as they appeared most comfortable with lane swaps of Group B's teams. This would help them gain some sorely needed ground in the standings.

Snake Esports' ability to understand map pressure and how and when to manipulate freezes during and following a lane swap has given them massive early leads. The recent introduction of Im "T0M" Jaehyeon to Snake's lineup could allow them to reintroduce Ceng "U" Long, giving them the strong solo lane synergy between Li "Flandre" Xuan-jun and U that made their teamfights powerful in 2015.

All eyes will be on QG Reapers, EDward Gaming, and Royal Never Give Up. QG have been able to counter the prevailing lane swap style by freezing side waves and making comebacks, but the more aggressive approach of Group B might be what it takes to unsettle them.

EDward Gaming's noted discomfort in lane swaps may make them easy prey for some of the more proficient Group A teams, but their late game team fighting approach will allow them to slot into the other dimensions of Group A clashes simply. Royal's jungler doesn't ward; they could lose out in a lane swap easily without knowledge of the opposing team's movements, but they may also take advantage of late scaling compositions by forcing skirmishes.

Despite having different styles, Royal and QG tend to distribute their resources in a similar fashion.

This should prove to be one of the more interesting clashes of the second round. QG are more refined in their playstyle, but Royal have a better understanding of how to maintain momentum than any team in the league. If they can adopt Snake Esports' knowledge of the lane swap meta, they look to become a truly menacing threat. Royal don't face Snake until the last week of Round 2. They need to figure it out before then if they want a chance against QG.

Note: Most data in this article are hand-collected from VODs, but end-game gold data are compiled from match pages on Carry6.

Kelsey Moser is a staff writer for theScore eSports. You can follow her on Twitter.