VALORANT: VCT, EU and NA, Operator usage and efficiency

A thought inspired by a combination of too much VALORANT and too much NBA content lately: how efficient are pro team Operator buys in VALORANT in terms of kill conversion (i.e. KPR per Operator holds/buys)?

To start out with: yes, there are problems with KPR for measuring the impact of a buy, since an AWPer could get zero kills and just lock down half the map and be valuable, or could go 1-for-1 pushed every time and therefore be extremely harmful to their team. It’s still an interesting thing to look into, so I’m going to do it.

Stats here were taken for every team in EMEA and NA VCT Stage 2 Challengers as of 20th June (i.e. including the first week of playoff games but not the second week).

Team usage/efficiency

Usage rate is calculated as picks/rounds; this figure can hence theoretically be 100%+ if a team double or triple-AWPed in enough rounds, but that’s unrealistic in practice.

NA as a whole has a slightly higher usage rate — 12.5% against 11.2% on attack, and 36.3% against 34.9%. This probably overall undersells the difference; 100 Thieves’ complete refusal to buy AWPs drags the NA mean down, while three of the four teams alive in the EU bracket make minimal use of it (Guild on defense being the exception).

The numbers on 100T are surprising; even knowing their reluctance to AWP, I wouldn’t have guessed it would be that low.

It doesn’t feel like a total coincidence that the teams playing on worse ping (BBL being the obvious one) are lower since AWPs just can’t deal with high-ping wide-peeks etc.; will be interesting to see if there’s some adaptation necessary after a long/infinite LAN layoff for whoever qualifies out of FPX and M3C.

In general, the more you use an AWP, particularly on defense, the better you’re going to be with it, although the data gets noiser the lower you go (Sentinels have an insane attack KPR, but only played 11 rounds; 100T similarly only played 2).

It did occur to me late on in writing this title that it might be interesting to see how often teams win rounds with or without an AWP, so here’s a quick scan of that:

Note that this is undifferentiated and hence prioritises defensive usage because of the relative volumes (and also I believe double-counts for double-AWP rounds and so on). Most teams are unsurprisingly better holding an AWP, with Sentinels, FOKUS, and M3C as downside outliers, and G2 as a big, big upside outlier (which probably speaks well to what mixwell is doing even though his rate stats have been poor).

What about player stats?

Player usage/efficiency

We put the bar for a ‘qualifying’ AWPer here at a 5% total usage rate; the only AWPer this excludes of any note is BABYBAP on FaZe, who has a 1.13 KPR on 16 rounds (FaZe played 442, more than any other team in either region). This gets us to 35 AWPers across the 24 teams.

Seven of the most efficient ten in terms of KPR are NA, though the data here is a little noisy. In particular, note what’s going on with TenZ and his 17 kills in 6 rounds; I’m not sure what caused that, but my guess would be something to do with weapon juggling with Kanpeki on Fracture. In any case, my general takeaway would be that Cryocells, yay, and Jamppi have a statistical footprint that stands above most of the competition.

Listing the bottom 10 seems a little gratuitious, but for reference, 0.72 KPR was the average total, with KPR slightly higher on attack mainly because of TenZ and corey’s figures. Most of those below that line were not regular AWPers and had low usage; the one worth picking out as interesting is d3ffo, who’s at just 0.58 KPR currently (22 kills in 38 AWP rounds)

What about usage?

yay and Cryo stand out again, and bdog also looks good. JUGi makes sense; bear in mind that FOKUS are 1–4 and hugely reliant on him, so he’s never going to look efficient here. However, we of course we have to ask…what the hell is going on with cNed and tex? I can’t see a data error there unless RIB data has completely failed.

tex might be a juggling case, but cNed clearly isn’t, and it probably speaks to some deeper-rooted dysfunction within Acend. Attack AWP is a heavily situational thing in Valorant, but cNed has made it work historically; there’s an argument to be made that cNed’s ability in that respect was one of the things that set his previous teams and Acend apart in 2021 and put them over the top. To go back to the NBA analogy, he’s an unicorn player like e.g. Nikola Jokic or Giannis Antetokounmpo; his playstyle and skillset is extremely unusual, which can make it hard to fit him into a system, but if he’s built around and harnessed correctly, it can create something that we haven’t seen before and might not be seen since.

However, the game is evolving; and, while cNed still obviously has value — note that he’s still right up there on defense (6th among all AWPers on defense in KPR, behind Johnqt, Kanpeki, Cryocells, yay and Derke) — it seems like he’s not getting away with what he used to on attack, and Acend haven’t been able to integrate him into the attacking plan more generally.

I did run the numbers quickly on K/D for EU out of curiosity (across all weapons); most players are better on defense (only four — Shao, AsLanM4shadoW, AvovA, and Mistic are +0.10 or better on attack, and 38 are -0.10), but cNed was 52nd out of 61 players in the league — 0.89 K/D on attack, 1.23 on defense.

It’s an interesting thing to think about, and perhaps something to look to see if Acend can do something about coming into LCQ.