Drift0r reveals preliminary results from Black Ops Cold War SBMM study

Liam Mackay

A group of content creators is doing a detailed study on skill-based matchmaking in Call of Duty: Black Ops Cold War and Drift0r revealed some preliminary results.

Skill-based matchmaking (SBMM) has been a controversial topic in Call of Duty since Modern Warfare’s release in 2019. At its core, the system is designed to match players with others of a similar skill level.

Members of the community active on social media have long complained that SBMM is too strong in Cold War, taking casual fun away from public matches.

A team of content creators are doing an in-depth study on SBMM in Black Ops Cold War.

Previous study

Last year, content creators Drift0r and TheXclusiveAce did a comprehensive study of SBMM in Modern Warfare. They wanted to discover if the skill of the account has any bearing on connection by collecting data from accounts of varying skill.

Drift0r drew these conclusions:

  • Modern Warfare never prioritizes skill over connection
  • High stats players do not match together
  • There is matching based on recent performance
  • There appears to be a hidden matchmaking rank (MMR) system
  • Lobby difficulty changed from an experiential point of view but not mathematical

Check out the full video for detailed explanations of these conclusions.

Preliminary results

They are repeating the study in Black Ops Cold War, this time with YouTubers JGod and S0ur.

The team has now collected all their results and will be releasing their full report soon. Drift0r gave us an insight into what they’ve discovered, however.

“Our data indicates that skill influences ping and matchmaking time,” Tweeted Drift0r. “This is different from MW19 or MW2 where it didn’t seem to affect connections.”

We can’t draw any conclusions of our own from these preliminary results, but Cold War’s SBMM appears to be a different beast to what we saw in Modern Warfare.

Black Ops Cold War's 2v2 Gunfight

YouTuber Gamer Central asked Drift0r how big a sample size they’re taking for their studies, as they felt last year’s was too small.

“Our same size is 9 accounts playing 10 games each,” Drift0r replied. “Every conceivable variable controlled. API doesn’t let us auto pull stats anymore so it was all manual data entry.”

He admitted that “this sample size is almost certainly too small” but was the best he could do given “limited resources.”

We’re likely to see the full results the weekend beginning December 11, but it could be later.