field notes · june 2026
A Decade-Old Game Just Set Its All-Time Player Record. Here Is What Discount Revivals Mean for Servers
In June 2026, The Forest, a survival-horror game that entered Early Access ten years ago, hit 109,242 concurrent players on Steam. Not a nostalgia bump. An all-time record, set on a sub-$5 discount. Its sequel surged alongside it at its lowest price ever. This is the cleanest demonstration yet of something worth understanding if you run, rent, or sell game servers: in co-op survival, demand spikes are price events more often than content events, and they have a shape you can plan around.
The record nobody predicted
The numbers first, because they are remarkable. The Forest (2018, Early Access 2014) was discounted below five dollars in early June 2026 and promptly posted 109,242 concurrent players, the highest figure in the game's history, ahead of every launch-era peak. Coverage rightly framed it as a ten-years-later record. At the same time, Sons of the Forest hit its lowest price ever, roughly $8.99 at 70% off, and its player count multiplied by around two and a half. All of this landed before the Steam Summer Sale, which opens June 25 and runs into early July, meaning the wave has a second, larger crest still ahead of it.
Nothing shipped. No update, no expansion, no overhaul. The games are the same software they were in May. The only variable that moved was price, and the demand response was the strongest either game has ever recorded.
Why price moves co-op survival like nothing else
Co-op survival has a property most genres lack: it is bought in groups. A single-player game at $30 is one person's decision. A co-op survival game at $30 is four or six people's decision, made simultaneously, with the group's least-enthusiastic member holding an effective veto. That coordination problem suppresses sales of older co-op titles at full price far below their actual appeal.
A deep discount dissolves the veto. At $5, "just buy it so we can play tonight" works on everyone, including the skeptic. The result is that aged co-op survival games hold enormous latent demand that only price can unlock, which is why a decade-old title can out-peak its own launch: the 2026 audience of potential friend-groups is far larger than the 2014 one, and the discount reaches all of them at once.
The shape of a revival wave
Price-driven waves are consistent enough to sketch: a sharp ramp during the sale window, a peak at or just after the deepest discount, then a two-to-six-week decay toward a baseline that usually settles somewhat above the pre-sale level, because some fraction of the wave converts into a lasting community. The June case has an unusual structure: the publisher discount produced the record before the platform-wide Summer Sale (June 25 to early July), so the decay of wave one will collide with the ramp of wave two. Expect elevated populations across co-op survival into mid-July, then the usual settling.
The other side: reversion data from the same month
The same player-count data that shows the revival also shows what happens when event-driven spikes end, and it is worth holding both in view:
- Conan Exiles gave back roughly a third of its players in the weeks after its UE5 "Enhanced" relaunch spike faded, a spike that had tripled its population.
- V Rising reverted about 30% after a free weekend and half-price sale ended.
- Windrose, a 2026 release, continued shedding players post-launch despite strong sales figures, the ordinary decay every launch sees.
None of these are failures. They are the other half of the same physics: event-driven demand is a wave, not a step. Relaunches, free weekends, discounts, and launches all borrow players from the future; some stay, most recede. The error is not the spike, it is planning as if the peak were the new baseline.
What it means for server demand
Player-count waves translate into server demand with a short lag and a filter. The lag: friend groups peer-host first, then hit the familiar walls, the world is offline when the host sleeps, the host's connection advantages them, saves live on one person's machine, and start shopping for persistent hosting within days. The filter: only groups that intend to keep playing rent servers, which makes server demand a higher-quality, slightly delayed echo of the player spike, typically cresting one to two weeks after the population peak and decaying on the same curve.
For anyone operating in this market, the June wave is therefore predictable in both directions: co-op survival server demand rises through late June and July, concentrated in the discounted titles, and recedes toward late summer. The Summer Sale's June 25 opening, colliding with Palworld's July 10 full release in the same genre-adjacent space, makes this the densest server-demand window of 2026 so far.
The sizing playbook for a wave you know is temporary
- For communities: rent monthly, not yearly, during a revival. Peak-enthusiasm contracts are how groups pay for empty slots in September. Scale the slot count to the people who show up in week three, not week one.
- For operators and self-hosters: elasticity beats provisioning for peak. The wave's shape is known; capacity that can shrink gracefully in August is worth more than capacity that handles the July crest with headroom to spare.
- For anyone reading store pages: treat a discount-driven all-time peak the way you would treat a free-weekend number, as proof of latent appeal, not of steady-state population. The number that matters for a server is who is still logging in six weeks after the price goes back up.
The Forest's record is a genuinely happy story: a good game finding its largest audience ever, ten years on, because the price finally matched the group-purchase reality of its genre. The server demand that follows is real, valuable, and temporary in exactly the proportions the last decade of sale waves has taught. Plan for the wave; do not mistake it for the sea level.