@selixThe Cities Skyline Paradox
Why the sequel stumbled
and how a new studio might save it
An Update on Cities: Skylines II
In mid-November 2025, Paradox Interactive and long-time partner Colossal Order said they would go their separate ways. After fifteen years, Paradox is moving Cities: Skylines II to Iceflake Studios, an internal Finnish team. Colossal Order (CO) will ship one last “Bike Patch” and an asset-editing beta, then step away. Players had seen this coming: Cities: Skylines II (CS2) launched in October 2023 with technical problems, design missteps and no mod support, and many of those issues lingered a year later.
In this post, I’m trying to sort out why CO stalled, why Iceflake gets the keys, and which parts of the sequel can realistically be fixed.
A brief history of the series
Cities: Skylines (2015) emerged from the rubble of Maxis’ SimCity reboot, combining approachable city-planning mechanics with modding openness. Colossal Order, based in Tampere, and Paradox kept the mechanics approachable and let mods flow through Steam Workshop. The first game sold well, gained dozens of expansions and a massive mod scene. CO was only about thirty people; that small-team feat still impresses me nine years later.
Unity Entities Graphics feature matrix lists Burst Occlusion Culling as “Experimental”
In 2023 CO tried to leap ahead with a sequel. CS2 was built on Unity’s High Definition Render Pipeline (HDRP) and promised per-citizen simulation, a dynamic economy and cross-platform modding. Paradox warned before launch that performance might disappoint. They were right: the PC release shipped with heavy GPU bottlenecks, slow simulation and a thin economy. The detail that stuck with me from Paavo Huhtala’s post-mortem is that every pedestrian model carried 6,000 vertices, teeth and all, and pallet stacks rendered in full even when hidden. Entities Graphics’ occlusion culling was still marked experimental, so CO’s ECS/HDRP glue effectively shipped without a reliable occlusion pass. Those high-resolution shadows and assets let the game issue “an innumerable number of draw calls,” choking GPUs while CPUs sat mostly idle.
One Year Later - Cities: Skylines II Is Still a Broken, Lifeless Mess
The simulation wasn’t much healthier. A year after launch, players were still complaining about drone-like citizens, empty parks and traffic AI that chose bizarre routes; the promised dynamic economy felt “nonexistent.” CO postponed the console release and paid DLC to focus on fixes, but slow-downs and path-finding issues kept popping up through 2024 and 2025. In my own launch-week save the camera still hitches when I pan over industry, even after the later patches eased the worst stutter.
Paradox Mods FAQ; Hallikainen on missing mod support
Modding – central to CS1 – was mostly MIA at launch. Paradox and CO said CS2 mods would live on Paradox Mods, a single hub meant to work on PC and consoles, and that Steam Workshop wouldn’t be supported. That business choice delayed the Editor and kept PC modders on the sidelines; deep monkey-patching was technically possible but brittle while the internals churned. In February 2024 CO CEO Mariina Hallikainen called the missing mod tools their “biggest regret”: “The biggest regret we have is that modding support is not yet available for the game… we simply ran out of time as the focus had to shift from modding support to all hands on deck to fix the performance.”
The facts of the change
Paradox’s November 17 2025 post lays out the plan: the Cities franchise moves to Iceflake Studios, an internal team in Tampere, Finland. Iceflake takes over “all existing and future development” of CS2 – free updates, the in-game Editor, console versions and future expansions. CO will ship one more update, the Bike Patch, with bike lanes, Old Town buildings and bug fixes, plus a beta of the asset-editing tools. Iceflake assumes full control in early 2026.
The public quotes are cordial. Hallikainen thanked Paradox for fifteen years and said they were “excited to channel our experience, creativity, and passion into new projects.” Paradox deputy CEO Mattias Lilja promised to “provide [Cities players] with more content and new experiences.” Iceflake’s Lasse Liljedahl called the handoff “an immense honor and a great responsibility… we will humbly carry this legacy with respect.” Optimistic words, but also the usual PR gloss.
The quieter admission came earlier. In October 2024 Lilja said launching CS2 in October 2023 was a mistake: they thought “iterating this live was probably the right way to go” but “should probably not launch that early”. They shipped knowing it was unfinished. Two years later the sequel still carried that baggage, and moving it to an internal studio lets Paradox reset without cancelling it.
A few things went wrong at once:
Technical overreach
CO aimed high: per-citizen simulation, dynamic economy, cinematic visuals. They were still a 30-person studio, and Unity HDRP punished their asset choices. Ordinary models were rendered at absurd detail, with no lower-detail meshes. The ECS/HDRP setup effectively lacked real occlusion, so millions of polygons drew even when off-screen. Shadows stayed high-res everywhere. In theory you can fix that with aggressive LODs and asset rework; in practice it’s months of grind for a small team already fighting fires. Plenty of players would happily accept CS1-level visuals and a rock-solid simulation over HDRP lighting that cooks a GPU.
Simulation ambitions also ran ahead of implementation. Citizens moved like drones, parks stayed empty, traffic AI picked odd routes. Public transport hardly dented congestion. CO issued an Economy 2.0 patch in mid-2024, but core behaviors still felt unfinished by late 2025.
Management and business constraints
CO was juggling PC release, console ports and DLC while also building a brand-new cross-platform modding system to satisfy Paradox’s Paradox Mods plan. That meant one pipeline for PC, Xbox and PlayStation, plus console security rules. It displaced Steam Workshop, the de-facto home for CS1 mods, and it simply wasn’t ready, so the game launched without it. Hallikainen later called that the team’s “biggest regret”.
Paradox wanted revenue from DLC and consoles and agreed with CO to “iterate live”. That turned paying players into beta testers. Patches fixed one issue and surfaced another, and each console delay chipped away at trust.
Human factors
CO had been on city-builders for more than a decade. Fatigue is hard to quantify from the outside, but stepping away lets them stop being “the Cities studio” and try something fresher. On my end, replaying the same launch map to test each patch got old fast; muscle memory from CS1 still has me reaching for Network Multitool-style tools in CS2’s UI and Paradox Mods, even though they are not there yet.
Why Iceflake might succeed
Past work: Surviving the Aftermath review
Iceflake Studios isn’t a household name, but it has relevant miles. Founded in 2007 and picked up by Paradox in 2020, the team built Surviving the Aftermath, a post-apocalyptic colony builder that entered early access in 2019 and hit 1.0 in November 2021. It launched rough, improved steadily and ended up with “mixed or average” reviews around 69/100 on Metacritic – not a smash, but proof the studio can grind a management game into shape.
Because Iceflake is inside Paradox, the publisher can drop in specialists and control the roadmap more tightly. Iceflake inherits the source, the toolchain and the docs. They’re less emotionally tied to the old design, so they may be quicker to cut features that don’t pull their weight. CS2 already has larger maps, better road tools and flexible zoning. If Iceflake can add proper LODs and occlusion, and keep chipping at traffic and economy logic, the game could finally feel stable for regular players.
To be fair, CS2 does try things CS1 never could: citizens hold specific jobs and commutes instead of teleporting to whichever slot is free, companies import and export resources and need to turn a profit, and the lighting aims for realism. Those are hard to bolt onto CS1’s architecture and part of why a sequel existed at all.
There are real limits, though. Iceflake can’t swap out Unity HDRP. Paradox Mods stays the official hub, and deep monkey-patching on PC is still possible, but the shifting internals make those mods brittle and far from CS1’s stable API plus Steam Workshop. Expect performance work, bug fixing and steady tweaks, not sweeping reinvention. If their first patch just halves the draw calls and stops citizens from teleporting between tram stops, I’ll happily trade a few high-detail trees to get there.
Paradox’s course correction
Paradox has been on an apology tour after rough launches like Bloodlines 2, Prison Architect 2 and the cancelled Life By You. Lilja has said they misjudged hardware and released too early. Moving CS2 in-house fits that course correction: more control, fewer surprises.
There’s also the boring business reality: paying a team for a year to refactor CS1 without a new box price is hard to justify. DLC tails taper; a numbered sequel gives marketing something big to sell again. That does not excuse CS2’s launch, but it explains why “just keep patching CS1” rarely wins inside a spreadsheet.
Short term, the roadmap is simple: Bike Patch, asset-mod beta, console work. After that Iceflake will share its own plan. Paradox keeps nudging players to link accounts for cosmetics, signaling they’re not sunsetting the series, just trying to steady it.
Cities: Skylines II is what happens when ambition outruns capacity. A small team pushed high-detail assets, shipped without mod tools and hoped to finish live. Paradox let it launch anyway. Players balked, and two years later the sequel still needs work.
Is it doomed? Maybe not. Iceflake inherits a game with big maps, flexible roads and a community that wants it to succeed. Performance fixes (better LODs, occlusion, asset passes) are grindy but doable. Traffic and economy tweaks are harder yet not hopeless. What CS2 won’t be is CS1 with stable, open modding; Paradox Mods and console parity keep the gate narrow, and brittle internals mean deep mods break often. If Iceflake can deliver stability and believable simulation, that might be enough to make the sequel finally feel finished.