At first it was only an impression, a feeling. For few months, the matches from my favorite teams weren’t overwhelming me anymore, frags from my star players weren’t giving me the chills. It seemed like something wasn’t right with Counter Strike. Since we try to do good work here at VaKarM, we tried to dig, to find a reason to this feeling. Going through many changelogs, statistics, discussions, a common thread appeared, at least for rifle players.

If the frags seem less impressive, it’s also because there seem to be less headshots than before. But before beginning to explain why we’re focusing on headshots, we still had to confirm or deny the feeling.

Un style de jeu qui impressionne

Thankfully, headshots are one of the rare easily accessible in game statistics, either through eBot or HLTV data. After gathering results from more than 3,160 matches and analysing the scores of 118 players who made more than 250 frags in 2013 and 2014, we noticed a very surprising stat: the average headshot percentage for top players decreased by five points between 2013 and 2014 (note that those results change only a little if we remove the minimum of 250 frags). You’re gonna say that it’s not THAT impressive, right ? Well you’re wrong ! In reality, this result is pretty significant.

We’re talking about pro players’ (or at least semi-pros) performance, those who spend several hours per day training themselves on the game. Their aim skills should then, logically, improve and sharpen. But if it was the case, this increase would have an impact on their accuracy stats, unfortunately not available, and their headshot ratio. But it’s not the case here.   We’re talking about an average, that take into account very different situations and in the end hide many changes.  Thus, on 118 players, only 16 had an increased headshot percentage in 2014. Among these, we have two very big increases that significatively made the average higher : Harts, whose percentage of headshot in 2013 was pretty low (10 points under the average) and smn… banned for cheating since then. Some top players also had a significant decrease like NBK for exemple ! And if ScreaM is pretty stable ("only" a 3 points drop), it’s to the cost of a disappointing ingame efficiency.   We are talking about a figure that should be relatively stable and shouldn’t vary that much. Indeed the percentage of headshot is intimately linked to the gamestyle/aimstyle of a person and should only progressively change over time. Also, it is way less sensitive to other factors (like the positioning on a map, etc…) than other statistics. The K/D ratio for example could really be affected by how your team perform : it's a lot harder to get a good K/D ratio when you always end up 1vs5.

Some examples of the changes of HS % between 2013 and 2014.

For those who are interested in the details of these calculations, this Excel file gives you all the data and different sheets used.

These few calculations show that the initial impression is confirmed by the results : the frequency of headshots is declining.

But why are we giving so much importance to that percentage ? Because without access to more precise statistical data, this one is the only one that permits us to judge the evolution of the aim of a player. Indeed, what better technique to kill than a HS that, except with the M4, kills with one bullet ? All things being equal, the best thing to do for a player is to HS as much and as fast as possible to get rid of his opponents. But the statistics speak for themselves : this rate seems to decrease. There is therefore a paradox between theory and reality.

Contradiction to be explained. But we then don’t stick the facts anymore, to enter the realm of hypothesis. I do not think that the players’ aim has substantially fallen. First of all it would have been surprising that the aim of many players evolve in the same direction at the same time: individual performance declines can be explained, but when it concerns more than one hundred players over more than one year, this assumption no longer holds.

My explanation proposition is different. The top players, who have almost all previously officiated on 1.6 or Source ended up adapting, consciously or not, to CS:GO and the various updates made by Valve. Assuming top players look for  efficiency in the frag, the logical conclusion would be that the HS is not the most effective way anymore. If at the beginning of CS:GO some gameplay elements like aimpunch were included, some updates have reinforced this phenomenon :

• 2nd October 2014 : Major tagging update, which in the end amplifies this effect, particularly through a cumulative effect.

• 12nd February 2014 : Addition of the CZ-75, automatic pistol with a very high rate of fire and mostly used in melee situation that doesn’t particularly encouraged to aim for the head.

Thus, the combined actions of these three components make it in principle more interesting now to look for a hit, regardless of the location of the affected body, as soon as possible to prevent the opponent from aiming if he doesn’t have a kevlar (aimpunch) and moving (tagging) than trying to HS. The very low bullet dispersion of the silenced M4 could be added to this list, which doesn’t penalizes much the automatic firing and thus also encourages less accuracy.

Is that a bad thing? One could say that Valve is simply trying to give a different style to the game. Except that basically, what makes the beauty of a frag is partly the technical difficulty there is in realizing it. And that will then require a great speed and precision of execution to the players.

Some sprays are remarkable

But mechanically speaking, the smaller the target area is, the harder it is to hit it. Logically, the hitbox of the head is smaller than the rest of the body, so it is more difficult to HS than hit the body. Which means that, with a game and updates that encourage players to hit the body instead of HS, we reduce the aiming difficulty of the game and therefore inevitably, other things being equal, the skill-cap, considering that the aim is one of the most important part of the gameplay; which is not desirable for a competitive game.

Thanks to Sephi and Jcoste for the translation