ROX Tigers – A Statistical Exploration

This is a guest article by Denis “Peaches18” Beausoleil. He is a professional League of Legends data analyst working with uLOL and NACS teams. Denis is from Vancouver, Canada.

With the LCK Summer Playoffs just finished and the Championship Points race all locked up, we now have two of the three Korean teams slotted for Worlds. So I figure now is a good time to start actually contributing to the discussion about who are the best teams in each region and why. To do this, I will introduce a bunch of stats that I feel are critical in explaining overall team prowess. A bunch of stats you say? Oh yes, we’re going straight to the deep end my little kiddies. Grab your water wings, let’s dive right in.

In the 2016 LCK Regular Season, the ROX Tigers had a dominating presence over the competition with: the highest winning percentage, the highest gold per minute, the highest gold differential per minute, the third highest Early Game Rating, the highest Mid to Late Game Rating, the second highest amount of time with 52% or more gold, the second lowest amount of time with 48% or less gold, the third highest share of jungle monsters killed, the second highest Gold Shift Events won, the second highest Gold Shift Events won when significantly ahead, and the fourth highest Gold Shift Events won when significantly behind.

I can see it now. Your eyes have glazed over and your hair is blown straight back.

I got you fam, let’s break it all down.

Every great team has an identity, a style of play which shapes that identity, and the ROX Tigers are no different. When we think of them  we think about excellent laners throughout the team and their jungler “Peanut” who invades the enemy jungle looking to apply pressure through skirmishing instead of warding. The term Battle Ward Peanut makes people think the ROX Tigers get these huge early leads and then dominate from there. And ya, at times that happens. But their statistical signature is that they aren’t as imposing in the early game as we might think. They get First Blood in 50% of their games, which is average. They get First Tower at 47.5% of the time, also average. Well what about First Mid Tier 1 rate? They must get that objective at a high percentage and then snowball from there, right? Well I wouldn’t go as far as that, they get the First Mid Tier 1 in 50% of their games. Certainly they get First Dragon a lot? Wrong again, they get First Dragon 47.5% of the time.

Don’t you see what’s going on here? No First Mid Tier 1, no First Dragon. The only thing standing between a ROX win and us is a thin layer of gabardine. And that my friends is where it all gets fascinating.

So how do they do it? How do they win so many games if they aren’t dominating those traditional early game stats such as First Blood, First Tower, and First Dragon?

Well, let’s take a step back and as they say, begin at the beginning.


Average early game stats


Let’s use regular season data from standard lane games only, since that’s now our new reality on Patch 6.15. At the 6 minute mark, the ROX Tigers are behind in XP at -82 and only slightly ahead in Gold at +10. At the 10 minute mark, they are now ahead in XP at +65 but behind in Gold at -20. Therefore, a typical ROX game is to be average at the 10 minute mark. They have average XP, average Gold, average First Blood, and average First Tower and First Dragon (which are maybe a bit early at 10 minutes). Moreover, the CS @ 10 battle isn’t as suffocating across all lanes, only Smeb and PraY have CS leads at 10 minutes, Kuro is typically tied (as well as Peanut).

So if it isn’t in the first 10 minutes the next stat to look at is their Early Game Rating, which takes into account factors such as side choice, gold difference, and drake difference at the 15 minute mark. As I said earlier in this post, the ROX Tigers have the third best Early Game Rating in the LCK. Ah, so that’s where they start to make their jump. Average as average gets pre 10 minutes, starting to separate from the rest between 10 and 15 minutes. Watch any ROX Tigers game and pay attention between the 10 and 15 minute mark to see if they start to create a small snowball. Remember though, they are likely still average at First Dragon and First Tower so if those objectives are still up they will only take them at an average rate.

(That last statement may not necessarily be true. I won’t go into an extensive rant but one thing to look into is the timing of their First Dragon and First Tower. As in, maybe they get First Dragon and First Tower less than half the time before 10 minutes and then more than half the time after 10 minutes, giving themselves and overall average rate of each objective.)


Not so average anymore


Now we’ve arrived. The first 15 minutes are over and from this point forward is precisely the time where that small ROX snowball turns into an avalanche.

As I said in the intro, the ROX Tigers have the best Mid to Late Game Rating in the LCK, which is the statistic we use to summarize the remaining portion of the game. The ROX Tigers go from being: average at getting First Blood to having the highest Kills per Death, average at First Tower and First Mid Tier 1 to having the highest Tower Percentage, average at First Dragon to having the second highest Dragon Percentage. And, not talked about yet, they also have the second highest Baron Percentage. Impressive.

Which now brings us to two all-encompassing stats, stats that look at the entire game. I have a soft spot for these. These are: 1) Gold Per Minute and Gold Differential Per Minute and, 2) Gold Shift Events. The ROX Tigers have the highest Gold Per Minute and Gold Differential Per Minute. They also have, barely, just barely, the second highest Gold Shift Events won by percentage, with SKT being just 0.3% better. Put these two together and that is all the statistical argument needed to make a case. The ROX Tigers generate more gold than anyone, they are ahead by more gold than anyone, and they win important Gold Shift Event at a championship level.

(Gold Shift Events as those such events that swing a game in the favor of one team. This stat is the snowball stat. The best teams in every region have the highest Gold Shift Events percentage.)

But how did the ROX Tigers win the LCK? Why didn’t KT Rolster or SK Telecom win? Two reasons, lane swap and gold production.

In the regular season the ROX Tigers won 76% of games with a standard start, the best in the LCK and now clearly an advantage on our new Patch 6.15 where lane swaps no longer happen. KT Rolster started in a lane swap more often than ROX or SKT or SSG but in standard lanes they had a lower winning percentage than all of those Top 4 teams.

As for gold, it is precisely when looking at these two “entire game” stats that tell the difference. In playoffs, SK Telecom literally won 0% of Gold Shift Events when they were behind. Meaning, when they were losing the game by a significant margin they couldn’t come back. This is a far cry from the regular season when they were the best in the LCK. In playoffs, the ROX Tigers had the highest Gold Per Minute and Gold Differential Per Minute. They didn’t have the highest percentage of Gold Shift Events won, that goes to KT Rolster. But, and this is key, the ROX Tigers won 100% of every Gold Shift Event when they were significantly ahead. KT Rolster? Close but not quite, they only won 96% of those Gold Shift Events. One of those gold shift events was…yup, you guessed it.

Game 5. The game has been close all along, gold has been nearly even, objectives are hotly contested or traded. It’s been a great series and a phenomenal start to our final game. And then in the 32nd minute an extended team fight occurs around the mid lane where Battle Ward Peanut goes full-on Suicide Ward Peanut when he, as Hecarim, presses R and goes Onslaught of Shadows right over that Taliyah Weavers Wall to go for the kill onto Fly’s Aurelion Sol, only to find himself effectively cut off from the rest of his team and subsequently die. Two other members from ROX Tigers also die only one for KT Rolster. Both teams take a breather, reset the map, and then find themselves jockeying over position for the Baron where another team fight breaks out in the jungle near midlane and Peanut dies again. Poor Peanut (after the game he looked so upset). From there, KT Rolster rush straight to Baron.

But then Score, one of the best junglers, decides that even though Peanut is dead and therefore there is NO OTHER SMITE on the map, that he will go for that perfect smite to secure Baron. Only, Score miscalculates and misses it by 2 points.

MonteCristo exclaims “Two! Two!”

I guess that’s how you was raised.



League of Analytics

Oracle’s Elixir

Games of Legends