Predict your next
2K with confidence
Recent benchmark + current state + taper window. Get a forecast 2K with confidence band and a tailored taper plan, anchored to Concept2 power ratios.
Quick answer
Convert your recent benchmark to 2K-equivalent watts via Concept2 ratios, multiply by your current-state factor (+1.5% tapered, −2.5% fatigued), apply a small staleness drift if the test is >30 days old, then convert back to a 2K time. Confidence band of ±2.5% on watts captures normal day-to-day variation. The taper plan adjusts to your “days until 2K”.
Frequently asked questions
How does the forecast work?
Three steps. (1) We convert your recent benchmark to 2K-equivalent power using empirical Concept2 ratios (1K → ÷1.10, 5K → ÷0.78, 6K → ÷0.74, 30:00 → ÷0.70, etc.). (2) We apply a state multiplier: tapered athletes peak ~1.5% above baseline, fatigued athletes sit ~2.5% below. (3) If the benchmark is more than 30 days old we apply a small staleness drift (~−0.05% per day past 30) to reflect typical fitness variation between tests.
Why is this different from "Test → 2K Equivalent"?
That calculator is a pure ratio: it tells you what 2K time matches a given test under identical conditions. The forecast layers in your current state and the time elapsed since the test, so the prediction reflects what you can actually pull on a target day rather than what you could have pulled the day of the benchmark.
How accurate is the confidence band?
±2.5% on watts is the typical day-to-day variation for trained rowers when state and pacing are honestly estimated. Real outcomes can fall outside this band when sleep is poor, when the venue is hot or cold, when the erg drag factor differs from training conditions, or when the athlete races a poorly-paced piece (positive split, blow-up at 1500m). The band is a 1-sigma honest expectation, not a hard floor or ceiling.
How should I pick "Tapered / Normal / Fatigued"?
Tapered = at least 3 days of reduced volume, including a rest day, and you feel snappy on the warmup. Normal = your usual training pattern, no recent rest week, current sessions feel routine. Fatigued = inside a high-volume block (>16h/week for serious athletes) or coming off a heavy lifting / capacity phase with no sharpening. If unsure, pick Normal — it returns the cleanest baseline without flattering or punishing the prediction.
What about the taper plan?
A coaching-style suggestion based on your "days until target" input — short tapers (≤2 days) emphasise nervous-system priming, mid-window tapers (3–5 days) cut volume to ~50% and add a single sharpening session, longer windows (6+ days) maintain quality through the week and start cutting volume in the final 7–10 days. Treat it as a starting point; specific volumes depend on your training age and history.
Will you ever build the proper Bayesian forecast?
Once we have saved erg data per athlete (recent UT2 averages, training-load, freshness signals from heart-rate variability, etc.), the forecast can update via Bayes from each new session — a true posterior on race-day 2K rather than this single-test heuristic. That requires the saved-sessions backend, which is a follow-up effort.