Skullgirls 2nd Encore

There is a version of fighting game history where Skullgirls is a footnote: a pretty indie curiosity with eight characters, a weird art style, and a studio that didn’t survive the decade. That’s not what happened. What happened is that a small team built one of the most technically literate 2D fighters of its generation, and the rest of the industry spent the next ten years quietly adopting its ideas.

The lineage here matters, because it explains both the game’s polish and its scars. Skullgirls began at Reverge Labs, designed by a competitive player who treated tournament-level frame data as a design input rather than an afterthought. Reverge collapsed. The team reformed as Lab Zero Games, crowdfunded expansions, shipped 2nd Encore as the full edition — English and Japanese voice tracks, individually voiced story modes for every character, Survival, Trials, and a tutorial suite that remains the genre’s reference point — and then imploded publicly. Custody passed to Hidden Variable Studios, under publisher Autumn Games, which now maintains the game and ships new characters. That’s three studios for one product, and the seams are less visible than they have any right to be.

Market position is the awkward part. At $24.99, Skullgirls costs well under what Capcom or Bandai Namco charge, and it delivers more single-player content and better training tools than either. But it competes for the same finite thing every fighter needs: bodies in the matchmaking queue. Its Steam reception is strongly positive, which is good news — and also a sign that the population is measured in thousands, not hundreds of thousands. The page carries a Steam flag for off-topic review activity following an update that revised some character art and stage details, which drew an organized campaign unrelated to how the game plays.

Set in the Dark Deco world of the Canopy Kingdom, where every seven years the wish-granting Skull Heart corrupts whoever claims it, this is a game with a real aesthetic point of view. It is also, mechanically, one of the most demanding things you can buy.

• Hand-drawn animation at a frame count few match: Every frame of all 14 fighters was drawn by hand rather than rendered from 3D models or pixel-pushed. The result is cartoon fluidity — squash, stretch, and anticipation frames — inside a system where each of those frames is a hitbox with consequences.
• Custom assists: Build a team of one to three, then designate any move in a character’s kit as their assist call. Most tag fighters hand you three preset options. This one hands you the whole move list, which turns team building into a design exercise instead of a menu selection.
• Anti-infinite detection: The engine watches for repeated combo loops and automatically grants the victim an escape. It’s an elegant answer to the oldest problem in tag fighters — you get combo creativity without matches stalling out in long touch-of-death sequences.
• GGPO rollback netcode: Not delay-based. Real rollback, shipped years before the genre’s giants treated it as table stakes.
• The tutorial suite: Not a control-scheme walkthrough. An actual course in fighting-game fundamentals — spacing, reversals, resets, blockstrings — that teaches concepts transferable to every other game in the genre.
• Ratio-based team economy: One character means more health and damage. Three means assists and tag pressure with less of both. The trade is legible and it defines how you play.

The Gameplay Loop

Skullgirls runs on a simple, brutal premise: your team size is a resource, and you spend it. A solo fighter hits like a truck and survives a beating. A trio hits softer, dies faster, and gets two additional support calls plus tag-in mixups that a solo player simply cannot generate. Neither is the correct answer. That’s the point — it’s the first decision you make and it colors everything after.

What separates this from Marvel vs. Capcom lineage is the custom assist system, and it is genuinely under-appreciated. In most tag games, the assist roster is a designer-curated list: here’s a projectile, here’s an anti-air, here’s a launcher, pick one. Skullgirls says pick anything. Want a character’s slow, ugly command grab as a support call? Fine. Want a specific normal attack because its hitbox happens to cover a gap in your point character’s pressure? Also fine. This pushes real design authorship onto the player, and it means two people running identical teams can be playing meaningfully different games. It’s the kind of feature that only ships when the people making it play at a high level themselves.

The roster is 14 fighters and every one of them is strange. This isn’t a lineup with four shotoclones and a grappler. Peacock is a cartoon gag character built around item drops and screen control. Painwheel is a slow, self-damaging pressure monster. Double is a shapeshifter who borrows from everyone. Big Band is a brass-instrument parry specialist. The cost of this design philosophy is an enormous matchup-knowledge burden — you cannot transfer your understanding of one character to another — and the benefit is that the game never feels like it’s repeating itself.

Onboarding Friction

Here’s the contradiction the marketing won’t tell you. Skullgirls has arguably the best fighting-game tutorial ever shipped, and it is still a hard game to learn.

The tutorial is excellent: it teaches what a reset is, why throw-teching works the way it does, how to punish on block, how to think about neutral. Hidden Variable and Lab Zero before it treated teaching as a feature with a budget rather than a compliance checkbox. Trials and challenges extend the curriculum into specific, tough execution problems. Training mode displays hitboxes and frame data directly, which is the tool serious players used to need third-party mods to get.

None of that changes the fact that the execution ceiling is steep and the people still playing online are, overwhelmingly, the ones who climbed it. This is the structural problem with a long-running fighting game with a modest population: the casuals churned out years ago. What remains is a skilled, dedicated, and small community. New players will learn fast, because they will be losing constantly to people with years of muscle memory. Whether that’s a feature or a wall depends entirely on temperament.

Matchmaking and the Population Problem

Rollback netcode solves latency. It does not solve headcount. Quick match, ranked, and lobbies all work as designed, and the rollback implementation is built to hide the jitter that delay-based systems turn into dropped inputs — the design is still among the most reliable in the genre. But outside peak hours in a populated region, waits are inevitable. That is the single most common complaint in the Steam reviews and it is the correct complaint.

The offline content is unusually strong compensation. Every character has a fully voiced story mode with actual writing behind it, not a placeholder arcade ladder with two dialogue boxes. Survival Mode has real teeth. Trials will occupy players for weeks. A large set of unlockable color palettes gives the grind a cosmetic payoff. Bought purely as a single-player object, this would still be worth the price — which is not true of most fighters at three times the cost.

The Anti-Infinite System, Reconsidered

This deserves singling out. Infinite combos are the structural failure mode of freeform tag fighters; the usual fix is a hit-count damage scaler or a hard combo timer, both of which punish legitimate creativity along with the abuse. Skullgirls instead watches for repetition — the same sequence looping — and grants the defender an escape option when it detects one. Long, inventive, non-repeating combos survive untouched. Degenerate loops die. It’s a scalpel where the genre reached for a hammer, and the fact that nobody else copied it in all these years says more about industry inertia than about the idea.

The Dark Deco direction — 1930s American cartoon linework crossed with art deco architecture and an unsettling undertone — is a complete aesthetic, not a filter. Character animation carries weight and personality through frame count alone: hair reacts, coats trail, idle animations have breathing room. Because every frame is drawn rather than interpolated, attacks read instantly. You can see a move’s startup as a visual event, which matters enormously in a game where reactions are measured in frames.

Michiru Yamane’s score does exactly what a Castlevania veteran’s score should — jazz-inflected, moody, memorable, and sitting well below the sound effects in the mix where it belongs.

Technically, a 2D game of this scope is light on hardware, and years of post-release bug-fixing have gone into the build. The demands it places on a machine are modest by modern standards.

The UI is where the polish thins out. Menus are functional and dated — navigable, clearly organized, and visually a decade behind the fighting itself. Team select and assist configuration in particular involve more menu layers than they should. It’s not broken. It’s just the one part of the package that looks its age.

Steam is the default recommendation and it isn’t close. The largest player population lives here, arcade sticks and hitboxes are fully supported, remapping is comprehensive, and the frame-data and hitbox displays are easiest to read at desktop resolution and refresh rates. If you own a PC and intend to play online, buy it here.

Nintendo Switch is the interesting compromise. Portability suits a game with this much offline content extremely well — Trials and story modes are ideal handheld material, and the hand-drawn linework stays crisp on a small screen. The problems are input and population. The stock Joy-Con d-pad is poorly suited to precise quarter-circle and charge motions, and while the Pro Controller improves things considerably, serious play means bringing a compatible fightstick. The online pool is also smaller and more fragmented than Steam’s, which compounds the matchmaking wait that already exists.

Buy on Switch for the single-player. Buy on Steam for the fights.

Skullgirls 2nd Encore is a better fighting game than its commercial footprint suggests, and the gap between those two facts is the whole story. It shipped rollback netcode when the biggest publishers in the genre were still defending delay-based online. It built a teaching tool that the community recommends to newcomers playing other games. It solved infinites with a scalpel. It put more hand-drawn frames on screen than anything before it.

What it never got was the audience that volume of craft deserves, and no review can conjure one. Buyers are joining a small, sharp, welcoming-but-unforgiving community with a deep offline package to fall back on. At this price, against this competition, that’s a clear recommendation — with the caveat that the hardest part won’t be learning to play. It’ll be finding someone to play against.

Years on, the genre has caught up to Skullgirls on netcode. It still hasn’t caught up on everything else.