Haksal’s MVP Play as Genji: The Value of a Final Blow in Overwatch League

Yong-Cheol "Imt" Jeong
Andbox
Published in
6 min readMar 19, 2021
Genji ‘Did You Miss Us?’ from NYXL

There have been no evaluation of the value of a final blow in the Overwatch League (OWL) so far. However, the value of final blows that happen in a 6 v 6 teamfight and a 6 v 1 teamfight should be weighed differently. This means that the value of final blows can be different even if two players record the same amount of final blows. We can say that the one who records higher value played better than the other.

How can we evaluate the value of a final blow? I believe that it is plausible to say that it is correlated to the change of relative combat power (RCP) resulting from a final blow. I’ve posted about the RCP in my previous post. Again, here is the RCP table of all possible teamfight situations.

All possible RCPs of blue team and red team

To understand ΔRCP, let’s take a look at an example. In a 6 vs 6 teamfight between blue team and red team, the current RCP=0. If a blue player earns one final blow, than it makes 6 blue vs 5 red teamfight, thus the RCP changes to 1.8. In this case, ΔRCP is 1.8 which means the value of a final blow is 1.8.

According to the RCP changes (ΔRCP), all possible values of a final blow is as follows.

All possible FB values

The table above is the FB values of blue team’s point of view. For example, if a blue player gets one final blow in case of 1 blue vs 1 red, than the FB value is 1.0, while the FB value in 2 blue vs 1 red would be 0.5.

It is interesting that the FB value in 6 vs 6 (first final blow of a teamfight) is almost 2, which means that it is as impactful as two final blows. It is 9-fold valuable in case of 6 blue vs 1 red (0.2), presumably because the latter has almost no effect on the result of a teamfight.

Similarly, the ΔRCP in a 1 blue vs 6 red is only 1 (exactly 1.033), despite it being definitely the most difficult final blow to make among all cases. This is because the final blow that makes a 1 vs 6 fight a 1 vs 5 does little to change the result of the teamfight.

Additionally, a Death risk is the opposite concept to the FB value. If a blue player dies in 6 vs 6 teamfight, a red player gains 1.8 FB value while the dead blue player gains 1.8 Death risk. Higher death risk brings higher risk losing the teamfight.

Let’s take a look into some real match data. This is the table of basic stats of the match vs GZC in Countdown Cup.

Basic Stats in Eichenwalde

In Eichenwalde (Map 2), Haksal earned 18 final blows while only dying 2 times, which stands out compared to other players. How would we evaluate Haksal’s performance using FB value and Death risk?

FB values and Death risks in Eichenwalde

Haksal recorded the highest FB value at 26.37 while at the same time having the lowest Death risk at 5.0. A point of interest here is the fact that Nenne recorded a higher FB value (5.67) than BiaNcA (4.67) even though they earned the same number of final blows (4), which means the final blows that Nenne made were more valuable than those that BiaNcA made. Also, Haksal had the lowest death risk with 5.0, lower than Mano (6.8) and JJonak (5.6) who had the same number of deaths (2). This indicates Haksal’s deaths were less significant in regards to the results of the teamfight compared to the other players.

You can find the FB value gained by each hero below.

FB value gained by each hero in Eichenwalde

As I presented in my last post about RCP, Haksal showed an incredible Dragonblade-Super charger combination play in Eichenwalde.

Haksal’s crazy Dragonblade in Eichenwalde

He earned 6.4 FB value in this 4th teamfight in Eichenwalde. This is remarkable.

FB value gained at 4th teamfight in Eichenwalde

Thinking about the FB value and Death risk, I believe the ratio between them can indicate one of the value of player’s performance like Final blows/Deaths ratio (FB/D). (FB value/Death risk) and (FB/D) seems to be linearly correlated as shown below, but the difference between the two is whether each indicator implies the ‘value’ or not of each nominal stat.

(FB value / Death risk) / (Final Blows / Deaths) in Eichenwalde

Let’s take a look at the (FB value/Death risk) between NYXL and GZC.

(FB value/Death risk) in Eichenwalde

Haksal recorded 5.27 in terms of (FB value / Death risk) and this consisted about half of the entire value on the NYXL. Considering that Happy recorded 0.96 and accorded for 2/3 of the entire value on the GZC, it is not strange to say that Haksal dominated Eichenwalde.

However, Haksal’s outstanding performance did not persist throughout the entire match. He peaked at Map 2 (Eichenwalde), but his performance level dropped sharply in Map 3, 4 and 5, due to various reasons such as composition mismatches. While Haksal was struggling, Nenne stepped up. His (FB value / Death risk) ratio took the shape of a V curve: taking a sharp dip from Map 1 to Map 3 and bouncing back up from Map 3 onwards.

(FB value/Death risk) throughout the entire match

Lijiang Tower, the last map, Nenne recorded 1.32 and it was almost one-half of NYXL’s (FB value/Death risk).

(FB value/Death risk) in the last map

Here are the overall FB values earned for each hero played by Haksal (NYXL) and Eileen (GZC).

FB value earned by players separated by each hero

Haksal earned 52.52 FB value throughout the match as Genji and this was the top in terms of a single hero.

As a result, Haksal won play of the match (POTM) in this match, and this is plausible in terms of FB value he gained. He definitely carried the earlier two maps, and Nenne (and also the others) did in the later maps, and these performances won us the match.

  • English Revised by Minseong Kim (NYXL Player Manager)

--

--

Andbox
Andbox

Published in Andbox

Andbox brings people together across the New York Tri-State area to celebrate their passion for gaming with live events, digital content, competitions, & apparel. Home to NY’s professional esports teams & franchises and some of the leading gaming content creators in the world.

Yong-Cheol "Imt" Jeong
Yong-Cheol "Imt" Jeong

Written by Yong-Cheol "Imt" Jeong

Head of Data @NYXL. PhD in Neuroscience @KAIST, South Korea.