Overwolf Logo

Spawnpeeking; a data-driven analysis

This article is a more in-depth companion piece to our post on the official Rainbow 6 subreddit.

tl;dr

On average, a Siege player is likely to die (either on the ATK or DEF side) within the first 18 seconds of a round in approximately 0.66% of rounds. This does not appear to differ significantly depending on your Rank.

Intro

Hey everyone! My name is Bl0n, and I'm the developer behind R6 Analyst. As more and more people are using the app, we've collectively compiled a rather interesting dataset of over 25000 Ranked matches, all played in the past two months.

I’ve been diving through this dataset, looking for some tasty nuggets of information. And I have found some interesting stuff that I thought I’d share!

Today, I’d like to highlight some statistics about Spawnpeeking in Siege.

Summary

An analysis of 143031 rounds from 25240 Ranked matches across 2224 players from all over the world, shows that the average Siege player is likely to experience dying by a spawnpeek in approximately 0.66% of Ranked rounds. Experiencing any sort of spawnpeeking personally (doing it and succeeding or failing, or dying to it or getting an early pick as an attacker) occurs in ±2.14% of rounds for each average player.

Defining Spawnkills and Spawnfails

For this analysis, I made the following definitions. These are just my own invention, you may have other opinions:

  • I considered any kills within the first 18 seconds of the Action Phase as a spawn-related kill.
  • A spawnkill is a kill a defender makes on an attacker.
  • A spawnfail is when a defender is killed by an attacker, also known as an ‘early pick’.
  • A spawndeath is any death in the first 18 seconds, on either side.

I dove through my dataset of 25240 Ranked matches, tracked by 2224 unique players. The players are from all over the world (at least 87 different countries) and play on all three major data centers (APAC, EMEA, NCSA).

A note about the dataset: There are various possible biases in the dataset, especially with regards to MMR and region. The table below shows the number of rounds tracked per Rank category.

A note on the graphs

The graphs in this article deal with very small numbers. Please keep in mind the Y-axis! I hoped to strike a balance between readability of the graph, and representation that the events that the bar graphs describe are not present in the majority of Ranked rounds.

How prevalent is Spawnpeeking?

My data seems to suggest that the average Siege player is likely to die within the first 18 seconds of the Action Phase in approximately 0.66% of rounds. Your rank does not appear to have any statistically significant effect on the prevalence of spawnpeeking. Whether you're a Copper or Platinum player, you're almost just as likely to get a spawnkill or die by it. (Not enough data for Diamond/Champion.)

Experiencing any sort of spawnpeeking personally (doing it and succeeding or failing, or dying to it or getting an early pick as an attacker) occurs in ±2.14% of rounds for each average player.

Note that, with 5 players on each team, the chance of anyone dying at the start of any particular round, is much higher.

NOTE: The Y-axis only goes up to 10%, not 100%. Error bars at 2σ.

Spawnpeeking per Map

Here’s a graph of every Ranked map, and the prevalence of spawnkills/spawnfails.

NOTE: The Y-axis only goes up to 5%, not 100%

With the dataset split into twelve maps, a lot of data is needed to make any hard conclusions. However, as you can see, Bank appears to be the most dangerous Ranked map in terms of spawnpeeking, with an average player spawndeath event occurring in 3.11% of rounds. The safest map looks to be the reworked Theme Park, which has spawndeath events in only 1.40% of rounds.

It’s interesting to try and determine whether it is worth it for Defence to actually attempt a spawnpeek on any particular map. Of course, the preferred outcome is to get a ‘free’ kill early on. The map that comes closest is Theme Park, where Defenders had 1.75x more spawnkills than Attackers. Consulate and Chalet are also contenders, with ±1.5x more spawnkills than spawnfails. However, the dataset is not large enough to verify this holds true overall with a statistically significant amount of certainty. More data is required to determine this.

Spawnpoint analysis

The dataset is not (yet) big enough to make statistically significant conclusions on individual Spawnpoints, but I still thought it’d be fun to show an overview of all Spawnpoints, and how dangerous they are:

Click on map-name to zoom in.
Hover over Spawnpoint to see % rounds in which spawnkilled on average.
Click center to go back to start.

I personally expected Kanal Sailboats to have the top spot, but it’s actually near the very bottom! 

Conclusion

In total I regret to say that, with the exception of some spawnpoints (like Clubhouse: Construction Site and all Bank: Alley Access), you’ll probably come out on top as a defender, if you do decide to spawnpeek.

However, a very important thing to note here is that the data as tracked only takes into account kills and deaths in the first 18 seconds. There is much more going on. Any spawnpeek attempt that does not pan out for the defenders (no kill/death) is likely an advantage to the attackers. They can pin or pinch the defender and kill him easily afterwards.

Thanks for reading! Stay tuned for some more tidbits of information that I’m pulling from the dataset. Next topic: Overtime. If you have any questions or comments, or other burning questions you think the dataset might be able to help answer, feel free to ask them on the Reddit post or through Twitter!

Want to improve in Siege in a data-driven manner? Try out the app!