How to calculate training load with sRPE

sRPE (session-RPE) is calculated by multiplying the effort score a player gives the whole session, on Borg’s CR-10 scale, by the session’s duration in minutes. The result, in arbitrary units (AU), estimates internal load. In football it has validity support against heart rate, but it measures something different from GPS and does not replace it (Impellizzeri et al., 2004; Impellizzeri et al., 2019).

What internal load is and why sRPE

Internal load is the player’s psychophysiological response to training; external load is the work performed. They are distinct constructs, not interchangeable, and neither is perfectly predicted from the other (Impellizzeri et al., 2019). That is why sRPE is read alongside GPS rather than instead of it.

A review of 36 studies confirmed “validity and good reliability and internal consistency” of the method across several sports and ages. It can be used as a stand-alone method, although some authors recommend combining it with physiological parameters such as heart rate (Haddad et al., 2017).

The calculation, step by step

1. Ask. The player gives a single global score for the whole session. The original question is “How was your workout?” (Foster et al., 2001), on Borg’s CR-10 scale with the verbal anchors modified by Foster (Haddad et al., 2017). RPE is that score.
2. Multiply by the minutes. Haddad et al. (2017) give this example: an 87-minute session with an RPE of 4 gives 87 × 4 = 348 AU.

When and how to ask

Foster asked for the score about 30 minutes after the end of the exercise (Foster et al., 2001). Impellizzeri et al. (2004) applied it in football with the same rule: the data were collected about 30 minutes after each session “to ensure that the perceived effort was referred to the whole session rather than the most recent exercise intensity”.

Published club practice varies slightly. The staff of a Premier League club collected it about 20 minutes later, from each player separately, on the CR-10 (Gaudino et al., 2015; Scott et al., 2013).

The 30-minute window is flexible, and the evidence is mixed. Rodríguez-Marroyo, Blanco and Foster (2022) conclude that sRPE is valid between 5 minutes and 72 hours after the session. It differs from the 30-minute score only in hard sessions (sRPE > 5), when asked immediately or 1–4 weeks later; in easy or moderate sessions the timing makes no difference. Note that this is a study of 25 athletes from other sports, not football. It follows that what matters most is not changing the rule mid-season, especially for hard sessions.

How it is validated against heart rate and GPS

• Heart rate. In 19 young footballers and 479 sessions, individual correlations between sRPE and three heart-rate-based methods were significant, from r = 0.50 to r = 0.85 (p < 0.01). The authors call it “a good indicator of global internal load of soccer training” (Impellizzeri et al., 2004). That study was not in adult professionals.
• GPS in elite players. In 22 Premier League players (1,892 sessions, mean 57 ± 16 min), adjusted correlations between RPE-based load and external variables were r = 0.11 with HSR (>14.4 km/h), 0.45 with impacts and 0.37 with accelerations (>3 m/s²); the authors describe them as “moderate” predictors (Gaudino et al., 2015). In another study of professionals (15 players, 97 sessions), total distance, low-speed activity and player load correlated with sRPE at r = 0.71–0.84, and HSR and VHSR at r = 0.40–0.67 (Scott et al., 2013).

The r = 0.11 with HSR illustrates why sRPE does not replace GPS: it measures internal load, and external load tells another part of the story.

From the session to the week: load, monotony and strain

Foster proposed three weekly calculations from the daily scores (Foster, 1998; Clemente et al., 2020):

• Weekly load = sum of the daily scores.
• Monotony = daily mean ÷ standard deviation.
• Strain = weekly load × monotony.

Foster’s example training diary gives a weekly load of 3,400 AU, monotony of 1.26 and strain of 4,284 (Foster et al., 2001); 3,400 × 1.26 = 4,284. High monotony indicates little variation within the week (Clemente et al., 2020).

What do they predict? In 25 athletes (not footballers), a high percentage of illnesses was explained when each athlete exceeded individual thresholds, mainly of strain (Foster, 1998). That is the evidence: individual thresholds, in a different population, about illness rather than injury.

What values are seen in professional squads

The published ranges come from specific teams and are not a universal reference:

• In an elite European team, mean daily sRPE was 331 AU in mesocycle 1 and 239 AU in mesocycle 10. All internal and external variables were significantly lower the day before the match (p < 0.01) (Oliveira et al., 2019).
• In a top elite European team, comparing starters and non-starters, sRPE monotony had maximums of 7.2 and 7.0 in week 1 and a minimum of 1.5. Maximum strain was 8,498 AU in starters and 15,263.9 AU in non-starters, in week 41, with no significant differences between groups (Oliveira et al., 2021).
• In 20 professional players over 20 weeks, monotony and strain were higher in the middle and at the end of the season. The correlations between internal and external measures change sign depending on the period (r = +0.608 and −0.506 for monotony), which supports always measuring together rather than relying on a single metric (Nobari et al., 2022).

Limits and common mistakes

• There is no magic threshold. No universal threshold for monotony (for example > 2.0) or for “dangerous” weekly AU was verified. Foster speaks of individual thresholds, in athletes who were not footballers (Foster, 1998).
• There is no exact minute. 30 minutes is the original, but the evidence allows anything from 5 minutes to 72 hours, with caveats for hard sessions (Rodríguez-Marroyo et al., 2022).
• It does not replace GPS. It measures internal load; its correlation with HSR was low (Gaudino et al., 2015; Impellizzeri et al., 2019).
• It is not validated against heart rate in adult elite players. The classic football study is in 19 young players (Impellizzeri et al., 2004).
• It has not been shown to predict injury. Foster’s association was with illness in 25 athletes (Foster, 1998).
• There are no “normal” AU values for a squad. They depend on the club, the calendar and minutes played (Oliveira et al., 2019; Oliveira et al., 2021).

In practice

A staff setting up an sRPE routine decides three things and keeps them fixed: who is asked (each player separately), how long after the session, and on which scale. Then each player is compared against their own history and against GPS, not against someone else’s table. In Athlo, the player records their RPE after each session from their phone, and staff can load responses in bulk. You can see how it works in Wellbeing and sRPE.

Sources

• Foster C, Florhaug JA, Franklin J, et al. (2001) “A new approach to monitoring exercise training” Journal of Strength and Conditioning Research 15(1):109–115.
• Impellizzeri FM, Rampinini E, Coutts AJ, Sassi A, Marcora SM (2004) “Use of RPE-based training load in soccer” Medicine & Science in Sports & Exercise 36(6):1042–1047.
• Foster C (1998) “Monitoring training in athletes with reference to overtraining syndrome” Medicine & Science in Sports & Exercise 30(7):1164–1168.
• Gaudino P, Iaia FM, Strudwick AJ, et al. (2015) “Factors influencing perception of effort (session-RPE) during elite soccer training” International Journal of Sports Physiology and Performance 10(7):860–864.
• Scott BR, Lockie RG, Knight TJ, Clark AC, Janse de Jonge XAK (2013) “A comparison of methods to quantify the in-season training load of professional soccer players” International Journal of Sports Physiology and Performance 8(2):195–202.
• 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.
• Rodríguez-Marroyo JA, Blanco P, Foster C, et al. (2022) “Expanding knowledge about the effect of measurement time on session rating of perceived exertion” Journal of Strength and Conditioning Research 37:230–233.
• Impellizzeri FM, Marcora SM, Coutts AJ (2019) “Internal and external training load: 15 years on” International Journal of Sports Physiology and Performance 14(2):270–273.
• Nobari H, Martins AD, Oliveira R, Ardigò LP (2022) “Seasonal variations of the relationships between measures of training monotony and strain in professional soccer players” Scientific Reports 12:10930. https://doi.org/10.1038/s41598-022-15278-4
• Clemente FM, Silva R, Castillo D, et al. (2020) “Weekly load variations of distance-based variables in professional soccer players: a full-season study” International Journal of Environmental Research and Public Health 17(9):3300.
• Oliveira R, Brito JP, Martins A, et al. (2019) “In-season internal and external training load quantification of an elite European soccer team” PLoS ONE 14(4):e0209393.
• Oliveira RFS, Vieira LHP, Martins AD, et al. (2021) “In-season internal and external workload variations between starters and non-starters, a case study of a top elite European soccer team” Medicina 57(7):645.