Friday, September 21, 2012

How to make a readiness monitoring using a simple wellness questionnaire. Part 2



How to make a readiness monitoring using a simple wellness questionnaire.
Part 2

In this part I am going to cover how to collect the data in the Excel and how to numerically analyze it.

This involves:
  1. Color-coding individual categories (fatigue, sleep...)
  2. Color-coding total score (sum of categories)
  3. A way for calculating base-line and trends using rolling average of the last six measurements
  4. How to calculate the difference of the current score to the individual tendency (which is calculated by rolling average) 
Since the wellness questionnaire is basically a nominal scale, the method of calculating difference might be simple subtraction (Difference = Current Score – Rolling Average). Other calculations that are based on ratio scales might involve percent change (Difference[%]  = (Current Score – Rolling Average)/ Rolling Average x 100) or Z-score (Z-Score = (Current Score – Rolling Average)/Rolling Standard Deviation) and all of these might vary on how you calculate baseline, whether it is a rolling average (what amount of measurements should be taken into account?), or just plain average of all measurements, or even an average of measurements in some period (for example pre-season).

Interesting point is that calculating the baseline allows us to deal with players who score higher-or-lower than normal, while calculating Z-scores allows us to deal with players who show higher or lower variability in their scoring.

To be clearer, we have three players who score:
 








Average
SD
CV
Athlete A
15
14
16
15
14
15
16

15,0
0,82
5%
Athlete B
10
9
11
10
9
10
11

10,0
0,82
8%
Athlete C
15
13
17
15
13
15
17

15,0
1,63
11%


 Athlete A tends to always report lower scores, while Athlete B tends to always report higher scores. Calculating simple difference score (in this case score minus average) provides a method to deal with this scenario instead of relying on absolute numbers as a sign of reduced readiness.
 


Difference







Athlete A
0
-1
1
0
-1
0
1
Athlete B
0
-1
1
0
-1
0
1
Athlete C
0
-2
2
0
-2
0
2



As you can see both Athlete A and Athlete B have same difference scores compared to their average (baseline). But what about Athlete C? He is always double.

Both Athlete C and Athlete B have same average score (15), but way different difference scores. What you can judge from the example is that for each change in Athlete B’s score Athlete C scored double. Thus, Athlete C has higher variability in his scoring (see his SD and CV).

As some athletes tend to report normally higher-or-lower scores, some athletes tend to have higher-or-lower variability as well. Does this means they are automatically more or less ready to train, more or less tired? Not necessary so. I guess we need to take into account their natural variability and find a way to calculate it and take into account.

One solution would be to use Z-score:

Z-score







Athlete A
0,00
-1,22
1,22
0,00
-1,22
0,00
1,22
Athlete B
0,00
-1,22
1,22
0,00
-1,22
0,00
1,22
Athlete C
0,00
-1,22
1,22
0,00
-1,22
0,00
1,22



As you can see, all three athletes deviate the same from the average when we use the Z-score. Thus you see that none of them is more tired or more ready compared to whole group.

In this installment I used simple difference score (since the wellness is based on nominal scale and not ratio scale), but you can play around and implement Z-scores. In that case you would need additional tab to calculate rolling SD (standard deviation) the same way we calculated rolling averages. In that case GREEN zone might be from 0 to – 1SD, ORANGE zone from -1SD to -2SD and RED everything below -2SD.

For more info I suggest checking the following papers from JASC (thanks to Dan Baker for sending them)

Fatigue monitoring in high performance sport: A survey of current trends.
J. Aust. Strength Cond. 20(1)12-23. 201

Monitoring overtraining in athletes: A brief review and practical applications for strength and conditioning coaches. J. Aust. Strength Cond. 20(2)39-51. 2012




In the next installment I will cover one simple way to visualize this for better decision making, along with designing a very simple dashboard.
 

No comments:

Post a Comment