On April 26th VALVE updated the Active Duty pool, removing Overpass and bringing back Dust2. This decision has been criticized by most of the community, where the general feeling is that Vertigo should have been the one to go. In this piece, we will take a deep look at the evolution of map vetoes since the release of Counter-Strike 2, and whether Overpass was the right map to be removed.
Eight months have gone by since the official launch of CS2 and, after the initial excitement, the dust has already settled and teams have been able to test and decide which maps are more suitable for each roster. To analyze the evolution in map priorities over this period we will study best-of-3 (BO3) vetoes from 1102 matches, played in 70 different tournaments, both online and LAN, across different regions (North America, South America, Europe, CIS, Asia, and Oceania). Data was extracted manually from HLTV and a brief description of the matches included can be found in my GitHub. In this first part, we will try to answer the following questions:
- How likely is a given map to be either picked or chosen as decider in a BO3?
- Which maps are more likely to be removed in the first ban wave of a BO3?
- Which maps are more likely to be picked in a BO3?
- Which maps are more likely to be removed in the second ban wave of a BO3?
- Has there been any noticeable alteration in the veto pattern over the course of the eight months considered?
- How often does a given team win the map he picks? Does it depend on the map?
How likely is a given map to be either picked or chosen as the decider in a BO3?
Ancient is a mainstay
To answer this question, a common approach is to look at the number of times that each map is being picked or left as deciders overall. This gives a good general understanding of how popular certain maps are among the top-tier teams. In our database, the most picked map is Ancient by far, with Anubis following behind. Mirage, Nuke, and Overpass are the next three, with Inferno and Vertigo being considerably the two least popular maps to be picked by either team in a BO3.
When it comes to deciders, Ancient appears again as the most popular map, almost doubling the rest of the map pool. The next most popular map is Overpass, followed closely by Anubis and Nuke. Inferno and Mirage appear slightly after and Vertigo is, with a fairly high distance, the least popular map to be chosen as the decider.
This information confirms what most of the community has stated in the last few weeks: Overpass was way more popular than Vertigo, which should have been removed instead. However, this approximation presents some issues that we must address.
Let’s focus on each phase of the map vetoes, starting with the first ban wave. In that case, it does not matter whether a map is vetoed by the first or the second team, what really matters is which maps are removed and, therefore, not available to be picked. As a result of this, we should describe the first bans as the number of matches where a given map was vetoed in the first ban wave, divided by the number of matches. This will result in a percentage, which will add up to 200% since there are two maps vetoed initially.
When it comes to maps picked, things get tricky. The usual approach used in this section is to represent the percentage that each map was picked by either of the teams, dividing the number of times each map was picked per the number of matches, and this is where the devil lies. Since we have already undergone a ban phase, the pool of matches where each map can be picked is not the same for all maps, and that affects the real probabilities of each map getting picked. Therefore, the percentage of matches where a map is picked has to be estimated as the number of matches where a given map is picked by either one of the teams, divided by the number of matches where the given map was not vetoed in the first ban wave.
Consequently, when we want to estimate the percentage of bans for each map in the second map wave, the pool of matches is reduced to matches where the given map was neither banned in the first ban wave nor picked by any of the teams. Having all of this in mind, we can start answering our next set of questions.
Which maps are more likely to be removed in the first ban wave of a BO3?
Nobody likes Vertigo
Studying maps banned in the first ban wave is a good proxy for understanding which maps are being permanently banned more frequently since teams tend to keep their first ban stable and focus on a six-map rotation. As expected, Vertigo is the map banned more frequently, being removed early in 40.74% of the matches. Ancient and Overpass are the least banned maps, being removed only in 14.52% and 16.06% respectively. The remaining four maps present similar ban rates: Anubis is banned in 32.76% of the matches, Inferno in 29.85%, Mirage in 36.48% and Nuke in 29.58%.
This data indicates that almost half of the teams across all regions had Vertigo as their permanent ban. In fact, the percentage may be even higher if we take into account that once the first team has removed Vertigo, if the second team does not play Vertigo either they can use their ban on a different map. On the contrary, Ancient and Overpass were the more common maps in the map pool for elite teams, which may indicate that most teams will have to re-evaluate their map pool with the inclusion of Dust2.
Which maps are more likely to be picked in a BO3?
But everyone loves Anubis
This section reflects on which maps teams feel more comfortable, once the initial wave ban has removed 2 options out of 7. As we can see in the figure, the differences between maps are not extremely high. Ancient (44.80%), Anubis (45.34%), Mirage (43.71%), Nuke (40.72%) and Vertigo (41.04%) are picked more than 40% of the time. Inferno comes next, with an 8% decrease sitting at 34.28%. Finally, Overpass is the least picked map when available, with only 31.46%.
Combining the information from the first section with this section, we can see a clear difference between Ancient and Overpass: both are rarely banned in the first wave, but while Ancient is one of the most picked maps, Overpass is the least picked when available. This may suggest that, even though Overpass is in most team’s map pool, it is far from being at the top of their comfort zone. One possible explanation could be that if you picked Overpass, you are starting T side on one of the most CT-sided maps in CS2.
Which maps are more likely to be removed in the second ban wave of a BO3?
Overpass is gone now, but this may be partially why
Lastly, we would like to know which maps are more likely to be discarded as deciders. These maps are at the bottom of the team’s map pool and, even if they are not perma bans, teams prefer to avoid them. In the following figures, we can observe that Overpass (73.82%), Vertigo (76.88%), and Inferno (73.82%) are removed in more than 70% of the matches where they are still available. Ancient (48.46%) is by far the least banned map, with being chosen as the decider in more than 50% of the matches where it is still available. Mirage (68.02%) and Nuke (66.09%) are frequently removed, with around 65% of getting removed for both. Finally, Anubis is the second least banned map with almost 59.51%.
Finally, with the results included in this section, we get a good general understanding of why Overpass may have been one of the better maps to have removed. Most teams do not remove it in the first ban wave, but it is one of the maps with a lower probability of getting picked by any team when it is available. In addition, it is removed most of the time in the second ban wave. Vertigo presents similar values for ban waves but is picked more often when it is available. Therefore, teams with a strong Vertigo can take advantage when the other team does not remove it in the first ban wave, whereas Overpass is just waiting to be discarded sooner or later.
Either way, both Vertigo and Overpass were the clear candidates to be removed on this basis, and it was obvious that Inferno needed some work to gain more priority in the map pool. The results in this section contrast with our initial idea that Overpass was a more popular decider map than Vertigo.
Has there been any noticeable alteration in the veto pattern since Counter-Strike 2’s release?
There are some takeaways
For this section, the idea is to split the matches included per month and study if there is any appreciable difference. The final dataset includes tournaments from 8 consecutive months (October 2023-May 2023), but since ESL Pro League S19 is the only tournament left in May playing with the old map pool, I made the decision to include those matches in April. This leaves us with a final timeline of 7 months. November is the month with the highest amount of matches (269), and December is the month with the least amount (102) corresponding with the winter break.
Starting with the first ban wave, the percentages seem fairly stable, with Vertigo presenting a slight downward trend in the last 5 months. This suggests that Vertigo was slowly being removed by some teams as their permanent ban, with FaZe being the most notorious example that comes to mind. On the contrary, Mirage seems to be in an upwards trend, surpassing Vertigo as the most banned map in the first ban wave over the last two months. Finally, it is noticeable as well that Inferno started CS2 with a marked increasing trend, but started decaying since February.
Moving to maps picked when available, two maps presented a clear downward trend: Inferno and Overpass. In addition, Anubis and Ancient were following a more moderate downward trend as well. Vertigo is the map that has seen the biggest change, duplicating the percentage of times that is being picked when available since December. Mirage and Nuke seem to be following an upward trend in the last couple of months as well.
Lastly, maps banned in the second veto rounds present visible patterns as well. Vertigo is getting more popular as a decider, with Anubis and Overpass being the maps more affected. This matches the results that we have been seeing in the previous parts.
To sum up, we can confidently say that the map pool has been evolving since the launch of CS2. With several big updates, and teams getting more time to practice and test maps in officials, we have some winners and losers. Overpass and Inferno needed some changes since they were less popular map picks and more frequently removed as deciders. Vertigo is the clear winner, being banned initially less frequently, picked more often, and appearing more as a decider.
How often does a given team win the map they pick? Does it depend on the map?
Anubis is surprising
After studying what maps are more likely to be played, we ask ourselves if some maps are more likely to be won by the team that picks into them. Overall, 53.18% of the maps are won by the team who chose it. This may seem problematic, but it is important to notice that whenever a team wins a BO3 2-0, the percentage is 50% since they are winning their map and their rival’s map, whenever teams swap victories in each other maps the percentage is 0% and whenever each team wins their map the percentage is 100%. If we assume that all results are represented fairly even, we would expect the percentage to be around 50%.
In terms of which map has the best win rate for teams that pick into it, the differences are relatively small. The best map is Nuke, with a 56.01% win rate, followed by Vertigo (54.85%), Inferno (53.96%), Mirage (53.59%), Ancient (52.37%), Overpass (51.20%) and with the worst win rate, Anubis (50.89%). This differences may contribute to explain the change in map veto trends that we described earlier: Anubis is the least reliable map and teams could have decided to avoid it as their pick, and Vertigo is the second best, which may explain why it is increasing its popularity.
It makes sense why Overpass left the map pool
Though that doesn’t mean we don’t miss it
We have studied several aspects of map selection in BO3s. Initially, we delve into which maps are more popular, trying to understand if Overpass was the right map to go. In that regards, we have to decide which aspects give CS2 a more interesting map pool. Overpass is a long, complex map that was used usually as a decider and was in the bottom part of the map pool for most teams. On the contrary, Vertigo is a fairly plain map, usually removed in the first ban wave, but it offered a greater advantage in terms of map pool to those teams able to master it.
Moreover, Overpass and Anubis were in a decreasing trend in terms of maps picked when they are available, while Vertigo, Mirage and Nuke were trending upwards. These tendencies could be explained partly by the differences observed in win rates for teams that pick into a given map.
There are several concerns about the data used that have to be addressed. First of all, I have tried to select a fairly heterogeneous set of tournaments to avoid data being biased by certain teams. However, it is obvious that certain teams that tend to go further in tournaments will play more matches, and therefore influence more the tendencies. This is partly avoided by including closed qualifiers to main events, where top teams do not play, but the level of the matches remains high with teams playing for a good incentive. Moreover, since each region has its own qualifier, we can get more data from teams of less common regions such as Asia or Oceania.
In the second part of the article, we will focus on the differences among regions. Which maps are more popular in each region, what win rate when two regions face each other, and several more interesting ideas. If you have any new questions leave them in the comment section below and we will try to answer them in the next piece as well.
