The Mid Season Invitational of 2016 has so far defied all our expectations. SKT was mediocre, NA was far better than EU, and RNG somewhat figured out Euro style lane swaps. Yes, I want to talk about all the switcheroos we did not see during the group stage at MSI. I went ahead and noted the approximate position of every ward placed by teams before 2 minutes (Riot does not make stats like ward position available, this was fun). I mapped them on Summoner’s Rift and tried figuring out why we saw more and more successful swaps as the days went by. You can find the map below. Each point on the map represents a ward placed. You can filter for the side, matchup, or team you want on the right side of my visualization. I also wrote short summaries on interesting situations for each team.
Many who follow the Western League scene can appreciate lane swaps on a strategic level. However, they make for a sometimes tedious first 10-15 minutes of the game. In the NA and EU LCS, teams would invade deep into enemy territory on opposite sides of the map. Instead of splitting the map at the river, they would basically split it along the mid lane until their comps reached specific power spikes. Lane swaps are largely based on the information a team can gather in the first 2 minutes of the game. If a team notices anything after that it means it already screwed up somewhere. Here is what Summoner’s Rift looked like at the end of day 5 at MSI. (Unfortunately, if you are reading this on your smartphone, you will see and image instead of a map with filters)
As we can see, the large majority of wards was placed throughout the river in a style similar to what we know from solo queue. When professional teams placed wards like this, their lane swaps largely came down to intuition and maybe some VOD-reviews. Remember, most teams had very little time to prepare for MSI after playoffs ended. So for the first few days, teams either guessed wrong, or more likely, overwhelmingly preferred to simply see where they are at in terms of raw skills and head-to-head matchups. In the larger scheme, this might point to a lack of overall international competition throughout the year. You could also say that every participant stuck to what they already knew. NA’s CLG is the only team that attempted deep invades more than twice. G2 and RNG did it twice and SKT once. As the round-robin went by, teams succeeded more often in their swaps. Working of the information they gathered through days 1 and 2, they managed their swaps differently and looked for favorable matchups. This is why we saw only one successful lane swap based on precise information over the first two days. Throughout the rest of MSI, teams switched around creatively ten times and started tweaking their early game movements, just like they adapted their picks and bans. These are all general trends, so let us break it down on team-by-team basis.
Counter Logic Gaming
CLG stuck to their guns when it comes to macro-play. They initiated all but one deep invade to get a lane swap. They did this versus G2, RNG, SKT, and SUP. It is reasonable to assume that CLG think this is the optimal way to play the game right now. But it was not always the team to move for a deep invade first. Both SKT and G2 start off against CLG by deep invading in one game. Clearly, CLG likes to switch it up, but is adept enough at reacting to what other teams do in the early game. There are the exceptions of Huhi and Stixxay. Both of them were caught out multiple times when teams came from unexpected angles. Stixxay had to burn flash against SKT because Faker, Wolf, and Bang ganged up on him in CLG’s tri-brush. Faker was missing from mid lane, but Huhi pinged way too late. Huhi himself on the other hand almost got killed by RNG when Mata and Wuxx swung around CLG’s red buff. While they could basically autopilot their way to the first wave in the NA LCS, this was not the case at all during the Invitational’s group stage.
If we ignore the cost that came with G2’s vacation, they actually did not behave very differently over the first 120 seconds of the game than any other team at MSI. They mainly warded the entire river in a defensive way. There are some things that went wrong for them (aside from losing). First of, Perkz was caught out on Day 2 against RNG and almost gave them a kill. This set his Ryze behind in lane quite heavily. The outcome for Hybrid was even worse. He gave CLG first blood 98 seconds into their meeting on Day 2. This time CLG was the team switching things up by sending their botlane through mid, where Perkz recalled at the minute mark. He had already placed his ward far up in mid lane towards CLG’s blue side. So aphromoo and Stixxay snuck through together with Huhi and killed Perkz, who also had no business walking from top to mid at that point in time.
The Flash Wolves surprised many when they won against SKT, but this turned out to be SKT underperforming heavily over the first few days of the tournament. FW was one of the teams to actually almost never invade. The only notable exception was against Super Massive’s Achuu when they tried the kind of invade Bang, Faker and Wolf did against CLG earlier. However, they did not time it right with Maple. NL and SwordArt walked out of botlane into SUP’s tri-brush too early and their gank only led to a blown Flash by Dumbledoge instead of Achuu. Dumbledoge could have even saved his summoner, but he had no information on who else was there from FW. Another notable exception was against RNG. FW started on blue side and set Looper pretty far behind to start of that game by hiding their intentions from RNG. They probably used what scouting RNG had done on them to their advantage here.
Royal Never Give Up
RNG did shine when it comes to teamfights and skirmishes as we expected. Despite their solid performances throughout the group stage, they did show a lack of experience with lane swaps. Even though they exploited Super Massive’s lack of sparring partners, they were badly outsmarted by CLG and FW. In the first game of MSI, CLG invaded deep into their blue side jungle and built a large lead before throwing the game on multiple occasions. However, they showed progress in their second meeting on day 4. CLG coming from blue side warded deep for a lane swap. RNG was content with simply clearing CLG’s red and raptors. Now comes the real tweak. Mlxg did not join Mata and Wuxx to push bot tower. Instead he ganked mid and got an easy first blood with xiaohu’s Zed on an overextended Huhi. Against FW, RNG tried a Euro swap and failed miserably. They took 25 seconds longer than FW to secure their first tower. In the meantime, the Wolves did not back after they got the bot outer turret. Instead, they swung around to midlane and killed xiaohu’s Ryze at almost the same time for first blood. This could have happened to CLG or G2, however both Western teams would very likely not have been as far behind, thus limiting FW’s window of opportunity.
The start to MSI 2016 went according to plan for SKT. But after losing four games in a rowthe LCK-champs looked out of sorts. No one knows exactly what was going on there, so let us focus on what we could observe. One trend from day one was that Bang and Wolf attempted late invades on their own to mess with the enemy team’s duo lane at gromp or golems. However, this was not really based on vision prior to the invade. It seems likely that they simply wanted to test their competitions nerves. Apart from this, SKT also tried creative invades based around their enemies’ tendency to simply stand in tri-brush when starting on blue-side. SKT did initiate one deep invade against CLG at their second meeting in a game SKT won pretty handily. However, they guessed wrong and actually did not get their Sivir the time to scale into late game the way they wanted. So their hustle did not even result in what they had intended. This might be attributable to CLG simply knowing what SKT wanted and making the right call.
The Wildcard-team put up a good fight. However, several of their tendencies were exploited by their opponents. First of, SUP tend to ward defensively very late into our 2-minute window. Even more than other teams, they simply did not want to get invaded on while doing their camps. Dumbledoge got killed for not respecting the time RNG had on day one to roam towards dragon after they had warded their top side jungle. It was again Naru who placed his ward into the brush by dragon way too late. At this point, four members of RNG had already roamed down. Achuu’s ward in tri-brush got cleared by Mata and Wuxx. Dumbledoge simply did not know which RNG members were there and died for the lack of information. Antoher example was in their last meeting with RNG. Starting on red side, Stomaged got caught out on Elise at his own red buff. Coming from their blue side, Mlxg, Mata, and Wuxx simply walked up into mid lane and up to red-side raptors. Naru had not warded there yet and got pressured by xiaohu. So they snuck into the river from mid and were able to simply kill him because he was up too far. SUP’s jungler got punished 40 seconds into the game for the entire team not placing their wards about 5 seconds earlier on average. This may point to two things. First, Wildcard-teams still lack practice partners to develop a more diverse skillset of strategies. Second, maybe the lack of better competition led to SUP relying too heavily on mechanical skill and not information, especially in the early game.
This is of course a rough sketch and there simply is not enough data to take this information as evidence. But it was interesting to see how teams tried creating advantages for themselves before they knew their opponents and after they had some time to play against them. The better teams were able to adapt over the course of the tournament and were thus rewarded with better results. You can look through the entire group stage in the graphic I created above. You can filter according to sides, matchups, and teams.