The Performance Manager TM by vko15646

VIEWS: 0 PAGES: 25

									The Performance ManagerTM

        Andrew R. Coggan, Ph.D.
    Cardiovascular Imaging Laboratory
  Washington University School of Medicine
           St. Louis, MO 63021
Scientific studies using mathematical modeling to
quantitatively relate training load to performance

    • Approximately 30 English language papers

    • Many different sports studied (i.e., weightlifting,
      hammer throwing, running, swimming, cycling,
      triathlon)

    • Variety of mathematical approaches used (e.g., from
      simple regression to neural networking)

    • Vast majority have relied upon Eric Banister’s
      impulse-response model or some variation thereof.
Impulse-response model of training adaptation




                        Banister et al., Aust J Sports Med 7:57, 1975
Impulse-response model: effect of “square wave”
     increase in training load to 100 units/d
 Prediction of training-induced changes in
performance using impulse-response model




                         Busso et al., J Appl Physiol 92:572, 2002
Limitations to the impulse-response model

• Mathematically complex, yet overly simplified

• Requires frequent, quantitative measurement of
  performance (i.e., 20-200 times every 60-90 d)

• Parameter estimates still may be insufficiently stable
  (precise) to permit highly accurate prediction of future
  performance

• Inter-study and inter-subject variability in parameter
  estimates (esp. ka (k1) and kf (k2)) limits ability to apply
  “generic” version of model
           Representative studies from the literature
                                 Training
   Study          Subjects                     τa (τ1)   τf (τ2)     ka (k1)       kf (k2)
                                 program

                                 Constant-
                                    load
                  Initially        cycling
Busso et al.,
                 untrained                      38±9      2±2      0.048±0.019   0.117±0.114
   1991                          60 min/d,
                 men (n=8)
                                 4 d/wk, for
                                   14 wk

                                   Interval
                                 cycling 40-
Busso et al.,   Recreational                                         0.0021,       0.0078,
                                  60 min/d,    60, 60     4, 6
   1997         cyclists (n=2)                                       0.0019        0.0073
                                  3-5 d/wk
                                  for 14 wk

                                   Interval    3 d/wk:   3 d/wk:      3 d/wk:       3 d/wk:
                  Initially                    41±15       9±6     0.019±0.006
                                 cycling 40-                                     0.015±0.008
Busso, 1993      untrained
                                  60 min/d     5 d/wk:   5 d/wk:      5 d/wk:       5 d/wk:
                 men (n=6)
                                  for 15 wk    35±12      13±3     0.021±0.006   0.021±0.006
  Representative studies from the literature (con’t)
                                 Training
   Study          Subjects                     τa (τ1)   τf (τ2)     ka (k1)      kf (k2)
                                 program

                                  Running
                   Initially       40-100
Morton et al.,
                  untrained       min/d, 7     40, 50    11, 11       1, 1         1.8, 2
   1990
                  men (n=2)      d/wk, for 4
                                     wk

                 National and
                                 Swimming
                 international
 Iñigo et al.,                    35-40
                     level                     41±4      12±6      0.062±0.04   0.128±0.055
    1996                         km/wk for
                  swimmers
                                  44 wk
                    (n=18)



                                 Swimming
                 Olympic level
 Hellard et                       45-50
                  swimmers                     50±8      19±8      0.01±0.01     0.05±0.03
 al., 2005                       km/wk for
                    (n=7)
                                    4y
           Performance Manager
Impulse-response model of training adaptation



             




                                                        Coggan, 1975
                        Banister et al., Aust J Sports Med 7:57, 2004
Performance Manager: result of “square wave”
    increase in training load to 100 TSS/d
   Uses for the Performance Manager

• Determining optimal long-term training load

• Identifying periods of severe overreaching that may
  lead to illness or overtraining

• Identifying periods of “training stagnation”

• Assuring the progressive overload principle is applied in
  a rational manner

• Planning a taper in an attempt to peak for a particular
  event
     Performance Manager chart
for an elite track cyclist (2002 season)
         Performance Manager chart
for an elite track cyclist (2002 season) (con’t)
Performance versus TSB: effect of duration
          Duration of effort       TSB at time of PB
                  5s                      35
                 10 s                     35
                 20 s                     35
                 30 s                     35
                1 min                     34
                2 min                     34
                5 min                     33
       5 min (normalized power)           34
                10 min                    6
       10 min (normalized power)          5
                20 min                   (-19)
       20 min (normalized power)          34
                30 min                    6
       30 min (normalized power)          34
                60 min                   (-10)
       60 min (normalized power)          34
Performance Manager chart for a masters cyclist
              (2004 season)
Performance Manager chart for a masters cyclist
            (2004 season) (con’t)
Performance Manager chart for a masters cyclist
              (2005 season)
Performance Manager chart for a masters cyclist
            (2005 season) (con’t)
Performance versus TSB: effect of duration
    Duration of effort       TSB for 2004 PB   TSB for 2005 PB
            5s                     11                13
           10 s                    11                 6
           20 s                    11                 3
           30 s                    11                13
          1 min                    19                13
          2 min                    8                 14
          5 min                    4                  9
 5 min (normalized power)          6                 10
          10 min                   0                 (-7)
 10 min (normalized power)         19                10
          20 min                   0                 (-4)
 20 min (normalized power)         19                 6
          30 min                   10                (-2)
 30 min (normalized power)         1                  6
          60 min                   3                 (-2)
 60 min (normalized power)         3                 (-5)
TSB at time of personal best for power
            (all durations)
TSB at time of personal best for power
               (<5 min)
TSB at time of personal best for power
              (>10 min)
        Caveats and limitations

• Accuracy of predictions depends upon:
   – accuracy/completeness of underlying data
   – use of appropriate time constants (esp. for ATL)

• “Composition” of the training load still matters

• Training Manager helps you view the “forest”, but
  you should never lose sight of the “trees”
         Additional resources


• www.cyclingpeakssoftware.com/power411/
  performancemanagerscience.asp
• www.cyclingpeakssoftware.com/power411/
  performancemanager.asp
• www.cyclingpeakssoftware.com/power411/
  howtoperformancemanager.asp
• www.cyclingpeakssoftware.com/support/
  WKO+2_1_user_guide.pdf
Special thanks to the “beta testers” of the
          Performance Manager
   Hunter Allen       Dave Martin
   Tom Anhalt         Scott Martin
   Gavin Atkins       Phil McKnight
   Andy Birko         Rick Murphy
   Lindsay Edwards    Terry Ritter
   Mark Ewers         Ben Sharp
   Sam Callan         Alex Simmons
   Chris Cleeland     Phil Skiba
   Tony Geller        Ric Stern
   Dave Harris        Bob Tobin
   Dave Jordaan       John Verheul
   Kirby Krieger      Frank Overton
   Chris Merriam      Lynda Wallenfells
   Jim Miller         Mike Zagorski
   Chris Mayhew

								
To top