MUNICH • 11 OCTOBER 2026

Road to 3:20

Planned sessions, completed work, and what it means for race day.

Intervals.icu sync Loading…

GOAL READINESS

Loading

— / 100
Readiness trend —
—

SELECTED WEEK

—planned km
—completed km
—completion
Running time by HR zone —
—
Key sessions in this week

CURRENT TRAINING STATE

Loading

—
Long-term fitness—
Short-term fatigue—
Weekly fitness change—

CURRENT RACE POTENTIAL

Fresh-day performance prediction

Flat course, normal weather and a proper taper—not a promise or today’s fatigued result.

—
Marathon—
—
—
Half marathon—
—
—
10K—
—
—
5K—
—
—

Loading model evidence…

RUNNING HEART-RATE ZONES

Your working intensity map

— maximum HR

VOLUME TRAJECTORY

Planned versus completed

Core plan 100K stretch Actual

DAILY SCHEDULE

Week

WORKOUT ASSESSMENTS

Latest completed work

BEFORE THE BLOCK

Starting position

HOW THE GAUGE WORKS

Direction, not destiny

The score is not a probability of running 3:20. It combines plan adherence, quality-session execution, volume progression, marathon-specific readiness, and fitness/recovery. The automated assessment updates the components from Intervals.icu evidence.

A strong score can still be overridden by injury, illness, extreme weather, or a failed dress rehearsal.

HOW THE PREDICTION WORKS

Two models, one cautious estimate

The 5K, 10K and half-marathon estimates come from a robust, recency-weighted speed–duration curve fitted to your strongest recent evidence. Your own fatigue exponent is estimated from the data rather than fixed to a generic lookup table.

The marathon estimate blends that speed potential with an independent eight-week model based on actual running volume and mean training pace. This prevents a fast interval session from pretending that marathon durability already exists.

Races and time trials carry the most weight. Structured intervals are recovery-discounted and lower-confidence. Heat-affected or clearly non-maximal efforts are down-weighted, and every daily refresh preserves the historical prediction trend.