How competitive teams deal with balance patches and their impact on competitive diversity
Article Written by Ruben 'Daisyx' Korte
As you all know, League of Legends changes every 2-3 weeks. These patches are often designed to balance the game. While some particularly severe nerfs do affect aggregate solo queue statistics, most of the time, players in solo queue won’t even notice them happening.
Because of this, the patching is mostly done to affect competitive play. Because in competitive play the difference in individual player skill is much smaller, drafting (as I outlined in my previous article) and thus the strength of picks, is far more important. Due to this, rather minor changes to a champion can have a large impact in how desirable it is to pick the champion. This is especially true for S-tier champions, whose ability to be blinded in every situation often relies on them having decent matchups into all opponents, which can drastically change with even minor nerfs.
As a result, every patch, teams (and most particularly coaches/analysts) scramble to figure out what is good and what isn’t. Usually, the timeline of a patch and teams adapting to it goes as follows: on Tuesday/Wednesday the patch comes out. At this point, the coaching staff looks at the patch notes and looks at reviews of community figures of those patch notes (most notably phreak and LS) and then sits down and thinks about which champions will go up in (relative) value.
After this, the coaching staff evaluates for themselves what they think the big winners and losers will be. Based on this, they will have meetings with individual players to gauge their opinion to the changes and come up with a list of champions that they should try out in solo queue to see how strong they are. Meanwhile, teams will continue scrimming on the tournament realm as they often have competitive matches on the previous patch on the weekend after the new patch happens. After these matches, teams will get together and decide, based on the players experience with the aforementioned champions and sometimes (Korean) soloq scouting which champions will be played in scrims.
Following that, the usual process of refinement happens and players and coaches see which champions work in scrims and which don’t. In LCS, most teams scrim maybe 2-3 different teams each week, so then they make their own little meta based on their scrim results. 1.5 weeks after the patch hits, the first competitive games will be played on it. After this, people will often copy what works best and after that will copy the Koreans when they play on a new patch.
This process repeats itself every 2-3 weeks. Sometimes patches bring large changes (like 5.4 with the new banner and a lot of item changes) and other times changes are relatively small. Having said that, sometimes small changes to champions can have far reaching consequences - because if a champion that gets nerfed is the counter to another champion, then that champion in turn becomes more valuable, which can have a ripple effect across other lanes, too.
These ripple effects are often very hard to predict and tend to be ignored. This leads to the meta being very much based on what Riot buffs/nerfs, since patches are often the starting point for coaches’ discussions. Besides that, there is little room for innovation with people trying out picks that aren’t being changed, since the amount of time that pros have to test out new things in solo queue is very limited. Even if you are a great coach/player and are able to accurately predict that a non-meta pick can be meta, having your players try it out just isn’t worth it since you will only be right a very small amount of time and all the time a player spends picking up a new champion in solo queue (usually you need ~5 games to have decent mastery of a champion to see how good it is) is time that they can’t spend practicing regular picks.
This in turn leads to a relatively stale meta where the only changes in picks happen because of Riot changes, which usually leads to the same picks being cycled in and out of competitive play because Riot balances based on what is played in competitive.
I have often heard people suggest that Riot increase the time between patch cycles, especially after week 3-4 of LCS since often by then the horribly overtuned champions which arrived from the mid season or pre-season patch have been nerfed. This would allow teams to spend more time on the same patch, which would increase their mastery of the game, but also give them more room to experiment with uncommon picks, which would in turn benefit teams that are able to efficiently and creatively experiment. Similarly, this would allow teams to better define their own style because different teams have different gold/resource allocations which would then allow them to tailor their picks on that. This would also lead to more diversity between different regions as teams become more confident in their own style and test it more in scrims as opposed to just blindly copying what is being played in Korea. This would lead to a broader pool of champions being played across the world with the occasional introduction of (old) new champions, which would also enhance the viewing experience.
A good example of this hypothesis is the Season 6 World Championship, where SSG brought out the MF support (in combination with Ashe ADC) pick as a counter to Zyra, with MF being a champion that nobody had played in a very long time, especially not as a support.
As the above example shows, a time when this could happen is the Worlds patch, where the best teams in the world scrim and play each other for ~2 months on the same patch. To see if this is indeed true, I have compared the pick rate of 8 different moments: 3 different weeks within Worlds 2016, 3 different weeks within Worlds 2017 and for comparison’s sake, I have included data from week 1 and week 9 of the spring split of 2018. To assess the diversity, I have taken the total number of picks played in a certain week and divided that by the number of unique champions that have been played at least 1 time during that week. The lower this number is, the higher the competitive diversity is.
This data has been pulled from gamesoflegends.com and for the games from 2018, only games from EU/NA LCS, LCK, LPL & LMS have been used.
So what does this data tell us? Honestly, not much, because of the many difficulties in comparing the data - most notably between years, but also due to the lower number of matches in weeks 3 and 4 at Worlds.
To summarize, outside of the slight decrease in champion diversity that happens from weeks 1 to 2 in both Worlds, there are no patterns to be found. The one thing that surprised me the most was that there is barely any increase in competitive diversity between Season 6 and Season 7, which was when the 10-ban system was introduced. On average, the number of different picks increases by 8, some of which can be explained by a few of the new champions that were introduced between the two events.
Overall, it seems like the argument that longer competitive patches will lead to teams trying out more different picks and thus increasing competitive diversity doesn’t hold true. However if this effect does exist, it might be off-set by the fact that the more competitive games there are on a certain patch the more teams will copy each other.
If you enjoyed this article, follow the author on Twitter at @PNG_Daisyx.
Image credit: Lolesports Flickr.