Daily Flanker is a specialist Daily outside Today’s 6. Each date selects a concrete Flanker block with a fixed difficulty, source, visual family, spacing and deterministic trial order. Reopening the same date reconstructs the same block, so the daily result is about replaying one stable challenge rather than chasing a newly randomized arrangement every refresh.
The verifier does not test a few sample dates and extrapolate. It materializes every 2026 date, checks the block against its level contract, re-materializes it for deterministic identity, and checks that each concrete Daily identity is distinct across the year.
Completing Daily Flanker writes to the specialist flankerStats.daily history and streak. It does not add or replace a game in Today’s 6, and it does not mutate the frozen Today’s 6 sequence or daily completion records.
| Difficulty | 2026 dates | Trials | Condition contract | Window |
|---|---|---|---|---|
| Easy | 60 | 48 | 32 congruent / 16 incongruent | 900 ms |
| Standard | 180 | 64 | 28 congruent / 28 incongruent / 8 neutral | 700 ms |
| Challenge | 125 | 80 | 28 congruent / 36 incongruent / 16 neutral | 550 ms |
The schedule intentionally spends most of the year in Standard while retaining a substantial Challenge segment. The daily difficulty determines the trial-count and condition-count contract but never changes the basic answer authority: the center target still decides left or right.
During development, a first deterministic mapping produced only 363 distinct concrete identities across 365 dates. A second attempt reached 364. Those failures were kept as development evidence rather than being waved away. The final mapping assigns each date within its difficulty range a frozen source index and a display-variant combination in an injective way, while the date still deterministically controls trial ordering and global direction flip.
The final verifier therefore reports 365/365 deterministic, 365/365 independently valid and 365/365 distinct for frozen 2026. This is stronger than merely putting the date into a seed string: the identity check represents the concrete materialized block, not the text of the seed.
The Daily block retains its difficulty, source, symbol family, spacing, left/right target sequence after materialization, congruent/incongruent/neutral arrangement and response window. A replay can naturally produce different reaction times and errors because the player is different on each attempt, but the underlying challenge is reconstructed from the same deterministic inputs.
This makes a same-day replay useful when you want to separate “I misunderstood the task” from “I made a performance error.” It does not turn the result into a standardized assessment. Device latency, browser scheduling, input method, fatigue, distraction and strategy still matter.
The Daily result reports overall accuracy, congruent accuracy, incongruent accuracy, neutral accuracy when applicable, median correct reaction time, total errors or misses and the observed conflict cost when enough correct condition observations are available. Conflict cost is the incongruent median correct RT minus the congruent median correct RT.
Do not reduce the whole block to that one difference. If incongruent accuracy is much lower, a speed difference alone misses part of what happened. If overall accuracy is low, an apparently fast median can simply reflect a speed–accuracy tradeoff. The interface keeps these measurements visible side by side so the player can see the tradeoff instead of receiving an opaque composite score.
Target Spotlight is available in Daily as it is in Practice. It adds a visible outline around the center target and dims flankers. Turning it on marks the result Assisted. The Daily completion can still be recorded because it is a real play session, but Assisted status remains attached to the history record and does not masquerade as a Clean best.
If your goal is to learn the rule, an Assisted replay can be useful. If your goal is to compare your own unassisted runs, leave Spotlight off and compare several Clean attempts rather than treating one date as definitive.
DailyBrainArc has a frozen Today’s 6 rotation whose stability is independently checked. Specialist Dailies are additional, opt-in workflows. Flanker therefore gets its own Daily route and local streak but remains outside the six-game sequence. This prevents a new specialist from silently changing a previously frozen year schedule or rewriting historical Today’s 6 records.
The separation also gives the product a clearer purpose. Today’s 6 is a mixed daily set. Daily Flanker is one focused selective-attention task for someone who specifically wants that mechanic.
The frozen Flanker bank is project-local and abstract: each trial stores a condition and a left/right center target, while the concrete symbol family and spacing are materialized later. The project does not copy a commercial assessment sequence, scrape an academic trial bank, retrieve a remote clinical test, or import population norms. No precomputed “attention score” or diagnosis table is used.
The independent verifier reconstructs the center-target response and the visual relationship between target and flankers. Congruent must visually agree with the center response, incongruent must visually disagree, and neutral flankers must carry no left/right answer mapping.
Use Daily when you want one fixed challenge attached to the date and a specialist streak. Use Practice when you want to choose Easy, Standard or Challenge and generate a Fresh seed. Practice seed URLs are stable enough to replay the exact same concrete block, while Fresh block deliberately changes the seed.
Both flows remain local-first. They use the same frozen source authority and transparent response rule. The difference is scheduling: Daily derives its block from the date, while Practice derives it from the chosen level and URL seed.
Stroop creates attribute interference, Go / No-Go tests frequent response versus withholding, and Reaction Time focuses on speed without flankers. Flanker specifically keeps the required center response constant while surrounding spatial distractors agree, disagree or remain neutral.