KOM Informatics Login Signup Documentation Blog

Signup For Free!

The KOM Informatics service is free of charge during the beta period which is expected to run for a few years. Please click the checkbox on the CAPTCHA, and click [Submit]. We'll send you to the registration form for KOM Informatics.


KOM Informatics: Overview Of The Statistically Significant Correlations Between Workload And Performance 2022-2023 By Rider Phenotype

By David Brown

Interval Advisor AI The KOMInformatics.com SysAdmin has the ability to run a Training Effectiveness Report for a given time frame, workload window, and offset for each eligible user. The report extracts statistically significant relationships between performance and workload using PR watts on the performance side against kilojoules of work on the workload side. For background info on the report and these concepts see the first article in this series KOM Informatics: Raising 20 Minute Power Over 4 Weeks: What Actually Worked.

This article looks at Training Effectiveness data compiled from all of the PR time boxes, through the filter of rider phenotype.

Training Effectiveness Report Results (2022-2023)

As of 8/7/2024, The admin controlled Training Effectiveness report has run 9 times reporting on 6 month time frames (Jan-June, Jul-Dec) for 2022 and 2023. Five of those times the report used 28 days for the workload window, 0 days for the offset and 5 for the number of PRs. 28 days was thought to be a reasonable amount of time for most interventions to start showing some results. The data from the last time the 28 day window was run (Jan-June 2024) is not included in this analysis which is limited to 2022-2023.

Sort Order

You can change the sort order of the detail tables below by clicking on either the User Percentage or Watts Gain column headers. Clicking on the same column header twice changes the direction of the sort.

Column Explanations

KOM Informatics Rider Phenotypes

KOM Informatics Rider Phenotypes are derived from percentile ranks from all users from 12 PR timeboxes divided into 4 categories. Each user's data, is available to the user from the Rider Assessment report. My data is shown below. My phenotype based on averages from each of the categories is Long TTs-Distance with an average of 77.33. I'm just a little less strong in Short Climbs-Short TTs (76.33).



Phenotype Relative Performance - Which Group Is Best At Improvements?
Phenotype User Percentage Average Best PR Cat Intervention % (14 Day Interventions) Watts Gain (14 Day Interventions Intervention % (28 Day Interventions) Watts Gain (28 Day Interventions
Short Sprints 0.41 52.99 0.37 52 0.42 45
Long TTs - Distance 0.2 57.91 0.11 63 0.05 77
Short Climbs - Short TTs 0.2 70.54 0.3 51 0.32 65
Long Sprints - Attacks 0.19 65.29 0.22 55 0.21 42
Phenotype Relative Performance - Discussion

User Percentage in the Phenotype Relative Performance table looks at the prevalence of the phenotype in the database as a whole. 41% of users were classified as Short Sprint (SS), Short Climbs-Short TTs (SC-STT) and Long TTs-Distance (LTT-D) both came in at 20% and Long Sprints-Attacks (LS-A) came in at 19%.

So how well did the groups perform in achieving numbers of successful interventions? Averaging the 2 Intervention % columns, Short Climbs-Short TTs was the clear winner with 20% of the users but 31% of the interventions. Long Sprints-Attacks came in second with 19% of the users and 21.5% of the interventions Shorts Sprints was in third with 41% of the users and 39.5% of the interventions. Long TTs-Distance was last, with 20% of the users but only 8% of the interventions. The middle groups (SC-STT, LS-A) outperformed the groups at either end (SS, LTT-D).

How well did the groups perform in terms of watts gain? LTT-D was the clear winner averaging 70 watts. SC-STT was second with 58, SS and LS-A tied for third with 48.5.

14 Day Workload Window
  Workload Phenotype User Percentage Watts Diff Average R
NI Short Sprints 0.44 67 0.882911428571429
Z7 Short Sprints 0.56 54 0.882879545454545
Z4 Short Sprints 0.53 52 0.881828571428571
Z5 Short Sprints 0.43 52 0.889352941176471
Z6 Short Sprints 0.57 52 0.89514
Z3 Short Sprints 0.05 34 0.90975
NI Short Climbs - Short TTs 0.48 64 0.89231052631579
Z7 Short Climbs - Short TTs 0.46 55 0.902025
Z5 Short Climbs - Short TTs 0.35 54 0.877064285714286
Z6 Short Climbs - Short TTs 0.47 43 0.912848648648648
Z4 Short Climbs - Short TTs 0.42 36 0.888921212121212
Z3 Short Climbs - Short TTs 0.06 25 0.9099
Z5 Long TTs - Distance 0.18 87 0.881035714285714
NI Long TTs - Distance 0.18 83 0.881457142857143
Z7 Long TTs - Distance 0.11 48 0.906611111111111
Z4 Long TTs - Distance 0.14 44 0.885918181818182
Z6 Long TTs - Distance 0.16 42 0.900676923076923
Z3 Long TTs - Distance 0.04 17 0.887533333333333
Z4 Long Sprints - Attacks 0.32 91 0.909052
Z6 Long Sprints - Attacks 0.32 49 0.89164
Z5 Long Sprints - Attacks 0.32 47 0.89956
NI Long Sprints - Attacks 0.29 40 0.902678260869565
Z7 Long Sprints - Attacks 0.23 34 0.904338888888889
Z3 Long Sprints - Attacks 0.06 27 0.8983

14 Day Workload Window Specificity Analysis

Phenotype Workload Reliance % Average Watts Gain Discussion
Long Sprints - Attacks Z4 0.21 91
Long Sprints - Attacks Z6 0.21 49 This phenotype should be best at this workload. But these athletes were largely indifferent to the choice of workload. The watts gain for this workload (49) was only a little above the workload average (48). Specificity is not evident here either in workload choice or outcomes. Long Sprinters achieved their best watts gains through Z4 intervals, an area of relative weakness.
Long Sprints - Attacks Z5 0.21 47
Long Sprints - Attacks NI 0.19 40
Long Sprints - Attacks Z7 0.15 34
Long Sprints - Attacks Z3 0.04 27
Long TTs - Distance Z5 0.22 87
Long TTs - Distance NI 0.22 83 One might expect athletes of this phenotype to rely on large volumes of non-interval training and in fact the data shows the highest reliance percentage for this intervention at 0.23, along with the second highest watts gain (83). The highest watts gain for the group (87) came with Z5 intervals, an area of relative weakness. Z4 intervals should be another area of strength for these athletes, and they did show a modest reliance (.17) on this interval. Specificity is somewhat evident in these results.
Long TTs - Distance Z7 0.14 48
Long TTs - Distance Z4 0.17 44
Long TTs - Distance Z6 0.2 42
Long TTs - Distance Z3 0.05 17
Short Climbs - Short TTs NI 0.21 64
Short Climbs - Short TTs Z7 0.2 55
Short Climbs - Short TTs Z5 0.16 54 This phenotype may be the best at these workloads but they don't rely on the corresponding Z5 intervals all that much. The 16% reliance percentage was the second lowest for the group, with only Z3 (.03) being lower. When they did use Z5 intervals the watts gain (54) was higher than the average for the group(46.167). Specificity is not evident here. This group achieved their best watts gains working on their areas of relative weakness, NI (64) and Z7 (55).
Short Climbs - Short TTs Z6 0.21 43
Short Climbs - Short TTs Z4 0.19 36
Short Climbs - Short TTs Z3 0.03 25
Short Sprints NI 0.17 67
Short Sprints Z7 0.22 54 Between Z6 and Z7 intervals, Short Sprinters showed a heavy reliance (.44) on their area of strength.The NI workload was the best for them (67) in terms of watts gain. Specificity is evident here in the reliance percentage. Short Sprinters achieved their best watts gains (67) utilizing the NI workload, an area of relative weakness.
Short Sprints Z6 0.22 52
Short Sprints Z4 0.21 52
Short Sprints Z5 0.17 52
Short Sprints Z3 0.02 34
28 Day Workload Window
  Workload Phenotype User Percentage Watts Diff Average R
Z4 Short Sprints 0.33 64 0.897419230769231
NI Short Sprints 0.3 45 0.870591666666667
Z5 Short Sprints 0.39 42 0.894483870967742
Z6 Short Sprints 0.58 41 0.88894347826087
Z7 Short Sprints 0.44 41 0.896131428571428
Z3 Short Sprints 0.06 34 0.88496
Z3 Short Climbs - Short TTs 0.05 82 0.872075
NI Short Climbs - Short TTs 0.41 78 0.893090625
Z7 Short Climbs - Short TTs 0.43 71 0.89835
Z6 Short Climbs - Short TTs 0.33 58 0.911696153846154
Z5 Short Climbs - Short TTs 0.14 51 0.867445454545455
Z4 Short Climbs - Short TTs 0.33 46 0.907896153846154
Z7 Long TTs - Distance 0.06 152 0.8784
Z4 Long TTs - Distance 0.05 68 0.87585
Z3 Long TTs - Distance 0.03 55 0.8692
Z6 Long TTs - Distance 0.04 50 0.930933333333333
Z5 Long TTs - Distance 0.05 37 0.92475
NI Long TTs - Distance 0.03 26 0.8705
Z6 Long Sprints - Attacks 0.18 78 0.868435714285715
NI Long Sprints - Attacks 0.14 43 0.899045454545455
Z5 Long Sprints - Attacks 0.34 38 0.901696296296296
Z7 Long Sprints - Attacks 0.16 37 0.877615384615385
Z4 Long Sprints - Attacks 0.23 27 0.890555555555555
Z3 Long Sprints - Attacks 0.01 14 0.8785

28 Day Workload Window Specificity Analysis

Phenotype Workload Reliance % Average Watts Gain Discussion
Long Sprints - Attacks Z6 0.17 78 This group showed only a moderate reliance on Z6 intervals, their area of strength. However these intervals led to by far the greatest watts gain (78) over second place NI (43). The watts gain here is markedly higher than those revealed by the corresponding 14 Day Workload Window section above. This suggests that Z6 based interventions for this group may take a while longer to show complete results.
Long Sprints - Attacks NI 0.13 43
Long Sprints - Attacks Z5 0.32 38
Long Sprints - Attacks Z7 0.15 37
Long Sprints - Attacks Z4 0.21 27
Long Sprints - Attacks Z3 0.01 14
Long TTs - Distance Z7 0.25 152
Long TTs - Distance Z4 0.2 68
Long TTs - Distance Z3 0.1 55
Long TTs - Distance Z6 0.15 50
Long TTs - Distance Z5 0.2 37
Long TTs - Distance NI 0.1 26 The low reliance on NI (0.1) and low watts gain (26) is in marked contrast to what is found at the 14 Day Workload Window above (0.22, 83). Instead, these athletes trained their weakness (Z7) relying on these intervals at a category high .25 while showing an impressive 152 average watts gain.
Short Climbs - Short TTs Z3 0.03 82
Short Climbs - Short TTs NI 0.24 78
Short Climbs - Short TTs Z7 0.26 71
Short Climbs - Short TTs Z6 0.2 58
Short Climbs - Short TTs Z5 0.08 51 This phenotype may be the best at these workloads but they don't rely on the corresponding Z5 intervals all that much. The .08 reliance percentage was the second lowest for the group, with only Z3 (.03) being lower. When they did use Z5 intervals the watts gain (51) was lower than the average for the group(64.33). Specificity is not evident here. This group achieved their best watts gains working on their areas of relative weakness, Z3 (82) and NI (78).
Short Climbs - Short TTs Z4 0.2 46
Short Sprints Z4 0.16 64
Short Sprints NI 0.14 45
Short Sprints Z5 0.19 42
Short Sprints Z6 0.28 41 At a 28 day Workload Window, Short Sprinters continued their reliance on areas of strength (Z6 .28), (Z7, .21). This approach yielded an average watts gain (41) lower than the average (44.5). When they relied on relative areas of weakness (Z4 64), (NI 45) they had greater average gains (54.5 vs 44.5). Specificity is evident here in terms of reliance percentage.
Short Sprints Z7 0.21 41
Short Sprints Z3 0.03 34

Discussion

The first part of the adage "train your weakness, race your strength" is largely corroborated by this data. Groups relying on an area of relative weakness to produce watts gain found success 6 of 8 times. (LS-A (14)(Z4), LTT-D (28)(Z7), SC-STT (14)(NI), SC-STT (28)(Z3), SS (14) (NI), SS (28) (Z4)). Reliance specificity - relying on an area of strength to produce optimal watts gains - only worked twice (2 of 8), for the LTT-D group at a 14 day workload window, and LS-A at the 28 day workload window.

These outcomes suggest a direction for athletes to modify their training when they reach a plateau.

+
Contact Us

Address

Hillsborough, NJ USA

Email Us