SyncState
Back to SyncState

Evidence

What this document is

It’s the scientific and methodological reference behind SyncState: why these variables are collected, how they should be read and what conclusions can’t be drawn from them.

It’s written for strength and conditioning coaches, heads of performance, coaches and academies who want to start monitoring without GPS, force plates or blood tests.

And it’s here for an uncomfortable reason: a valid instrument doesn’t make the product that uses it valid. What decides whether this is useful is the protocol, adherence, data quality and the professional judgement of whoever reads it. SyncState is not a medical device, does not replace external load and does not predict injuries.

Translated from the Spanish original, which is the authors’ version of record.

SportSync Institute · Article

Combined monitoring of internal load (sRPE) and perceived wellness in resource-limited teams: a narrative review of current evidence and new no-cost directions

Miguel Martínez-Moreno · Alejandro Segovia-Vilchez · Vincenzo Rago

Abstract

Resource-limited teams can’t afford external monitoring technology (GPS, force plates, blood tests), yet they account for most of the world’s athletes. This narrative review summarises the evidence on the two reference tools in these settings: internal load quantified through session RPE (sRPE) and single-item perceived wellness questionnaires.

The literature consistently supports (a) the validity of sRPE as a measure of internal load and of its weekly derivatives (monotony, strain), (b) the greater sensitivity of subjective over objective measures in reflecting the response to training, (c) interpretation against individual references (z-scores, traffic-light systems) rather than absolute thresholds, and (d) the conceptual fragility of the acute:chronic workload ratio (ACWR) as a predictor of injury.

The main barrier is not metric but practical: staff time, athlete adherence and closing the feedback loop. Emerging no-cost directions are discussed: differential RPE, countermovement jumps measured with validated mobile apps, heart rate variability through smartphone photoplethysmography, and the (still immature) role of machine learning. We conclude that the way forward is not to measure more, but to decide better: minimum viable systems, centred on the athlete and designed for the daily conversation.

Keywords: training load; sRPE; perceived wellness; athlete monitoring; team sport; limited resources; adherence.

1. Introduction

Training load monitoring has become standard practice in professional team sport (Akenhead & Nassis, 2016). However, most of the research has been carried out in elite settings with access to external load technology, while the vast majority of clubs operate on minimal budgets and small staffs. In these settings, practice is dominated by a combination of two pen-and-paper instruments: the session rating of perceived exertion scale (sRPE; Foster et al., 2001) and daily perceived wellness questionnaires (McLean et al., 2010; Hooper & Mackinnon, 1995).

This review has three aims: (1) to summarise the evidence on the validity and use of both instruments, (2) to examine how their data should be interpreted in day-to-day decision-making, and (3) to identify the barriers to implementation and the future directions available without a budget.

2. Materials and methods

A narrative review was carried out (not systematic; no PRISMA protocol was applied) through searches of PubMed/PMC, BJSM, IJSPP, Frontiers and key professional literature in August 2026, prioritising systematic reviews, meta-analyses, controlled trials and validation studies published between 1995 and 2026. Inclusion criteria: studies on sRPE, athlete self-report measures (ASRM), interpretation of monitoring data, the ACWR and practical implementation in team sport. As it isn’t systematic, the selection may contain reviewer bias.

3. Results

3.1. sRPE as a valid measure of internal load, and the choice of a 0-10 scale

The sRPE method uses Foster’s modified Category-Ratio 10 (CR10) scale and calculates a session’s internal load by multiplying the overall perceived effort by its duration in minutes: sRPE = RPE × duration. It was originally validated against heart-rate-based methods (Foster et al., 2001; Impellizzeri et al., 2004), and later reviews confirm its validity, reliability, ecological usefulness and low cost across sports, ages and competitive levels (Haddad et al., 2017; Foster et al., 2021).

The 0 to 10 scale isn’t an arbitrary range. It comes from Borg’s CR10 and uses anchors that are easy to understand, from 0, rest, to 10, maximal effort. It lets the athlete summarise, looking back, the perceived intensity of the whole session with a single overall question in a short format suited to everyday use. Using it requires familiarisation, stable verbal anchors, a consistent question and a collection protocol that avoids players influencing each other.

The result is expressed in arbitrary units. It describes the size of the perceived internal load, but it doesn’t measure external load or identify which technical, tactical, metabolic or mechanical component produced the effort. The same sRPE load can come from different combinations of intensity and duration, which is why SyncState keeps and shows RPE, minutes, session type and context separately. Daily load, weekly load, monotony and strain can be derived from these data, always as historical descriptors and never as diagnoses.

3.2. Subjective measures outperform objective ones

The systematic review by Saw et al. (2016) concluded that self-report measures reflect changes in the response to training with greater sensitivity and consistency than the usual objective measures. Short single-item questionnaires derived from Hooper and Mackinnon (1995) and McLean et al. (2010) are the dominant practice. However, Duignan et al. (2020) warn about the heterogeneity of scales, directions and anchors, and that adding items together into a single score cancels out their psychometric properties.

3.3. Wellness: scale direction and individual reference

There’s no universally superior short scale. The review by Duignan et al. (2020) found 5-, 7- and 10-point instruments and opposite directions: in 11 of 18 identifiable studies a high score meant better wellness, while in 7 it meant a worse state. The essential methodological condition isn’t a particular direction but keeping it consistent across every item, labelling the points with clear anchors and documenting it to avoid response and interpretation errors.

SyncState uses a 1 to 7 scale where 7 is the most favourable state and 1 the least favourable. It’s a design convention, not a universal scientific standard. It was chosen so that an upward trend always means something positive, and so that sleep, energy, muscle soreness, stress and mood all read the same way. Data from studies or tools that use the opposite direction must be recoded before comparing them.

Interpretation puts the player against themselves first. SyncState separates the recent 7-day trend from the rolling 28-day individual reference. Seven days roughly summarise the most recent microcycle; in regular weekly calendars, 28 days cover about four microcycles and reduce the proportional weight of an isolated response or an unusual week.

This window is operational, not a universal physiological boundary. Four weeks don’t guarantee that an adaptation has been completed, nor do they turn wellness or accumulated load into a direct measure of fitness: neural, metabolic, cardiovascular, muscular and connective tissue adaptations follow different time scales (Hughes et al., 2018). The 28 days are read as an individual pattern and recent history, not as evidence of physiological adaptation.

During start-up, the reference should show how mature it is: no reference for the first few days, an initial trend from one week, a provisional reference before four weeks are complete and an operational reference from 28 days only if adherence is sufficient. These states are conservative product rules, not clinical thresholds. Z-scores and traffic lights are statistical flags to check the context and start a conversation — never diagnoses or automatic decisions (Robertson et al., 2017).

3.4. The ACWR, and the operational choice of 7 and 28 days

The acute:chronic workload ratio (ACWR) compares recent load with a historical reference. The classic model uses 7 days as the acute period and 28 as the chronic one because they roughly represent one and four microcycles of exposure. This choice makes weekly calendars easy to read, but it doesn’t reflect a proven physiological boundary: the result changes depending on the chosen duration, the load variable, a rolling or exponential average and whether the most recent week is included in the denominator (Murray et al., 2017; Impellizzeri et al., 2020).

The historical association of acute load with fatigue and chronic load with fitness is conceptual, not a direct measurement of those states. Two players with the same 28-day load can show different adaptations, fatigue, capacity and tolerance. That’s why SyncState calls it recent exposure history, not fitness level or the load a player is prepared for.

Impellizzeri et al. (2020) documented conceptual problems, mathematical coupling and loss of information. The cluster randomised trial by Dalen-Lorentsen et al. (2021) found no preventive effect of ACWR-based load management. The most recent multilevel meta-analysis found a small to moderate association, but with very high heterogeneity — not enough to justify individual prediction or universal thresholds (Frontiers in Public Health, 2026).

SyncState therefore prioritises showing the last 7 days’ load, the 28-day history and their trends separately. If the ratio is kept, it must be labelled as a contextual indicator, with its formula and data availability explained, and without safe zones, injury language or automatic recommendations.

3.5. Implementation as the critical barrier

Akenhead and Nassis (2016) identified limited human resources and coach buy-in as the main barriers perceived by practitioners in high-level football. Saw et al. (2015) showed that successful implementation depends on social and system factors: a short instrument, a consistent collection time, athlete education and, above all, visibly feeding the information back. Without feedback, adherence drops and responses start being faked.

3.6. Weekly load, monotony, strain and calculation rules

Weekly load is the sum of the period’s sRPE loads. Foster’s monotony is calculated as the mean daily load divided by its standard deviation; strain multiplies weekly load by monotony. They’re useful for telling apart weeks with a similar total load but different day-to-day variability.

Monotony describes numerical regularity, not real variety of content. A week with different drills can come out as monotonous if duration and RPE are similar, and the calculation can end up dominated by duration. The historical value of 2.0 isn’t a universal normative threshold: the original research observed associations with illness when individual thresholds were exceeded, but didn’t validate a risk zone that transfers across sports, teams or players (Foster, 1998).

SyncState keeps weekly load, monotony and strain as descriptors of exposure and distribution, with these rules on show:

  • Planned rest days count as zero load.
  • Missing data stays missing and is never turned into a zero.
  • Double sessions are added together into one daily load before the week is calculated.
  • If the standard deviation is zero or close to zero, monotony is undefined or unstable and shouldn’t be shown as an ordinary value.
  • Colour bands, if used, are configurable descriptive references — not normative scales, diagnoses or injury predictors.
  • RPE, duration, calculated load, session type and context stay accessible so the aggregated indicator never hides where it came from.

The same caution applies to absolute load bands: easy, moderate or hard sessions aren’t universal categories, because they depend on duration, type of activity, age, phase of the season, involvement and the individual pattern.

4. Discussion

This review points to a deep shift in the field: the relative failure of the predictive project and the maturity of the decision-making one. For a decade, the implicit promise was that enough load data would anticipate injury; the ACWR case shows that this promise was built on statistical artefacts (Impellizzeri et al., 2020) and didn’t survive experimental testing (Dalen-Lorentsen et al., 2021). Even machine learning, the modern version of the same promise, shows no consistent advantage over logistic regression so far and suffers from small samples and a lack of external validation (Van Eetvelde et al., 2021; Bullock et al., 2022). The way forward isn’t to predict, but to spot things earlier and have better conversations (Robertson et al., 2017).

That leads to three practical consequences for settings without a budget:

  1. 1Subjective data stops being the cheap plan B and becomes the evidence-based plan A (Saw et al., 2016). The advantage doesn’t lie in buying sensors, but in designing the instrument well and protecting the process.
  2. 2The quality of the process matters more than how sophisticated the metric is: systems die from friction and mistrust, not from a lack of variables (Saw et al., 2015; Akenhead & Nassis, 2016).
  3. 3Less is more. A well-run minimum viable system (sRPE + 5 wellness items + explicit rules for missing data) beats an ambitious system abandoned in November.

New trends and tools at no meaningful cost

  • Differential RPE (dRPE): separates breathing, muscular and technical-cognitive effort using the same scale and at no cost (McLaren et al., 2016; James et al., 2021).
  • Countermovement jumps measured with mobile apps validated against force plates, as a one-off marker of neuromuscular fatigue (Balsalobre-Fernández et al., 2015; Sharma et al., 2023).
  • Heart rate variability through smartphone photoplethysmography, with acceptable agreement with ECG for RMSSD in short recordings (Plews et al., 2017).
  • Digital forms that feed back automatically the individual trend, putting the best-documented adherence factor into practice (Saw et al., 2015).

None of them replaces the sRPE + wellness pairing: they add to it as optional layers.

The final reflection is about the moral economy of data: in modest clubs, every minute of staff time and every player response is the scarce resource. Every piece of data that’s collected and not used in a visible decision erodes the trust the system depends on. The right design question isn’t “what else can we measure?” but “which decision tomorrow morning gets better with this data?”

5. Conclusions

  1. 1sRPE and single-item wellness questionnaires are the tools with the best evidence-to-cost ratio for load monitoring in resource-limited teams.
  2. 2Wellness should be interpreted item by item and against the player’s own pattern. SyncState uses 1-7, with 7 as the favourable state, and separates a recent 7-day trend from an operational 28-day individual reference.
  3. 3The 28 days cover roughly four microcycles and reduce the weight of isolated fluctuations, but they aren’t a universal physiological minimum or a direct measure of fitness.
  4. 4Weekly load, monotony and strain are useful descriptors of exposure and distribution. Their bands shouldn’t be presented as normative scales, diagnoses or injury predictors.
  5. 5The ACWR shouldn’t be used as an automatic predictor, a safe zone or a prescription rule. If it’s kept, it should be shown as a contextual indicator alongside the raw 7- and 28-day data.
  6. 6The critical barrier is implementation: brevity, consistent timing, education, adherence, rules for missing data and visible feedback decide how useful the system is.
  7. 7Emerging no-cost directions (dRPE, CMJ with validated apps, smartphone HRV and automated feedback) should be added as optional layers on top of a well-run minimum core — never as substitutes.

References

  • Akenhead, R., & Nassis, G. P. (2016). Training load and player monitoring in high-level football: Current practice and perceptions. International Journal of Sports Physiology and Performance, 11(5), 587-593. View
  • Balsalobre-Fernández, C., Glaister, M., & Lockey, R. A. (2015). The validity and reliability of an iPhone app for measuring vertical jump performance. Journal of Sports Sciences, 33(15), 1574-1579. View
  • Buchheit, M. (2025). Sports science 3.0: Ferrari tools, tractor jobs? Sport Performance & Science Reports. View
  • Bullock, G. S., Mylott, J., Hughes, T., Nicholson, K. F., Riley, R. D., & Collins, G. S. (2022). Machine learning for understanding and predicting injuries in football. Sports Medicine - Open, 8, 73. View
  • Dalen-Lorentsen, T., Bjørneboe, J., Clarsen, B., Vagle, M., Fagerland, M. W., & Andersen, T. E. (2021). Does load management using the acute:chronic workload ratio prevent health problems? A cluster randomised trial of 482 elite youth footballers. British Journal of Sports Medicine, 55(2), 108-114. View
  • Duignan, C., Doherty, C., Caulfield, B., & Blake, C. (2020). Single-item self-report measures of team-sport athlete wellbeing and their relationship with training load: A systematic review. Journal of Athletic Training, 55(9), 944-953. View
  • Foster, C. (1998). Monitoring training in athletes with reference to overtraining syndrome. Medicine & Science in Sports & Exercise, 30(7), 1164-1168. View
  • Foster, C., Florhaug, J. A., Franklin, J., Gottschall, L., Hrovatin, L. A., Parker, S., Doleshal, P., & Dodge, C. (2001). A new approach to monitoring exercise training. Journal of Strength and Conditioning Research, 15(1), 109-115. View
  • Foster, C., Boullosa, D., McGuigan, M., Fusco, A., Cortis, C., Arney, B. E., et al. (2021). 25 years of session rating of perceived exertion: Historical perspective and development. International Journal of Sports Physiology and Performance, 16(5), 612-621. View
  • Gastin, P. B., Meyer, D., & Robinson, D. (2013). Perceptions of wellness to monitor adaptive responses to training and competition in elite Australian football. Journal of Strength and Conditioning Research, 27(9), 2518-2526. View
  • Haddad, M., Stylianides, G., Djaoui, L., Dellal, A., & Chamari, K. (2017). Session-RPE method for training load monitoring: Validity, ecological usefulness, and influencing factors. Frontiers in Neuroscience, 11, 612. View
  • Hooper, S. L., & Mackinnon, L. T. (1995). Monitoring overtraining in athletes: Recommendations. Sports Medicine, 20(5), 321-327.
  • Hughes, D. C., Ellefsen, S., & Baar, K. (2018). Adaptations to endurance and strength training. Cold Spring Harbor Perspectives in Medicine, 8(6), a029769. View
  • Impellizzeri, F. M., Rampinini, E., Coutts, A. J., Sassi, A., & Marcora, S. M. (2004). Use of RPE-based training load in soccer. Medicine & Science in Sports & Exercise, 36(6), 1042-1047. View
  • Impellizzeri, F. M., Tenan, M. S., Kempton, T., Novak, A., & Coutts, A. J. (2020). Acute:chronic workload ratio: Conceptual issues and fundamental pitfalls. International Journal of Sports Physiology and Performance, 15(6), 907-913. View
  • James, C., Dhawan, A., Jones, T., & Girard, O. (2021). Minimal agreement between internal and external training load metrics across a 2-wk training microcycle in elite squash. Journal of Sports Science and Medicine, 20(1), 101-109. View
  • McLaren, S. J., Graham, M., Spears, I. R., & Weston, M. (2016). The sensitivity of differential ratings of perceived exertion as measures of internal load. International Journal of Sports Physiology and Performance, 11(3), 404-406. View
  • McLean, B. D., Coutts, A. J., Kelly, V., McGuigan, M. R., & Cormack, S. J. (2010). Neuromuscular, endocrine, and perceptual fatigue responses during different length between-match microcycles in professional rugby league players. International Journal of Sports Physiology and Performance, 5(3), 367-383. View
  • Menold, N., & Bogner, K. (2016). Design of rating scales in questionnaires (GESIS Survey Guidelines). GESIS - Leibniz Institute for the Social Sciences. View
  • Murray, N. B., Gabbett, T. J., Townshend, A. D., & Blanch, P. (2017). Calculating acute:chronic workload ratios using exponentially weighted moving averages provides a more sensitive indicator of injury likelihood than rolling averages. British Journal of Sports Medicine, 51(9), 749-754. View
  • Meta-análisis multinivel sobre la asociación ACWR-lesión. (2026). Frontiers in Public Health, 14. View
  • Plews, D. J., Scott, B., Altini, M., Wood, M., Kilding, A. E., & Laursen, P. B. (2017). Comparison of heart-rate-variability recording with smartphone photoplethysmography, Polar H7 chest strap, and electrocardiography. International Journal of Sports Physiology and Performance, 12(10), 1324-1328. View
  • Robertson, S., Bartlett, J. D., & Gastin, P. B. (2017). Red, amber, or green? Athlete monitoring in team sport: The need for decision-support systems. International Journal of Sports Physiology and Performance, 12(S2), S2-73-S2-79. View
  • Saw, A. E., Main, L. C., & Gastin, P. B. (2015). Monitoring athletes through self-report: Factors influencing implementation. Journal of Sports Science and Medicine, 14(1), 137-146. View
  • Saw, A. E., Main, L. C., & Gastin, P. B. (2016). Monitoring the athlete training response: Subjective self-reported measures trump commonly used objective measures: A systematic review. British Journal of Sports Medicine, 50(5), 281-291. View
  • Sharma, D., et al. (2023). Validity and reliability of “My Jump app” to assess vertical jump performance: A meta-analytic review. Scientific Reports. View
  • Estudio observacional de fiabilidad y validez concurrente de dispositivos de HRV en deportistas. (2025). Frontiers in Physiology, 16. View
  • Van Eetvelde, H., Mendonça, L. D., Ley, C., Seil, R., & Tischer, T. (2021). Machine learning methods in sport injury prediction and prevention: A systematic review. Journal of Experimental Orthopaedics, 8, 27. View

SportSync Institute, 2026.