R6 Siege analysis: in for the win
As mentioned in my previous blog post, CANA’s Inaugural Esports Tournament Lessons Learned From The R6 Showdown, we hosted our first annual esports event, the CANA R6 Showdown, this past April. The initiating spark to host an esports event grew out of our interest in supporting the esports industry through the use of data analytics. This event gave us a controlled data sample to assess our capabilities applied to esports. The selected game for the event was Tom Clancy’s Rainbow Six Siege.
The tournament gave us ample amounts of data. Upon analysis of the data retrieved from the gameplay, we decided to create our own metrics we believe offer critical information for players and teams. The three analyses done on the collected data were:
First Elimination Correlation to Round Won Analysis
Impact Kills Analysis
Bomb Site Analysis
We see the First Elimination Correlation to Round Won and Bomb Site Analysis as interesting and fun analyses aimed at engaging the fans, although they do offer critical information for the players and teams. The most substantial and informative analysis done was the Impact Kills analysis. More details on each analysis are provided below.
First Elimination Correlation to Round Won Analysis
This was a quick analysis of how often the team that got the first kill of the round ended up winning that round. In traditional sports, there are similar analyses done; an example is the determination that when team A scores first, it will go on to win 85% of the time. Many similar analytics can be applied in esports. For the R6 Showdown, we broke down the first kill correlation to rounds won by each game, each team, and then, overall.
After going through each game, we took a look at the first elimination/win percentage per team. The top tier teams from our event were in the 70th percentile range: NewHorizon Esports went 7/9, 77.78%, US Army Esports went 14/20, 70%, and our champions, Psych Ward, went 21/27, 77.78%. Combining all rounds for all games and teams, the team that got the first kill of the round won 45/67 rounds, 67.16% of the time. This is an excellent metric to pull from esports for general interest and information, but there are many more metrics to discover.
Impact Kills Analysis
Impact kills (IK) are exactly what they sound like: eliminations that had an impact on the round and game. Oftentimes, there are kills that have no impact on how the round plays out. For example, in a scenario where one player is left against four opponents and that solo player gets a few kills but inevitably dies and loses the round, those kills virtually mean nothing. In these situations, it is very common for the team with the major man advantage to use those extra bodies as markers to reveal the location of the last opponent. The team tends to rush into that area because it knows the one player can't hit all four opponents at once. The team sacrifices a few players to win the round. Consequently, those kills the solo player got really have no purpose or impact on the round.
This is why we wanted to create criteria that could measure these IKs. Measuring each team and player impact kills shows who the most valuable player was in each round, game, and team, from an elimination perspective. The criteria used to measure IKs are as follows:
First Kill (FK) - The first kill of the round gives an immediate man advantage.
Kill Even (KEV) - A player eliminates an opponent to even the number of players on each side. This evens the playing field instead of a team being down by two players.
Kill Advantage (KADV) - A player eliminates an opponent to give his/her team the man advantage in the round.
Trade (TRD) - A player eliminates an opponent very shortly after that same opponent eliminates his teammate. This essentially makes the opponents kill a wash.
Kill on Bomb Carrier (KBC) - A player on defense eliminates the bomb carrier. This is extremely important for the defensive team because once the bomb carrier is eliminated, the attacking team has to retrieve the bomb, but the defense knows exactly where the bomb is located. This means the defense can just continuously watch the bomb and wait for the attacker to try and pick it up. Without the bomb, the attacking team has to rely on eliminating all defensive players in order to win the round.
Kill while Planting the Bomb (KPB) - A defense player eliminates the opponent in the act of planting the bomb. This is crucial because it stops the attacking team from planting the bomb, thus requiring the defenders to defuse it. Having to defuse the bomb drastically decreases the chances of the defense winning the round, so eliminating the person while planting the bomb is extremely impactful.
Kill that Led to Victory (KLV) - These are kills that directly resulted in the round ending and that team winning the round.
Comeback Kills (COMEBACK) - These are kills that occurred in the scenario where there is one player left facing three or more opponents, and he/she eliminates all of them and wins the round (only counted if the solo wins the round).
Ownkill (OWNKILL) - A player kills their own teammate. This is impactful because it gives a major advantage to the opposing team and it did not do anything. It gives the opponent a free advantage.
Double - These are not double kills, but refer to kills that met two of the above criteria at once. For example, the first kill (FK) of the round could also be on the bomb carrier (KBC). They do not count as their own impact kills, but are recorded in the final impact kills individual player breakdown.
Triple - Triple kills refer to kills that met three of the above criteria at once. For example, player one eliminates player two and it was a KEV, TRD, and KBC. These do not count as their own impact kills, but are recorded in the final impact kills individual player breakdown.
There are situations where eliminations meet the criteria and are technically considered IKs through the metrics, but did not impact the round. In this instance, these kills came late and just prolonged the inevitable loss. These are not counted towards IKs.
Being able to break down individually how many kills actually had an impact on the game, and to detail what type of kill it was, provides great insight to the players and potential scouts on how “clutch” and valuable that player is. It also shows how many times a player was on the receiving end of an impact kill. This is a newly developed metric that is very useful to evaluate the true value of a player. It can be used for recruiting, practice, and strategy purposes, and for overall player improvement purposes.
Bomb Site Analysis
The initial hypothesis was that if the attacking team failed to locate the bomb site(s) during the round’s preparation phase, the team would be more likely to lose that round. The thought process behind our hypothesis was that when teams do not find the bomb site in the preparation phase, they have to use the action phase time to find the site. Starting the round in this way gives no advantage on where to go or the location of the enemy team. It significantly lowers the amount of time the team has to attack the correct areas. When attackers do locate the bomb site in the preparation phase, they start the actual round with a clear path and have the entire round to focus on getting to that area.
After going through the data, this hypothesis was disproved. For the overall event, the attacking team only won 32% of the rounds after finding the bomb in the preparation phase. On the other hand, the attacking team won 50% of the rounds when they did not find the bomb in the preparation phase. This was an extremely interesting discovery and we determined a few potential reasons for these results. The data that the bomb site was or was not located in the preparation phase comes from the game itself stating it. This means attackers’ drones could have been eliminated just before revealing the bomb site, but still unofficially relaying the locations of the enemies and bomb site.
For example, the players controlling the drone could have heard enemy footsteps and equipment being placed in a room adjacent to their drone. This would effectively let the players controlling the drone know where the enemies and the bomb site are without having to actually mark it. It is also worth noting the data may be skewed by competing teams with unequal skill levels. These lesser skilled teams stood no chance when it came down to pure gameplay skill, so finding the bomb site was irrelevant to their success.
This analysis is still important for teams. Even though our findings suggested low correlation between finding the bomb site in the preparation phase and winning the round, locating the bomb in the preparation phase still increases winning chances in the action phase. It is hard to plan an attack in an unknown area. Finding the bomb early gives the attackers a lead on where to go and allows them more time to focus on infiltrating the site.
These three analyses only scratch the surface of what our inaugural tournament’s Rainbow Six Siege data has to offer, not to mention the relevance to the entire esports realm. It continues to fuel CANA’s passion for esports. As the esports industry grows in every way possible, we plan to continue to support the industry through the use of data analytics. Team CANA, providing analytics intel for success!
Jack is an intern with CANA Advisors. To learn more about CANA’s internship program, please contact Ms. Cherish Joosteberns at firstname.lastname@example.org.