How to Calculate Training Monotony Index

In a landmark 2001 study, Foster et al. tracked 25 competitive athletes over a full season and found that weeks with a training monotony index above 2.0 were associated with a 3.5× higher incidence of illness and a measurable drop in performance benchmarks — even when total weekly load appeared reasonable. That single statistic explains why monotony tracking is no longer a niche academic exercise but a mainstream load-management tool.

This guide walks through the exact arithmetic behind Foster's training monotony index, explains what the numbers mean, shows how to design a training week that keeps monotony in a safe zone, and demonstrates how objective velocity data from an 800 Hz IMU sensor can make session-load estimates far more precise than RPE alone.

What Is Training Monotony?

Training monotony quantifies how similar every training day is to every other day within a week. A perfectly varied week — hard day, easy day, rest, moderate day — has low monotony. A week where every session carries the same load has high monotony, meaning the body never gets an easy day to consolidate adaptation.

The concept emerged from Foster's session-RPE (sRPE) method, which collapses a training session into a single internal load unit: sRPE (1–10 Borg CR-10) × session duration in minutes = session load (AU). Monotony and strain are then computed from the seven daily session loads within a training week.

Two metrics matter:

• Training Monotony (TM): Mean daily load ÷ Standard deviation of daily load. Higher values mean less day-to-day variation — more monotonous.
• Training Strain (TS): Weekly total load × Training monotony. Strain captures both the volume and the regularity of that volume, making it more predictive of illness than either metric alone (Foster, 2001).

Step-by-Step Calculation

Step 1 — Record Daily Session Load

For each day of the training week, immediately after the session, rate perceived exertion on the CR-10 Borg scale (0–10) and record session duration in minutes. Multiply them: Session Load (AU) = sRPE × Duration (min). On rest days, record 0 AU. Do not omit rest days — they are essential to computing standard deviation correctly.

Step 2 — Compute Mean Daily Load

Add all seven daily session loads and divide by 7: Mean Daily Load = Σ(Day₁ … Day₇) ÷ 7.

Step 3 — Compute Standard Deviation

Calculate the population standard deviation of the seven daily loads. Most spreadsheet apps (Excel, Google Sheets) can do this with =STDEV(A1:A7). A low standard deviation means daily loads cluster tightly around the mean — high monotony territory.

Step 4 — Compute Training Monotony

TM = Mean Daily Load ÷ Standard Deviation. If the standard deviation approaches zero (every day nearly identical), the formula produces very large values — a warning sign.

Step 5 — Compute Training Strain

TS = Weekly Total Load × TM where Weekly Total Load = Σ(Day₁ … Day₇). Strain values above 6,000 AU are consistently flagged in the literature as high-risk zones for non-contact injury or illness.

Worked Example

Mean = 2,515 ÷ 7 = 359.3 AU. SD ≈ 196.2 AU. TM = 359.3 ÷ 196.2 = 1.83. TS = 2,515 × 1.83 = 4,602 AU. Both values are within acceptable ranges for a competitive training week.

Threshold Values and Interpretation

Foster (2001) proposed thresholds that have been widely replicated across team sports (Putlur et al., 2004) and individual endurance athletes (Borresen & Lambert, 2009):

A high TM with a moderate total load is a subtler but still clinically important red flag: the athlete is repeating the same stimulus every day, blunting the hormetic adaptation signal. Conversely, a low TM combined with extremely high total load indicates high strain from volume alone.

Putlur et al. (2004) confirmed these thresholds in Division I female soccer players, finding that TM > 2.0 preceded increased URTI (upper respiratory tract infection) incidence by 5–7 days — a useful early-warning window for coaches to intervene with active rest or reduced session intensity.

Weekly Load Design to Control Monotony

The most practical lever for managing monotony is intentional load variation within the microcycle. A classic hard-easy-hard structure naturally reduces monotony by inserting low-load days between high-load sessions. Below is a sample weekly structure for a strength-power athlete who trains six days per week:

With this design, the high-day to low-day ratio keeps standard deviation elevated, which suppresses TM. Running the formula: mean ≈ 364 AU, SD ≈ 213 AU, TM ≈ 1.71 — solidly in the caution zone rather than the high-risk zone, and can be pushed to low-risk territory by lengthening Thursday's recovery emphasis.

A four-week mesocycle perspective matters too. Weeks 1–3 can progressively increase weekly total load (+8–12% per week), but TM should be recalculated each Sunday and should not trend upward across consecutive weeks. Week 4 deload cuts volume 40–50% while preserving intensity, which automatically lowers both TM and TS and allows structural adaptation to consolidate (Meeusen et al., 2013).

Using IMU Velocity Data to Sharpen Session Load

The classic sRPE method asks athletes to rate how hard a session felt 30 minutes after it ended — a process that is vulnerable to mood, motivation, and the well-documented retrospective positivity bias in elite athletes (Borg, 1998). Velocity-based training data provides an independent, objective proxy for internal load that can either validate or challenge the sRPE score.

Specifically, mean concentric velocity (MCV) at a given absolute load declines predictably as cumulative fatigue accumulates. If Monday's back squat at 100 kg averages 0.72 m/s but Thursday's identical load only produces 0.61 m/s, that 15% velocity drop indicates meaningful neuromuscular fatigue — even if the athlete rates Thursday's session as RPE 6 (same as Monday). Substituting the velocity-derived load estimate into the monotony formula corrects for this underreporting.

A practical protocol: record sRPE as usual, but flag any session where MCV at the reference load dropped >10% from the weekly baseline. For those sessions, add 1–2 points to the sRPE before computing session load. This prevents a falsely low TM from masking genuine fatigue accumulation across the week.

Common Calculation Mistakes

• Omitting rest days: Including only training days in the calculation artificially inflates the standard deviation and deflates TM, giving a falsely safe reading. Always include all seven days, with 0 AU on rest days.
• Using session duration as a proxy for load: Duration alone ignores intensity. A 90-minute technical session (sRPE 3) and a 90-minute max-effort session (sRPE 9) produce loads of 270 AU vs. 810 AU. The difference is performance-critical.
• Collecting sRPE immediately post-session: Foster's original protocol specifies 30 minutes post-session to allow acute sympathetic arousal to subside and RPE to stabilize. Immediate post-session ratings tend to be inflated by 0.5–1.0 points on the CR-10 scale.
• Treating thresholds as absolute: TM thresholds of 1.5 and 2.0 are population means derived mostly from endurance athletes. Collision sport athletes or those in high-fatigue mesocycles may have individually calibrated thresholds. Track each athlete's own TM-illness relationship over at least one full season before applying cut-offs mechanically.
• Ignoring cumulative strain across weeks: A single week of TM = 1.9 is manageable. Four consecutive weeks of TM = 1.9 is a different physiological reality. Calculate a rolling 4-week strain average alongside the weekly metric.