Splitting Splyce: A Tale of Two Halves

This article appears simultaneously here on League of Analytics and SplitPush.net, where you can find everything about professional League of Legends.

After showing some promise, Splyce finished the 2016 EU LCS Spring Split in 8th. While SplitPush.net and others kept them in that exact place in the Summer preview rankings, some people had Splyce pegged for minor improvements in the standings. No one would have expected such a big leap. The team currently resides in 2nd place and has shown a frightening amount of poise and consistency as of late.

Splyce showed promise in the first few weeks of the split, with Kobbe and Trashy both playing consistently well, Wunder advertising a large champion pool and marked improvement over last split, and Mikyx proving why he was swapped in during the spring split. Still, the team lacked consistency and especially Sencux’s play had its ups and downs. Fast forward to the second half of the split and the inconsistency and mediocrity seems like a distant past. Just look at the win-loss development below.

Splyce Win Loss

Today, I want to take a look at what changed, with a heavy focus on hard data. For this, I gathered all stats from League of Analytics’ EU LCS team and player pages and calculated them separately for weeks 1-4 and weeks 5-8. Furthermore, I checked out Games of Legends to also include some champion pick developments. Stats alone do not tell the whole story, but they can give a good insight that is less blinded by biases and can point out differences that might not have been obvious watching the games.


New champions emerge


While Kobbe and Trashy show heavy favoritism towards Caitlyn (12 picks – 75%-win rate) and Rek’Sai (15 picks – 66.7%-win rate) respectively, the other players have displayed a diverse set of champion picks and none of them has picked any single one more than six times. For Mikyx, this is mainly due to his strong Bard (6 games – 100%-win rate), which is the most banned champion against Splyce with 47% (vs. 35% ban-rate league wide).

Image courtesy of Riot Games

The general trend is consistent over the eight weeks, but differences emerge in individual champion picks, in particular for Sencux and Wunder. Focusing a lot on Gnar, Ekko and Fizz (which was subsequently banned against them in a lot of games) in the first four weeks, Splyce’s top-laner has shifted his focus. His picks now include current top-meta picks such as Irelia and Shen, but also two wins on Taliyah, Fiora and, amongst others, one on TahmKench. His proficiency on Shen and his recent performance on Fiora (especially in game 1 vs. Fnatic), force bans and allow for more flexibility in the draft.

Sencux also adapted to new patches and changes in the meta. His overall champion diversity is tied for first in the EU LCS, but has dropped in weeks 5-8 compared to the first half of the split. Versus G2, he was the first EU-player to bring out Malzahar and the champion has had a 100% presence in their games since — three picks and one ban by, as well as five bans against Splyce. Not only that, but Sencux is also responsible for 5 out of the 13 picks and 4 of the 6 wins on Kassadin, all while boasting an impressive 10.4 KDA on that champion.


Better start, increased control and more dominance


After winning 50% of their first 16 games, Splyce has since won a league-best 87.5%, with the one loss coming against powerhouse G2. The first four weeks show a team that struggled in the early game, with a first blood (FBP) and first dragon participation (FDP) of 43.8% and a gold difference at 15 minutes (GD@15) of -734 on average. Fast forward to the second half and things look completely different. While their FDP stayed the same, a first FBP of 62.5% means an improvement of over 18 percentage points and a GD@15 of 888, on average, results in a staggering 1622 gold more at the 15-minute mark. The GD@15 stat does not tell the whole truth though. While Splyce’s early game was lackluster in the beginning, the huge average deficit is mainly a result of some large outliers versus teams like G2 and Fnatic combined with average numbers in most other games. On the other hand, the positive GD@15 for the second half includes some problematic early games, especially versus Fnatic and H2k last weekend. That they managed to still win those games speaks to Splyce’s confidence, poise and ability to claw back from a deficit, all traits that are necessary to come out on top of the EU LCS.

Comparing their dragon control and first baron numbers reveal further improvements with regards to the teams’ objective control. While the former rose from 42.9% to 58% between the first and last four weeks, the latter jumped from a slightly below average 43.8% to an incredible 81.3%. Getting the first Baron in four fifth of your games (only games where any Baron is taken) is pretty darn good and should be a sign of a team that has impeccable control over one of the game’s major objectives.

How much of their dominance is a result of improved macro-play and objective control is not 100% clear, but the fact that they have been one of the two most dominant teams over the last few weeks is out of the question. After ranking second-to-last in percentage spent with 52-percent or more of the total gold with 8.7% during the first half, they improved to 29.8% — and second place behind G2 — in the second half. Furthermore, they have considerably reduced the time they spent with a large deficit (48-percent or less of total gold) from 22.5% to 8.8% (lowest from weeks 5-8). Another indicator of their dominance is their jungle control rate. Splyce has captured 54.1% of the overall jungle farm in recent weeks (1st in EU), compared to only 47.4% before. All this combined with a higher percentage of gold shift events (GSEs) — which are events where large sums of gold are rewarded — won, Splyce is putting themselves in better situations more often and, aided by this, perform better in important situations than before.


Sencux is stepping up


Sencux Splyce Damage
Image courtesy of Riot Games

Looking at the individual player statistics, there is one glaring difference between the halves. Compared to Kobbe, the team is relying a lot more on Sencux’s damage than before. Having, on average, dealt 24.4% of the team’s damage during the first 16 games, he was responsible for 29.9% in the latter part of the split. Kobbe on the other hand saw his share decline from 28.4% to 23.9%, a pretty sizeable drop. While some of this might be due to the change in champion picks by Splyce’s mid-laner, it is also a tribute to his improved play and consistency. Not only is Sencux dealing more damage overall and a larger share of the team’s damage, but he is doing so with only marginally more gold in his pockets (21.8% vs. 22.4% gold share). Kobbe on the other hand is transferring less of his gold into damage, which does not necessarily mean that his play has declined, but might simply be due to an improved Sencux or a shift in team priorities.

Going beyond Wunder’s improvements highlighted in a previous section, he is now responsible for more of Splyce’s deaths (23.5% vs. 28.3%). This is in no way a negative development for Wunder, but rather a result of the team’s overall positive development. He has actually died less than before (39 vs. 48 deaths), but Splyce overall has died even less frequently. His kill-death-assist (KDA) ratio has increased slightly from 2.8 to 3.6.

Mikyx and Trashy highlight the limitations of analyzing the team through hard data. One notable difference is their decline in kill participation, but this is true for every single one of Splyce’s players and cannot really be used against them. (A look at the team’s compositions might give further clues here.) The other big difference is an improved KDA. Mikyx’s KDA was 2.9 in the first four weeks and 4.5 in the last four and Trashy’s 4.8 and 9.1 respectively. Watching the games and listening to commentators and analysts, it is clear that especially Splyce’s jungler deserves a lot of praise for his smarter jungle pathing, better jungle control and greater synergy with his mid-laner.

Still, despite these praises and the fact that all of the members have improved, I think Sencux is the one who has shown the biggest progress during the split and is a major reason why the team is looking so dominant. Overall, the leap the team made cannot be explained by single developments or stats and overall team synergy seems to be at an all-time high. Splyce has proven to be one of the top contenders for a deep playoff run and their confidence, discipline, team-fighting and improved macro-play should put fear into their opposition. Given their recent performances under pressure versus some of the best teams in the region, there is no reason to think that they will slow down any time soon.


All images are taken from here and are courtesy of Riot Games.

Check Also

Featured Image

Improving on DPM — Introducing Average Damage per Minute Difference

Damage per Minute Difference, short DPM-D provides a new damage statistic that has some advantages to the traditional damage per minute (DPM).