Guide

Game plus-on-block frame advantage systems explained

Harbor Brawl's first pressure pass gave every rushdown character the same signature: a three-hit string ending +2 on block, a throw loop, and a safe special. Ranked data looked healthy — block damage accumulated, corners happened, rounds ended. Player surveys told a different story. Defenders described pressure as “one long blur” with no readable decision points. Mash reversals spiked in gold tier because every gap felt identical: not quite punishable, not quite safe to keep blocking. Attackers, meanwhile, could autopilot the same string from midscreen neutral to corner trap without adjusting spacing.

Frame advantage on block (often called plus on block) is the number of frames the attacker recovers before the defender can act after a blocked hit. It is the atomic unit of fighting-game pressure: +0 means both players can press on the same frame; +4 means the attacker gets four frames of exclusive action before a jab can interrupt. Designers tune not just whether a move is plus, but how plus — because +1 enables tick throws, +3 enables meaties, and −2 creates frame traps. Harbor Brawl split its pressure kit into explicit advantage tiers, re-tuned pushback per tier, and mapped punish windows for every character's fastest post-block button. Uninformed mash reversals fell 44%; intentional counterplay rose. This guide covers advantage math, blockstun interaction, tier design, effective spacing, frame traps from minus positions, the Harbor refactor, a technique decision table, pitfalls, and a production checklist.

How frame advantage is calculated

Every attack has three phases relevant to blocking: startup (wind-up before hitbox appears), active (frames the hitbox can connect), and recovery (frames after the last active frame before the character can act again). When the defender blocks, they enter blockstun for a move-specific duration instead of hitstun.

Frame advantage on block is:

advantage = blockstun − remaining_recovery_after_last_active_frame

If blockstun is 15 frames and the attacker has 12 frames of recovery left when blockstun ends, advantage is +3. If recovery is 17 frames, advantage is −2 — the defender's 4-frame jab beats any follow-up slower than 3 frames startup (accounting for input buffer). Publish this number per move in your frame data sheet; players and AI balance tools depend on it.

On hit vs on block

Hitstun and blockstun are separate columns. A move can be +5 on hit (long combo) but −1 on block (honest special). Never assume hit advantage transfers to block pressure — specials tuned for confirms often end minus on block by design, creating natural reset points in blockstrings.

Advantage tiers and what they enable

Treat frame advantage as discrete pressure tiers, not a continuous slider. Each tier unlocks specific offensive tools:

Advantage	Typical tools	Defender counterplay
+5 or higher	True blockstring links, guaranteed meaty on many knockdowns	Hold block; reversal only if invincible startup fits
+3 to +4	Meaty normals, layered frame traps, walk-forward throw	Delayed button, backdash, reversal
+1 to +2	Tick throws, re-stand pressure, fast special cancel	Jump, reversal, fastest jab (character-dependent)
0	Strike/throw mixup at neutral speed	Any button of equal or faster speed wins trades
−1 to −3	Frame trap bait (attacker expects mash)	Jab, short hop, backdash out of range
−4 or lower	Punish window; pressure must stop	Full combo punish, throw, or invincible reversal

Harbor Brawl's mistake was collapsing tiers: every route landed at +2, so tick throws, meaties, and frame traps all required the same defender response. The refactor spread normals across +1, +3, and −2 endpoints so pressure had rhythm — safe stretches, risky traps, and clear punish moments.

Pushback and effective advantage

Published frame advantage assumes point-blank spacing. Block pushback moves both characters apart during blockstun; at max range, a +2 string may fail to link because the next move whiffs. Conversely, at corner, pushback is clipped and effective advantage rises — a +2 midscreen can functionally behave like +4 when the defender cannot backdash out of throw range.

Document three values per move where possible:

Raw advantage — frame math at point blank.
Spacing band — min/max distance where links still connect.
Corner modifier — whether blockstun or pushback changes within one body-width of the wall.

Footwork interacts with advantage: a +1 move after walk-forward may land as +3 effective because the attacker closed distance during the opponent's blockstun. Test pressure with walk-forward held, not just standing strings.

Frame traps, resets and minus-on-purpose

Not every move should be plus. Frame traps are sequences that end briefly minus on block, baiting the defender to press a button that loses to the attacker's faster follow-up. A classic pattern: string ends −2, attacker delays a 5-frame normal; defender's 4-frame jab loses to the trap but beats a throw. Frame traps require telegraphing — audio cue on string end, distinct animation — or low-rank players experience them as random losses.

Intentional minus also creates reset points where offense stops and shimmy or backdash bait begins. If every link is plus, defenders never get a mental breath and resort to mashing reversals. Alternate +3 re-stands with −2 traps and 0-ending specials so conditioning loops have phases.

Punish windows and reversal math

When advantage is negative, the defender's fastest option defines the punish window. If a move is −6 on block and the defender's jab has 4-frame startup, a perfect punish requires the jab input within 2 frames of blockstun ending (before accounting for buffer). Wider punish windows (−10 or more) enable counter and heavy punish routes.

Reversals (invincible dragon punches, alpha counters) add a parallel clock: startup must finish before the attacker's meaty active frames, but after invincibility covers the trap gap. Document per-character:

Fastest normal punish at each minus tier (−3, −5, −8).
Reversal invincibility start frame and total startup.
Whether buffer allows reversal to beat tick throws at +1.

Harbor Brawl added an in-training frame overlay showing advantage tier color-coded on block — green +3, yellow +1, red minus — cutting “I got mashed for no reason” tickets 38%.

Harbor Brawl advantage refactor

The refactor followed four steps:

Audit — export every blocked hit's advantage; flagged 73% of rushdown strings within +1 of each other.
Tier assignment — each character gets one +4 tool, two +2 tools, one 0-ending special, and one −3 trap ender.
Pushback pass — +4 tools push less; −3 enders push more to reset midscreen.
Punish QA — automated script steps every minus value against fastest jab per character; flagged 12 impossible punishes and 8 free escapes.

Post-patch, average pressure sequence length before a reset or punish rose from 4.1 hits to 6.8 hits because defenders could identify tier shifts. Round length in platinum stayed flat — more decisions per minute, not longer staring at blockstun.

Technique decision table

Approach	Strengths	Weaknesses	Use when
Uniform +2 pressure	Simple to learn; stable blockstrings	Monotone; mash-heavy meta	Beginner-friendly casual modes only
Tiered advantage kit	Readable decisions; conditioning depth	More balance surface per character	Ranked and tournament play
Minus-on-purpose traps	Punishes mashing; creates mind games	Feels unfair without telegraph	Mid-to-high skill after tutorials teach tiers
Spacing-dependent advantage	Rewards footsies precision	Hard to document; online variance	Footsies-heavy fighters with clear hurtboxes
Chip-only pressure (no plus)	Defenders retain counter options	Low kill threat; timeouts	Zoning or timer-heavy formats

Common pitfalls

Publishing hit advantage as block advantage — separate columns in every movelist and API.
Ignoring buffer — +1 tick throws must lose to buffered jump or reversal when those options exist.
Corner-only testing — advantage and pushback differ materially at wall; test both.
All plus, no traps — pressure without minus endpoints trains reversal mashing.
Impossible punishes — −4 on block but fastest punish is 6-frame startup feels broken; widen window or reduce minus value.
Identical blockstun across normals — lights and heavies with same advantage remove texture from strings.
Rollback desync on block — advantage must be deterministic from blocked move ID and spacing; see rollback netcode.
Buffing pressure without meter cost — pair high-tier plus tools with meter spend or cooldown where appropriate.

Production checklist

Export startup, active, recovery, blockstun, and computed advantage per move.
Assign explicit advantage tiers per character kit; avoid clustering within +1.
Test every string link at min, mid, and max spacing plus corner.
Script punish validation: minus value vs fastest defender normal per character.
Validate tick throws at +1 against jump, jab, and reversal with buffer on.
Telegraph frame-trap endpoints with distinct SFX or animation cadence.
Document corner pushback modifiers separately from midscreen values.
Playtest mash rate and reversal rate by rank tier after each balance pass.
Expose advantage tier in training mode overlay or frame debugger.
Regression-test block advantage after any blockstun or recovery patch.

Key takeaways

Frame advantage on block is blockstun minus attacker recovery — the foundation of all pressure math.
Advantage tiers (+4, +2, 0, −3) unlock different tools: meaties, tick throws, mixups, traps, and punishes.
Pushback and corner geometry change effective advantage beyond the published number.
Intentional minus endpoints prevent monotone pressure and reduce reversal mashing.
Harbor Brawl cut uninformed mash reversals 44% by tiering advantage and visualizing punish windows.
Automated punish QA catches impossible windows before players report them as bugs.