I can see your Halo: player behavior and map metrics in Halo 5: Guardians

  • 28 March 2022
  • 0 replies

Userlevel 5

This content, written by Elliot Park, was initially posted in Looker Blog on May 24, 2018. The content is subject to limited support.

I’ve been playing Halo for a really, really long time.

I started with Halo: Combat Evolved, then Halo 2, but things really clicked when I started playing Halo 3 online. I never reached Rank 50, but I’ve spent a lot of time driving Warthogs off cliffs, getting stuck by plasma grenades, and playing games of Big Team Battle way past midnight.

As a new member of Looker’s Department of Customer Love, I was given the opportunity to build a data model from scratch off a dataset of my choosing. When I found out that for Halo 5 multiplayer data, it really didn’t take long for me to decide that this needed to be my dataset, no question.

Like a lot of competitive multiplayer games out there, Halo 5 is chock full of player performance metrics and statistics, and by digging into this dataset, we can draw a lot of cool insights regarding how players play Halo 5 and what makes some players really good (and others not so much).

The data

My dataset contains information on 1,068 unique players and 28,565 recorded matches, and is mostly comprised of matches from the Fall-Winter 2017 Season. In regards to my ETL process, I used a Ruby script to generate my list of gamertags and create my player table. Using those gamertags, I then pulled the 25 most recent matches for all my players to create my match table. Additional metadata on weapons, maps and gametypes were all stored as their own individual tables, as well.

The following analysis is not an end-all-be-all conclusions surrounding competitive Halo, but rather, an attempt to paint some very broad (but interesting) strokes about Halo 5 multiplayer data based on the sample data I was able to extract.

So without further ado, let’s dive into the data!

Weapon usage in slayer

Among all my ranked Slayer* players, by far the most used weapon is the Magnum. Based on my dataset, 100% of players at the Diamond, Onyx and Champion rank classes have more kills racked up with their Magnums compared to any other weapon. Note that as you go up each rank class, the percentage of players who primarily use the Assault Rifle or Melee/Spartan gets smaller and smaller.

While it’s no surprise that the Magnum is by far the most popular weapon among high ranking Halo players, it’s pretty crazy to see Magnum usage dominate to this extent. Bunching Magnum, Assault Rifle, and Melee/Spartan users up, let’s see how these groups stack up to each other using a variety of player metrics.

Looking at average win percentage (i.e. total matches won divided by total matches played), we can see that Magnum tend to lose a whole lot of their matches at the Bronze rank. As we climb the ladder, however, we see win percentages for Magnum users dwarf both our Assault Rifle and Melee/Spartan users.

We see very similar trend lines when comparing the average overall accuracy of each weapon group. Bronze-ranked Magnum users tend to struggle with landing shots, but at higher ranks of play, Magnum users prove to be the most accurate compared to their counterparts.

And in looking at average Kill Death ratios (i.e. total number of kills divided by total number of deaths), Magnum users again outclass the opposition. Note that player rank class and average Kill Death ratio seem to have an almost linear correlation in this case.

All in all, it’s apparent that Bronze-ranked Magnum users tend to struggle, but Magnum users at every other rank hit their targets more often, get more kills per death, and win matches more often than the competition. What this ultimately means is that while there may be a place for switching to your Assault Rifle or using your melee button, getting really good with the Magnum pistol and landing clean shots may be your best bet for climbing the ladder!

Map analysis

Halo 5 is no slouch when it comes to map variety. Taking a look at average Kill Death ratios across different maps, you’ll find some pretty interesting facts about how players of varying ranks perform on various map.

One observation here is that our Champion-ranked players show some absolutely dominant numbers on a handful of maps. Looking at the data, Champions tend to have their best matches on Stasis, Coliseum, Plaza and The Rig, averaging a Kill-Death ratio of 1.90 or better! This stands in stark contrast to how tough these maps seem to be for Bronze-ranked players, averaging Kill Death ratios right around 0.73 or less.

While it might be a stretch, personal experience tells me that these maps see a lot of long range skirmishes (note that three out of the four maps I mentioned have Sniper Rifle pick-up locations on them).

Ultimately, players who can advantage of the larger map sizes with well placed shots and map control over power weapon spawns can really take over in maps like these.

Diving into the nitty-gritty, let’s take a deeper dive into one of Halo’s most iconic maps, and one of my personal favorites: Truth (aka Midship, for any Halo 2 players out there).

Taking a look at average Slayer match duration by map, Truth tends to have some of the shortest matches among all of Halo 5’s multiplayers maps. Considering that you’re pretty much within eyesight of the opposing team seconds after spawning in, it makes sense that games tend to go shorter on Truth in particular.

Truth also tends to see an above-average percentage of players leaving Slayer matches mid-game. 8.41% of my recorded Slayer matches saw at least one player drop midway through the game, and while that’s not as high as the drop rates on Stasis and Empire (11.31% and 9.77%, respectively), it’s significantly higher than the numbers on Slag, Riptide and Eden where percentage of mid-game exits hover around 5%.

As a closing note, the average difference in team scores at the end of a Slayer match is larger on Truth than on any other map. In other words, winning teams tend to win Slayer matches by higher margins on Truth than every other map on average. However, based on my sample data, most Slayer matches tend to see winning teams about 11 to 12 kills ahead of losing teams by the end of a match, meaning most games don’t end up being particularly close, regardless of the map you’re on.

Closing thoughts

343 Industries has done a good job of making this all very accessible, and I definitely recommend checking the Halo 5 API out if you’re at all curious!

Thanks for taking the time to dive into the data with me, and I hope you enjoyed it. And to all you Halo players out there, “Good luck, and have fun!”

* Slayer - One of the hallmark modes of Halo multiplayer. Two teams compete against each other by killing opposing players. The first team to 50 kills wins.

0 replies

Be the first to reply!