Bungie spent the June 2 State of Play telling Marathon players Season 2 was "a fresh start." The data from the weeks after tells a harder story: the bump was real, brief, and then gone.
What Season 2 actually did
Per the PlayStation Blog, Bungie ran an Open Play Week from June 2 to June 9 with no PS Plus required, opening Tau Ceti IV to newcomers alongside the Season 2 launch. The free trial plus new content, including a Night Marsh map and a Sentinel defender class, worked in the short term.
Marathon climbed to roughly 40,686 concurrent Steam players during that window, per Dexerto's tracker. Against an average that had been sitting near 10,000, that's a four-fold spike. A free week will do that.
The hangover
The recovery didn't stick. By late June, SteamDB showed Marathon peaking around 25,807 in a 24-hour window, well off both the launch and the Open Play surge. Free-week numbers are borrowed, not earned, and the falloff once the paywall returned is the tell.
For context on the original disappointment: Marathon launched around an 80K peak and quickly settled near 10K, the kind of retention curve that ends live-service games rather than sustaining them.
The brutal comparison
The gut punch came from Bungie's own backyard. Destiny 2's final update on June 9 drew 167,867 concurrent Steam players, surpassing Marathon's all-time peak of 88,337. The studio's sunsetting game pulled nearly double the lifetime high of its expensive new one, on the same week Marathon was trying to mount a comeback.
The read
Extraction shooters are a crowded, unforgiving subgenre, and Marathon launched into skepticism about whether Bungie could land a brand-new IP after years of Destiny. Season 2 proves the team can still ship content that pulls curious players in. It doesn't yet prove they can make those players stay, which is the only metric a live-service game is graded on.
The honest version: a free week is a megaphone, not a foundation. Until Marathon posts retention without a giveaway propping the chart up, the 40K headline is a spike on a graph that keeps trending the wrong way.

