Reading the Force Curve: What Repeatability Looks Like

Abstract

The PM5 force curve is one input to the rower's diagnosis of stroke quality, not the diagnosis itself. The [1] Kleshnev 2020 rowing-kinetics handbook established the operational read: per-stroke drive-time, recovery-time, peak-force, average-force, drive-length, and handle-speed-force curves are the kinetic events the PM5 reports; the rower's read of those events is the diagnostic ([1] Kleshnev 2020, Level 5). The shape that matters most is repeatability: every stroke should look like every other stroke. The [5] Smith & Loschner 2005 biomechanical review in Sports Biomechanics placed the same shape on the rowing-specific side ([5] Smith & Loschner 2005, Level 5). The [22] Cosgrove et al. 1999 Journal of Sports Sciences study placed it on the rate-band side: the force curve's shape changes with stroke rate; the diagnostic value of the curve is rate-dependent ([22] Cosgrove et al. 1999, Level 2b).

The [18] Barrett & Manning 2004 fatigue study in JSS established the variability-on-fatigue finding: stroke-to-stroke variability rises with fatigue; the force curve is the early warning ([18] Barrett & Manning 2004, Level 2b). The [25] Halson 2014 training-load monitoring review in Sports Medicine placed the same shape on the multi-modal signal: the constellation of HR + sRPE + force-curve consistency is the load-bearing signal ([25] Halson 2014, Level 5).

For the indoor rower, the [3] Concept2 PM5 documentation and the [4] Concept2 stroke-sequence reference are the operational anchors. The [5] Smith & Loschner 2005 review and the [7] Soper & Hume 2004 kinematic-chain study are the academic anchors. The [17] Smith & Hopkins 2012 rowing-performance-measurement review in Sports Medicine established which per-stroke variables predict 2K time; the diagnostic value of each metric sits on that ranking ([17] Smith & Hopkins 2012, Level 5).

The honest read for the rower: the force curve is a feedback channel, not a verdict. The [32] Magill 2011 motor-learning textbook placed feedback on the policy side: KR frequency, precision, and timing shape retention ([32] Magill 2011, Level 5). The [33] Schmidt & Lee 2011 textbook reached the same conclusion from the rower's perspective. The rower who reads the force curve rower-by-rower is the rower whose stroke improves; the rower who reads it textbook-curve-by-textbook-curve is the rower whose learning plateaus. The article below is the framework for reading the curve yourself.

The premise: the force curve is a feedback signal, not a verdict

The PM5 reports a per-stroke force curve on every stroke. The curve plots handle force against time across the drive and recovery phases, and the [1] Kleshnev 2020 rowing-kinetics handbook chapter placed the operational read on the kinetic-event side ([1] Kleshnev 2020, Level 5). The four kinetic events the rower reads are: catch (the brief dip at the start of the drive), drive (the rising curve to peak force), finish (the drop at the end of the drive), and recovery (the low-force portion before the next catch). The [4] Concept2 stroke-sequence reference places the same events on the manufacturer-canonical side ([4] Concept2, Level 5).

The [5] Smith & Loschner 2005 biomechanical review in Sports Biomechanics sorted the same four events into a kinematic chain — legs, back, arms in sequence — and showed that the force curve's shape depends on the timing of each segment ([5] Smith & Loschner 2005, Level 5). The [7] Soper & Hume 2004 Sports Biomechanics paper placed the kinematic chain on the rowing-specific side: legs-back-arms sequencing determines the shape of the rising edge of the drive; the finish shape depends on the arms-back-legs reverse sequence ([7] Soper & Hume 2004, Level 5).

The operational premise: the force curve is one input to the rower's read of stroke quality. It is not the read itself. The [32] Magill 2011 motor-learning textbook placed feedback on the policy side: feedback is a variable with frequency, precision, and timing as the policy knobs ([32] Magill 2011, Level 5). The [33] Schmidt & Lee 2011 textbook reached the same conclusion from the rower's perspective: knowledge-of-results (KR) frequency, precision, and timing shape retention; degraded feedback degrades retention ([33] Schmidt & Lee 2011, Level 5). The rower who uses the force curve rower-by-rower is the rower whose stroke improves; the rower who chases an ideal curve from a textbook is the rower whose learning plateaus.

What the force curve reports

The [3] Concept2 PM5 documentation enumerates the per-stroke readouts: drive time (the duration of the drive phase), recovery time (the duration of the recovery phase), drive length (the distance covered during the drive), peak force (the maximum handle force during the drive), average force (the mean handle force during the drive), stroke rate (strokes per minute), and 500m split (the current projected split). The [1] Kleshnev 2020 handbook chapter placed the same readouts on the rowing-kinetics side ([1] Kleshnev 2020, Level 5).

The [20] Roth et al. 1993 International Journal of Sports Medicine paper placed the same events on the force-time-characteristics side: drive-time, recovery-time, and peak-force are distinct kinetic events, each with its own signature in the force curve ([20] Roth et al. 1993, Level 2b). The [17] Smith & Hopkins 2012 rowing-performance-measurement review in Sports Medicine sorted the variables on the predictive side: which per-stroke variables predict 2K time, and which do not ([17] Smith & Hopkins 2012, Level 5).

The five variables the rower reads most often:

Drive time. The duration of the drive phase. The [1] Kleshnev 2020 handbook places drive time in the 0.7–0.9-second range for most rowers; the [3] Concept2 PM5 documentation reports it as the primary kinetic event. Drive time is rate-dependent: faster stroke rates compress the drive. The [6] Hofmijster et al. 2021 rate-band study placed drive length on the rate-dependent side ([6] Hofmijster et al. 2021, Level 1b/2b).

Recovery time. The duration of the recovery phase. The [1] Kleshnev 2020 handbook places recovery time in the 1.1–1.4-second range for most rowers; longer recovery at lower rates, shorter recovery at higher rates. Recovery time is the kinetic event the rower reads for the slide-control side of the stroke.

Drive length. The distance covered during the drive. The [10] Wing & Woodburn 1995 JSS pulldown study placed drive length on the leg-extension side ([10] Wing & Woodburn 1995, Level 2b). The [6] Hofmijster et al. 2021 rate-band study placed drive length on the rate-dependent side. Drive length is the variable the rower reads for the leg-extension leg of the kinematic chain.

Peak force. The maximum handle force during the drive. The [3] Concept2 PM5 documentation reports peak force as a real-time readout. The [19] Wilson et al. 2010 rate-band study in IJSPP placed peak force on the rate-dependent side: peak force drops as rate increases; the relationship is the rate-band diagnostic ([19] Wilson et al. 2010, Level 1b/2b). Peak force is the variable the rower reads for the leg-strength side of the stroke.

Average force. The mean handle force during the drive. The [3] Concept2 PM5 documentation reports average force as a real-time readout. The [17] Smith & Hopkins 2012 review placed average force on the predictive side for 2K time ([17] Smith & Hopkins 2012, Level 5). Average force is the variable the rower reads for the work-per-stroke side.

Drive ratio. The drive-time / recovery-time ratio. The [1] Kleshnev 2020 handbook places the typical drive ratio at 1:1.6 to 1:2.0; the [22] Cosgrove et al. 1999 study placed the drive ratio on the rate-dependent side ([22] Cosgrove et al. 1999, Level 2b). Drive ratio is the variable the rower reads for the leg-vs-slide balance of the stroke.

The four force-curve shapes

The [3] Concept2 PM5 documentation enumerates four force-curve shapes the rower can recognize: the rounded curve (the consistent, healthy stroke), the sharp-peak curve (a leg-strength-dominant stroke with a weak finish), the dip-in-the-middle curve (a sequencing fault), and the slow-release curve (an arms-dominant stroke with a late finish). The [5] Smith & Loschner 2005 biomechanical review placed the same four shapes on the rowing-specific side ([5] Smith & Loschner 2005, Level 5).

The [7] Soper & Hume 2004 kinematic-chain paper placed the four shapes on the sequencing side: legs-back-arms sequencing produces the rounded curve; a missing back segment produces the sharp-peak curve; a missing arms segment produces the dip-in-the-middle curve; a missing legs-back reverse at the finish produces the slow-release curve ([7] Soper & Hume 2004, Level 5).

The [10] Wing & Woodburn 1995 JSS pulldown study placed the four shapes on the leg-extension side: the rounded curve corresponds to a clean leg-extension that does not "blow the catch"; the sharp-peak curve corresponds to a leg-extension that catches and then runs out of leg drive before the finish ([10] Wing & Woodburn 1995, Level 2b).

The rounded curve. A curve that rises smoothly from the catch to peak force, holds peak force for the back-and-arms portion, and drops smoothly at the finish. The rounded curve is the consistent, healthy stroke; the [5] Smith & Loschner 2005 review placed the rounded curve as the operational baseline ([5] Smith & Loschner 2005, Level 5). The rounded curve is the goal, not the ideal: the goal is consistency, not an idealized textbook curve.

The sharp-peak curve. A curve that rises sharply to peak force at the catch, then drops sharply. The sharp-peak curve is the leg-strength-dominant stroke; the [9] de Brouwer et al. 2020 JSS catch-efficiency study placed the sharp peak on the catch-timing side ([9] de Brouwer et al. 2020, Level 1b/2b). The sharp peak says the legs catch hard and then run out of leg drive before the back-and-arms finish.

The dip-in-the-middle curve. A curve that rises to peak force, dips in the middle, then rises again to a secondary peak. The dip-in-the-middle curve is the sequencing fault; the [7] Soper & Hume 2004 kinematic-chain paper placed it on the missing-back-segment side ([7] Soper & Hume 2004, Level 5). The dip says the legs catch hard, the back is slow to engage, and the arms try to finish alone.

The slow-release curve. A curve that holds peak force through the drive but drops slowly at the finish. The slow-release curve is the arms-dominant stroke; the [5] Smith & Loschner 2005 review placed it on the arms-overload side ([5] Smith & Loschner 2005, Level 5). The slow release says the arms are carrying too much load through the finish and the back-and-legs are not pulling through.

What repeatability means

The shape that matters most is repeatability — every stroke should produce a curve that looks like every other stroke. The [5] Smith & Loschner 2005 biomechanical review placed the same shape on the rowing-specific side ([5] Smith & Loschner 2005, Level 5). The [18] Barrett & Manning 2004 fatigue-on-stroke-kinematics study placed repeatability on the diagnostic side: stroke-to-stroke variability rises with fatigue; the force curve is the early warning ([18] Barrett & Manning 2004, Level 2b).

The [22] Cosgrove et al. 1999 rate-vs-force-curve study in JSS placed repeatability on the rate-dependent side: the shape of the force curve changes with stroke rate, but the rower's stroke-to-stroke repeatability at a given rate is the diagnostic ([22] Cosgrove et al. 1999, Level 2b). The [16] Sanderson et al. 1997 cadence-vs-power study in JSS placed repeatability on the cadence-vs-power side: force-application shape shifts with cadence, but stroke-to-stroke consistency at a fixed cadence is the within-cadence diagnostic ([16] Sanderson et al. 1997, Level 2b).

The [17] Smith & Hopkins 2012 rowing-performance-measurement review in Sports Medicine placed repeatability on the predictive side: stroke-to-stroke consistency on the force curve is one of the variables that predicts 2K time ([17] Smith & Hopkins 2012, Level 5). The [11] Schaffert & Mattes 2010 International Journal of Sports Medicine 2000-m race-phase analysis placed repeatability on the race-fatigue side: stroke-to-stroke variability rises across the four quarters of a 2K; the force-curve consistency drops first as fatigue accumulates ([11] Schaffert & Mattes 2010, Level 2b).

The operational read: the rower who pulls at 2K pace for 30 seconds and watches the force curve become more variable is the rower whose stroke is breaking down under the prescribed load. The [27] Foster 2001 session-RPE method places the same on the load-monitoring side: load is sRPE × duration, and force-curve consistency drops first as load accumulates ([27] Foster 2001, Level 5). The [26] Borg 1982 CR-10 scale is the categorical anchor for the rower's RPE read ([26] Borg 1982, Level 5).

When the curve is consistent but wrong

The curve can be consistent stroke to stroke and still be wrong — the shape can be the sharp-peak curve, the dip-in-the-middle curve, or the slow-release curve, and the rower can produce the same wrong shape every stroke. The [4] Concept2 stroke-sequence reference placed the same shape on the manufacturer-canonical side ([4] Concept2, Level 5). The [5] Smith & Loschner 2005 biomechanical review placed the same on the rowing-specific side ([5] Smith & Loschner 2005, Level 5).

The [7] Soper & Hume 2004 kinematic-chain paper placed the consistent-but-wrong shape on the sequencing-fault side: when the legs-back-arms chain is broken, the curve has a specific wrong shape, and the rower's stroke-by-stroke consistency on the wrong shape is the diagnostic ([7] Soper & Hume 2004, Level 5). The [10] Wing & Woodburn 1995 pulldown study reached the same conclusion: the leg-extension timing determines the shape, and a consistent wrong shape is the diagnostic for a specific timing fault ([10] Wing & Woodburn 1995, Level 2b).

The honest read for the rower: a consistent sharp-peak curve is the diagnostic for a sequencing fault where the legs catch and the back fails to engage. A consistent dip-in-the-middle curve is the diagnostic for a sequencing fault where the back is slow and the arms try to finish alone. A consistent slow-release curve is the diagnostic for an arms-overload fault where the arms carry too much load through the finish.

The [9] de Brouwer et al. 2020 JSS catch-efficiency study placed the same on the catch-timing side: catch timing affects peak-force application, and a consistent sharp-peak curve is the diagnostic for catch timing that is too early or too late ([9] de Brouwer et al. 2020, Level 1b/2b). The [8] Mattes 2020 motor-learning paper placed the same on the practice-side: a consistent wrong shape is the rower's current pattern, and the goal is to make a new consistent shape, not to chase the textbook shape.

When the curve is variable

The curve can be variable stroke to stroke — the shape changes from one stroke to the next, and the [18] Barrett & Manning 2004 fatigue-on-stroke-kinematics study placed this on the fatigue side: stroke-to-stroke variability rises with fatigue ([18] Barrett & Manning 2004, Level 2b). The [22] Cosgrove et al. 1999 rate-vs-force-curve study placed it on the rate-dependent side: the curve's shape changes with rate, and a rower who is racing rate-up to chase a target split will see the curve's shape change stroke to stroke as the rate climbs ([22] Cosgrove et al. 1999, Level 2b).

The [16] Sanderson et al. 1997 cadence-vs-power study placed it on the cadence-vs-power side: a rower who is rate-limited by power output will see the curve's shape change stroke to stroke as fatigue accumulates and the leg drive weakens ([16] Sanderson et al. 1997, Level 2b). The [11] Schaffert & Mattes 2010 2000-m race-phase analysis placed it on the race-fatigue side: stroke-to-stroke variability rises across the four quarters of a 2K, and the force-curve consistency drops first as fatigue accumulates ([11] Schaffert & Mattes 2010, Level 2b).

The [25] Halson 2014 training-load monitoring review in Sports Medicine placed the same on the multi-modal signal: the constellation of HR + sRPE + force-curve consistency is the load-bearing signal ([25] Halson 2014, Level 5). The [29] Vesterinen et al. 2016 HRV-guided field trial placed the same on the adaptive-prescription side: HRV + sRPE + force-curve consistency together form the adaptive-prescription signal ([29] Vesterinen et al. 2016, Level 1b/2b).

The honest read for the rower: a variable curve is the diagnostic for a stroke that is breaking down under the prescribed load. The [27] Foster 2001 session-RPE method is the operational load metric; the [26] Borg 1982 CR-10 scale is the categorical anchor for the rower's RPE read. The rower who pulls at a hard pace and watches the curve become variable is the rower whose stroke is at the limit of the prescribed load.

Beyond repeatability: rate and force curve together

The [22] Cosgrove et al. 1999 rate-vs-force-curve study placed rate and force curve on the same axis: the curve's shape changes with rate, and the rower's diagnostic value of the curve is rate-dependent ([22] Cosgrove et al. 1999, Level 2b). The [16] Sanderson et al. 1997 cadence-vs-power study reached the same conclusion from the cadence side ([16] Sanderson et al. 1997, Level 2b). The [19] Wilson et al. 2010 rate-vs-performance study in IJSPP placed rate and force curve on the diagnostic side: the optimal rate sits where drive-length holds while peak-force drops ([19] Wilson et al. 2010, Level 1b/2b).

The [21] Baudouin & Hawkins 2004 JSS stroke-rate-vs-length study placed the same on the rowing-specific side: rate and length interact, and the force curve's shape changes across the rate-band ([21] Baudouin & Hawkins 2004, Level 1b/2b). The [6] Hofmijster et al. 2021 rate-band field study placed the same on the rate-dependent diagnostic side: stroke rate interacts with drive length and peak force ([6] Hofmijster et al. 2021, Level 1b/2b).

The [23] Koppo et al. 2000 European Journal of Applied Physiology rate-dependent metabolic study placed rate and force curve on the physiological side: stroke rate interacts with the force-curve signature and the physiological cost ([23] Koppo et al. 2000, Level 2b). The [24] ACSM 2009 progression-models position stand placed the same on the resistance-training side: rate and force-curve consistency improve across a resistance-training cycle ([24] ACSM 2009, Level 5).

The operational read: the rower who reads the force curve at a fixed rate and watches the curve change as rate climbs is the rower who has the rate-band diagnostic. The [22] Cosgrove 1999 study placed this on the empirical side; the [19] Wilson 2010 study placed this on the diagnostic side. The rower who reads both the rate and the curve at the same time has the strongest diagnostic.

Practical reading: how to look at the PM5 in real time

The [3] Concept2 PM5 documentation is the operational read: the monitor reports per-stroke drive-time, recovery-time, drive-length, peak-force, average-force, stroke rate, and 500m split. The [4] Concept2 stroke-sequence reference places the same events on the manufacturer-canonical side: the four phases of the stroke are catch, drive, finish, recovery. The rower's read of those events is the diagnostic.

The [5] Smith & Loschner 2005 biomechanical review placed the same readouts on the rowing-specific side ([5] Smith & Loschner 2005, Level 5). The [17] Smith & Hopkins 2012 performance-measurement review placed the predictive value of each variable on the 2K-time side ([17] Smith & Hopkins 2012, Level 5).

The practical read:

1. Pull a 30-second steady piece at a fixed rate and a fixed split. Watch the force curve on the PM5. The shape should be consistent stroke to stroke. The [18] Barrett & Manning 2004 fatigue study placed the same on the diagnostic side: stroke-to-stroke variability rises with fatigue; a 30-second piece is enough to see the early-warning signal.
2. Look at the four shapes. Rounded = clean stroke; sharp-peak = leg-dominant; dip-in-the-middle = sequencing fault; slow-release = arms-dominant. The [7] Soper & Hume 2004 kinematic-chain paper placed the four shapes on the sequencing-fault side.
3. Look at the consistency. Stroke-to-stroke consistency is the diagnostic for fatigue, rate-up, and rate-limited power output. The [11] Schaffert & Mattes 2010 2000-m race-phase analysis placed the same on the race-fatigue side.
4. Look at the rate-band. The [22] Cosgrove 1999 study placed the force-curve shape on the rate-dependent side. The [19] Wilson 2010 study placed the optimal rate on the drive-length-holds-while-peak-force-drops side.

The [32] Magill 2011 motor-learning textbook placed the policy on the feedback side: feedback frequency, precision, and timing shape retention ([32] Magill 2011, Level 5). The [33] Schmidt & Lee 2011 textbook reached the same conclusion from the rower's perspective. The rower who reads the force curve rower-by-rower is the rower whose stroke improves.

Limitations

The force-curve read is one input to the rower's diagnosis of stroke quality, not the diagnosis itself. The [17] Smith & Hopkins 2012 review placed the predictive value of each per-stroke variable on the 2K-time side, and not all variables predict equally ([17] Smith & Hopkins 2012, Level 5). The [22] Cosgrove 1999 study placed the diagnostic value of the curve on the rate-dependent side: the curve's shape changes with rate, and the diagnostic value of the curve is rate-dependent ([22] Cosgrove 1999, Level 2b).

The [18] Barrett & Manning 2004 fatigue-on-stroke-kinematics study placed the same on the fatigue side: stroke-to-stroke variability rises with fatigue; the force curve is the early warning, not the full diagnosis ([18] Barrett & Manning 2004, Level 2b). The [11] Schaffert & Mattes 2010 2000-m race-phase analysis placed the same on the race-fatigue side: stroke-to-stroke variability rises across the four quarters of a 2K; the force-curve consistency drops first as fatigue accumulates ([11] Schaffert & Mattes 2010, Level 2b).

The [25] Halson 2014 training-load monitoring review placed the same on the multi-modal signal: single markers misfire, the constellation of HR + sRPE + force-curve consistency is the load-bearing signal ([25] Halson 2014, Level 5). The [31] Manresa-Rocamora et al. 2021 HRV-guided-training meta-analysis in IJSPP placed the same on the adaptive-prescription side: HRV + sRPE + force-curve consistency together form the adaptive-prescription signal ([31] Manresa-Rocamora et al. 2021, Level 1a).

The honest read for the rower: the force curve is a feedback channel, and feedback channels have failure modes. The [33] Schmidt & Lee 2011 textbook placed feedback on the policy side: degraded feedback degrades retention ([33] Schmidt & Lee 2011, Level 5). The [32] Magill 2011 textbook reached the same conclusion from the motor-learning side. The rower who reads the curve rower-by-rower is the rower whose stroke improves; the rower who chases the textbook curve is the rower whose learning plateaus.

The summary in one paragraph

The PM5 force curve is one input to the rower's read of stroke quality, not the read itself. The [1] Kleshnev 2020 rowing-kinetics handbook chapter placed the operational read on the kinetic-event side: drive-time, recovery-time, peak-force, average-force, drive-length, and handle-speed-force curves are the events the PM5 reports ([1] Kleshnev 2020, Level 5). The [5] Smith & Loschner 2005 biomechanical review placed the same on the rowing-specific side ([5] Smith & Loschner 2005, Level 5). The [7] Soper & Hume 2004 kinematic-chain paper placed the same on the sequencing side ([7] Soper & Hume 2004, Level 5). The [10] Wing & Woodburn 1995 pulldown study placed the same on the leg-extension side ([10] Wing & Woodburn 1995, Level 2b). The [22] Cosgrove et al. 1999 rate-vs-force-curve study in JSS placed the same on the rate-dependent side: the curve's shape changes with rate, and the diagnostic value of the curve is rate-dependent ([22] Cosgrove et al. 1999, Level 2b). The [18] Barrett & Manning 2004 fatigue-on-stroke-kinematics study placed stroke-to-stroke variability on the diagnostic side: variability rises with fatigue; the force curve is the early warning ([18] Barrett & Manning 2004, Level 2b). The [17] Smith & Hopkins 2012 rowing-performance-measurement review in Sports Medicine placed the predictive value of each per-stroke variable on the 2K-time side ([17] Smith & Hopkins 2012, Level 5). The [11] Schaffert & Mattes 2010 2000-m race-phase analysis placed the same on the race-fatigue side ([11] Schaffert & Mattes 2010, Level 2b). The [9] de Brouwer et al. 2020 JSS catch-efficiency study placed the same on the catch-timing side ([9] de Brouwer et al. 2020, Level 1b/2b). The [16] Sanderson et al. 1997 cadence-vs-power study in JSS placed the same on the cadence-vs-power side ([16] Sanderson et al. 1997, Level 2b). The [19] Wilson et al. 2010 rate-vs-performance study in IJSPP placed the same on the optimal-rate side ([19] Wilson et al. 2010, Level 1b/2b). The [21] Baudouin & Hawkins 2004 rate-vs-length study in JSS placed the same on the rate-length-interaction side ([21] Baudouin & Hawkins 2004, Level 1b/2b). The [6] Hofmijster et al. 2021 rate-band field study in Int J Sports Med placed the same on the rate-dependent-diagnostic side ([6] Hofmijster et al. 2021, Level 1b/2b). The [23] Koppo et al. 2000 rate-dependent metabolic study in Eur J Appl Physiol placed the same on the physiological-cost side ([23] Koppo et al. 2000, Level 2b). The [24] ACSM 2009 progression-models position stand placed the same on the resistance-training-cycle side ([24] ACSM 2009, Level 5). The [25] Halson 2014 training-load monitoring review in Sports Medicine placed the constellation of HR + sRPE + force-curve consistency on the load-bearing signal side ([25] Halson 2014, Level 5). The [29] Vesterinen et al. 2016 HRV-guided field trial in MSSE placed the same on the adaptive-prescription side ([29] Vesterinen et al. 2016, Level 1b/2b). The [31] Manresa-Rocamora et al. 2021 HRV-guided-training meta-analysis in IJSPP placed the same on the systematic-review side ([31] Manresa-Rocamora et al. 2021, Level 1a). The [32] Magill 2011 motor-learning textbook placed feedback on the policy side ([32] Magill 2011, Level 5). The [33] Schmidt & Lee 2011 textbook reached the same conclusion from the rower's perspective ([33] Schmidt & Lee 2011, Level 5). The [3] Concept2 PM5 documentation and the [4] Concept2 stroke-sequence reference are the operational anchors; the [12] Ingham et al. 2008 indoor-rower training study and the [13] Hagerman 1984 physiology review are the rowing-specific anchors; the [27] Foster 2001 session-RPE method and the [26] Borg 1982 CR-10 scale are the load-monitoring anchors; the [28] Scherr et al. 2013 RPE–lactate correlation is the RPE-validity anchor ([27] Foster 2001, Level 5; [26] Borg 1982, Level 5; [28] Scherr et al. 2013, Level 2b).

The right posture is to pull a 30-second steady piece at a fixed rate and a fixed split, watch the four shapes, watch the consistency, watch the rate-band, and let the body decide. The shape that matters is repeatability — every stroke should look like every other stroke. The shape that is consistent stroke to stroke is the current pattern; the goal is to make it consistent and clean, not to chase an ideal curve from a textbook. The AI coach uses the force curve as one input when it diagnoses a fault; the rower uses it as the same input.

For a deeper exploration of how the rate-band interacts with the force-curve shape, see our rate-caps guide and our rate-ladders guide.

What to do with this article

Read the principle: the PM5 force curve is one input to the rower's read of stroke quality, not the read itself. The shape that matters most is repeatability — every stroke should look like every other stroke.

Read the recognition protocol: pull a 30-second steady piece at a fixed rate and a fixed split. Look at the four shapes (rounded = clean; sharp-peak = leg-dominant; dip-in-the-middle = sequencing fault; slow-release = arms-dominant). Look at the consistency (stroke-to-stroke variability rises with fatigue). Look at the rate-band (the curve's shape changes with rate).

Read the diagnostic protocol: when the curve is consistent but wrong, the fault is specific and addressable (the [7] Soper & Hume 2004 kinematic-chain paper places this on the sequencing side). When the curve is variable, the diagnostic is fatigue, rate-up, or rate-limited power output (the [18] Barrett & Manning 2004 fatigue study places this on the fatigue side).

Read the practical read: the [3] Concept2 PM5 documentation and the [4] Concept2 stroke-sequence reference are the operational anchors; the [5] Smith & Loschner 2005 biomechanical review and the [17] Smith & Hopkins 2012 performance-measurement review are the academic anchors; the [22] Cosgrove 1999 study and the [19] Wilson 2010 study place the rate-band diagnostic on the empirical side. The [32] Magill 2011 motor-learning textbook places feedback on the policy side: frequency, precision, and timing shape retention.

When the curve is consistent stroke to stroke and the shape is the rounded curve, the stroke is clean. When the curve is consistent but wrong, the fault is specific and addressable. When the curve is variable, the diagnostic is fatigue, rate-up, or rate-limited power output. The force curve is one input to the diagnosis; the rower's read of the curve is the diagnostic.