CICADAMATA sphere 1-2 speedrun: Route & Timing Guide

The available leaderboard data identifies the relevant segment as Level 1-2 MANDALA, listed under the Any% Unrestricted ruleset. A run by sukie is recorded at 20.983 seconds, making this the clearest benchmark for focused practice on the requested section.

The same leaderboard page also records a Fullgame any% Demo run in 2:36.530 in-game time. Its displayed run time is 4:44.030, so runners should not compare those two values as if they were the same timing category. The moderator note explains that the in-game total was retimed from individual level segments.

For route development, treat the MANDALA time as a segment benchmark rather than proof that every attempt should immediately match it. A strong practice route first produces repeatable movement, then removes hesitation, unnecessary positioning, and failed transitions.

Reference: CICADAMATA Fullgame any% Demo leaderboard run

A fast segment is usually built around a small number of decisions: the opening movement, the safest line through the level, and the final action that ends the split. Because the available record data names the level and ruleset but does not provide an input-by-input route, this guide avoids inventing specific jumps, skips, or glitches.

Instead, divide MANDALA into repeatable checkpoints. Each checkpoint should answer one question: where should the character be, what action should be queued next, and which mistake causes an immediate reset?

Use a simple route notebook or spreadsheet with three columns: attempt number, split time, and failure reason. “Slow” is not a useful diagnosis. “Late turn at midpoint” or “hesitated before finish” gives you a specific correction for the next block.

The unrestricted ruleset is also useful for testing. If a faster line appears possible, compare it against a safer route over multiple attempts. A route that saves a fraction of a second but fails frequently may reduce full-session performance.

Speedrun improvement depends on measuring the right result. The leaderboard page distinguishes In-Game Time from Time, and the fullgame entry includes a moderator retiming note based on level segments. This distinction is especially important when comparing a local timer, an in-game display, and a submitted leaderboard result.

Create a split table for MANDALA with at least three fields: best segment, average segment, and reset count. A best time shows potential, while the average reveals whether the route is practical during a fullgame attempt.

Use short practice blocks rather than endless full attempts. After several resets at the same checkpoint, pause and rehearse the failed transition without pursuing a complete run. This keeps practice focused and reduces the tendency to repeat the same mistake.

The listed MANDALA result is a useful target, but speed comes from controlling variation. Start each attempt with the same setup, use the same route notes, and avoid changing several inputs at once. When a new line looks faster, test it in a separate practice block so your established route remains available.

A practical improvement cycle looks like this:

• Establish a baseline using the same route for several attempts.
• Identify the single most common failure point.
• Practice that point in isolation.
• Re-run the level without adding new changes.
• Compare the average, not only the fastest result.

For runners moving from segment practice into a fullgame category, remember that a fast MANDALA split is only one part of the larger run. The fullgame listing includes several level segments, and the moderator’s retiming note demonstrates why individual level performance must be added and interpreted consistently.

Use this checklist before submitting a new attempt or comparing your time with the leaderboard benchmark.

Before using a leaderboard result as a goal, confirm that your own timer and category match the relevant listing. The recorded page identifies the platform as PC, while the exact route details are not included in the supplied entry.