Matt Leung-Harrison@RiotPhroxzon
MSI Thoughts
Congratulations to HLE for winning MSI! This is one of the first times (at least that I can remember) that the first seed from each of the major regions made it to the top 4. Perhaps signs that the regions are all a lot more competitive with each other. The gap is closing?
The major talking point of the tournament was Mages bot, allowing teams to play a weakside bot strategy and send their support to roam around the map. At the end of the tournament, there were ~20% mage bot picks, primarily Mel and Ziggs (can survive easier by themselves) while they enable a roaming support like Pyke, Rell, Camille.
It's worth noting that there is a distinct difference in the types of Mage bot play we observe in regular play and the ones that we typically observe in Pro Play. The ones we observe in Pro Play are predominantly used to unlock a roaming support, whereas the ones in regular play are much more about playing something like Brand/Swain + engage support to kill you or pairing a bot like Seraphine/Ziggs with another skillshot support to throw a lot of abilities and if one lands, following up with chain CC for a kill.
My current take is that the Pro Play portion of Mages bot are in a decent spot; they are not at the level of ubiquity that we would really take major action here. For regular play, accounting for the high skill floor, winrates, level of presence and their mastery curves, I could see an argument for a slight tap down in some situations. That said, all the Mage bots are not equally powerful, but we do want to continue supporting a wide variety of picks here. It is desirable for bot laners to have a variety of different experiences when playing bot which should include things like Mages bot, melee carries, etc. as those spikes of intensity + variety are some of the things that keep League fun and fresh for the long term.
Another major talking point was the ubiquity of roaming supports, which was quite nice to see, given the meta shift at the start of the year being dominated by ranged, laning dominant supports. While we made some changes to reduce the effectiveness of these with role quests, the impact here seems pretty limited to Pro Play and making major adjustments here may end up having a disproportionately negative impact on regular play to solve a problem that may not be that much of a problem (or can be solved in other ways).
The last notable callout is the prevalence of both Hextech Rocketbelt and Sundered Sky at high levels of play. Hextech Rocketbelt we think is a bit too ubiquotous on immobile mages in particular that removes a core weakness from them (that they can be engaged on and killed). That said, it's relatively limited to Pro Play and does not find the same level of success in soloQ outside of the very highest levels. Regardless, we're still planning on adjustments that make it less effective on those users, but retain effectiveness on the melee champs that the item was originally designed for, especially because the core fantasy of rocketbelt is not really one that is strongly resonant for this use case (primarily rocketbelting for escaping).
Sundered Sky has been a strong item for a long time; and arguably too strong for too long. The Sundered Sky spike on 1 or 2 items in particular often results in champions being too heavily statted at those item breakpoints making it hard for other champions or subclasses from interacting with them. That said, a significant number of champions in the game purchase or have been tuned around this item, so we will approach the investigation and follow up balance carefully. It's pretty notable that the MSI jungle pool was quite heavily stacked with AD junglers, many of which use Sundered Sky.
Will of course discuss these more with the team when getting back to the office though as there are a diversity of viewpoints on these topics.
