In the penultimate part of this short series looking at maps in 2017 I’m looking at team trends in maps played. I’ll be looking at the top rates teams in our Glicko world rankings which includes the top NA team Cloud9, and also taking a detour to see what Virtus Pro have been up to.
There are some difficulties in separating interesting data from the mass when examining these trends. It tends to produce spaghetti line graphs which obscure trends.
After kicking the data around for a while I settled on an analysis method that allowed for clear representation of the map usage by teams and the identification of some more interesting trends or changes over the course of the year.
The best way to describe it is by looking at an example.
The blue bars are the monthly average share of all the maps played by Astralis. The orange is the month to month deviation of the played rate. In CS there is naturally a fairly high deviation, so the Inferno and Train ratios above are very consistently picked maps, cornerstones of the Astralis map pool.
None of the maps in the conversation for rotation out of the pool are a significant problem for Astralis, they are all pretty disposable.
The ones that are potentially more interesting are the maps where the deviation is much higher, and they might tell a story about how the team’s map pool has changed over time.
As Astralis’s leadership has been pretty stable over the course of the year there hasn’t been much in the way of wild fluctuation of their map pool. Nuke has moved around quite a bit with no particular pattern and continues to be a map they will play at a low level.
They played Overpass heavily around the mid year Major but it wasn’t a cornerstone map the rest of the time, although as the year has passed it’s probably trended upwards for them overall. Cache bookends their year, getting no play at the same time Overpass was played heavily but like Nuke it’s a map they find themselves on from time to time.
Despite having a major roster revamp in the middle of the year Faze’s consistency in their map pool really isn’t shaken that much. The only map that has a really high deviation is the irrelevant Cobble which they effectively never play. They don’t have as obviously strong preferences as some teams, although Mirage, Train and Overpass are obviously consistent staple maps.
The story here is the decline of Nuke in the Faze map pool. It’s obviously something they played consistently but have moved away from since the roster shuffle. Cache and Inferno remain options they can play but neither are particularly important for the team.
Like their counterparts at the top of the world rankings in FaZe, SK don’t seem to have much of a map pool because they’re so willing to play a wide variety of maps most of the time. Nuke is their go-to ban and they seem quite happy playing Cobble in contrast to some others, and Cache seems to be a map they’ll allow themselves to be pushed into playing from time to time.
Given the low variability of maps for SK Nuke is included just to prove they’ve had to play it at some point. Cache was something they played quite frequently in the run up to the PGL Major then rested, and it’s becoming more prevalent again now – perhaps it will be something they use again at the next Major. Similarly Overpass had a late summer/early fall hiatus but is back in the rotation.
G2 do not do things in the same way as the other highly ranked teams in our sample. Cache is a staple, Train is not a regular map and while others sometimes seem to play nothing but Mirage it’s not the case for G2 at all. Nuke is a pretty popular map for them.
The potential removal of one of Cache, Cobble or Nuke will rob G2 of a map they tend to play, although Nuke will have the least impact of those.
The obvious story here is the death of Train as a map G2 are willing to play from being one of their top choices, and the emergence of Mirage as an option albeit a very restricted one. It will be interesting to see if they’ve cooked up any surprises around their map pool for the next major.
Nuke has jumped around a lot but remains an option. Should Cache or Cobble get removed from the map pool then either Train will have to come back for G2 in a big way or the flirting with Mirage will have to turn into a full commitment.
Cloud9 are another team that’s had a roster shuffle recently and changes in the focus of their map pool have produced a fuzzy picture here, only Mirage continues to be a rock solid consistent pick for them in a period of change.
Even so there are signs of orthodoxy here with Train and Inferno both popular maps albeit inconsistent. When it comes to the imminent map rotation the removal of Cobble is probably the biggest threat to North America’s strongest team.
Overpass starts the year as a key map, is completely removed from their pool and has gradually made a comeback towards the end of the year. At the same time Cache has gone from a regular stamping ground to an after-thought. Like so many other teams Nuke started out the season as at least a semi-regular map but has become a perma-ban.
I wanted to take a look at VP because with a major coming up they probably become the most dangerous underdogs in all CS. They also have a pretty fuzzy map pool. Like most top teams they play plenty of Train and Mirage, and like G2 they have an unusual preference for Nuke. Overpass seems to be the only map they strongly avoid, but even then it gets more attention than most teams’ least played maps.
The story from their more variable maps is that there is no story beyond the chaos. Only Cache shows any sign of a trend, they didn’t play it at all in the spring but it’s re-entered their played maps in the second half of the year.
They’ve played a lot of what could be considered throwaway tournaments since then below the level the team is capable of. In search of some kind of pattern for VP I looked at the months when they’ve played well at a high level event this year – the July PGL Major and the October EPICENTER tournaments.
Although there are some signs of orthodoxy with Mirage and Train featuring reasonably prominently these still present a mixed picture. The big change from July to October is Inferno but there’s no particular sign of that being a long term trend. VP look to be a genuinely unpredictable team that will play almost any map. In that sense they’re not unlike the heavily favoured top dogs like SK and FaZe.
Although I hadn’t planned for a part 4 this has gone on so I’ll be putting something extra together with some leftover observations.