Name the test and the unit
“Fly 10 m, seconds, 20 m run-in” is better than writing “speed: 1.03.” The setup belongs with the result.
Data helps me see what memory misses.
I do not record training to turn every workout into an experiment. I record it because weeks blur together, personal bests are easy to remember, and ordinary days often explain more than the exciting ones.
PERSONAL DATA STUDY · FALL 2025
Getting stronger and jumping higher.
Did my back squat 1RM estimates and approach jump height rise during the same training block?
See my records and analysis ↗During this block, my squat and approach jump both improved. I wanted to see how they changed together, without assuming that one caused the other.
RECORDING WELL
Useful data starts before the graph.
A clean chart cannot repair inconsistent testing. These habits make one month comparable with the next.
Repeat the conditions
Use the same timing system, distance, surface, warm-up, and start rules when possible. Otherwise, label the difference.
Record the day around the number
Sleep, RPE, soreness, pain, weather, and training phase can explain why two identical workouts did not feel identical.
Do not save only highlights
Missing slow reps creates a flattering dataset and a weak conclusion. Ordinary and difficult sessions belong in the record too.
GRAPHS I ACTUALLY USE
Each picture answers a different question.
The sample charts below are illustrations. Their value is in choosing the graph that matches the question, not making the page look technical.
TIME SERIES · EXAMPLE
What changed, and when?
Plot results in date order and mark training blocks, injuries, and meets. This is the first graph I would make because sequence matters.
ROLLING AVERAGE · EXAMPLE
Is there a direction under the noise?
A short moving average smooths one-off highs and lows. It can reveal a trend, though it also reacts slowly when something changes.
SCATTERPLOT · EXAMPLE
Do two measures move together?
Compare strength with sprint time or sleep with RPE. A pattern can suggest a useful question; it cannot prove that one variable caused the other.
DISTRIBUTION · EXAMPLE
What is normal for me?
A histogram or box plot shows the usual range and unusual results. This helps decide whether a personal best is a new level or one exceptional attempt.
QUESTIONS I WANT TO INVESTIGATE
The point is a better next decision.
These are the kinds of problems that connect my interest in statistics with the choices I make in training.
01
Did getting stronger help me become faster?
- Record
- Estimated squat 1RM alongside flying 10m, 40-yard, and 100m time
- Look with
- Start with timelines and scatterplots, then examine correlation
- Be careful
- Strength, technique, maturation, and recovery changed together.
02
How long does a training block take to show up in performance?
- Record
- Weekly lifting and sprint volume compared with results 7, 14, and 28 days later
- Look with
- Rolling averages and lagged comparisons
- Be careful
- A small personal dataset can suggest timing without proving a mechanism.
03
Was a personal best a new level or one exceptional day?
- Record
- Repeated attempts collected under similar conditions
- Look with
- Mean, standard deviation, and intervals around the change
- Be careful
- A result becomes more convincing when it can be repeated.
04
When did progress slow down?
- Record
- Strength, sprint time, and jump height across training blocks
- Look with
- Moving averages and comparisons before and after a change in emphasis
- Be careful
- A flat trend can reflect fatigue, noisy measurement, or a real plateau.
05
What changed around injury and return?
- Record
- Training load, RPE, sleep, pain notes, and performance across the timeline
- Look with
- Segmented timelines and careful before–after summaries
- Be careful
- The record can describe my experience; it cannot provide medical or causal proof.
06
Does more work always produce more improvement?
- Record
- Weekly volume and intensity alongside later performance
- Look with
- Compare straight-line and curved relationships
- Be careful
- More complicated models can easily overfit a small dataset.
Another experience: returning to squats after track season
This reflection is separate from the fall 2025 study above. Its exact dates still need checking; it is not another measured dataset.
A SEPARATE TRAINING REFLECTION
When my squat returned, my vertical jump returned too.
After track season in June, I started squatting consistently again. Over the same period, my vertical jump moved back toward its earlier level. I did not record enough standardized tests to claim that one caused the other, but the overlap gave me a useful question to investigate.
THE QUESTIONDid rebuilding lower-body strength happen alongside the recovery of my vertical jump?
WHAT MY RECORD SHOWS01END OF TRACK SEASON 02JUNE · RETURN TO LIFTING 03LATER CHECKPOINT
A lower starting point
- Strength
- Squatting had become less consistent during the season.
- Jump
- My vertical did not feel or test at its previous level.
- Context
- Race fatigue, shifting priorities, and the end of a long season all mattered.
Squatting became regular again
- Strength
- I rebuilt the movement and training rhythm instead of chasing a maximum immediately.
- Jump
- Jump work stayed in the plan while I watched for signs of returning explosiveness.
- Context
- Training structure, recovery, and lower-body strength changed together.
Both measures moved back
- Strength
- My squat performance improved from the post-season starting point.
- Jump
- My vertical jump also returned toward its earlier level.
- Context
- This was an observed pattern, not a controlled experiment.
Squat performance and vertical jump recovery moved in the same direction during this period. That makes strength a plausible part of the story and gives me a pattern worth tracking more carefully.
The two changes happened at roughly the same time. A timeline and, once I have enough repeated tests, a scatterplot can show whether that relationship continues.
I cannot say that squatting alone restored my jump. Plyometrics, reduced competition fatigue, sleep, bodyweight, motivation, and test conditions also changed, and my older jump results were not all measured with the same setup.
For 8–12 weeks, record one weekly squat measure (top set or estimated 1RM), a standardized vertical-jump test, bodyweight, sleep, RPE, pain, and plyometric volume. Plot both measures by date before calculating correlation.
A SIMPLE CHECK
Is the recent window actually different?
Use 6–20 results collected with the same equipment, surface, warm-up, and timing method. Consistency makes the comparison more meaningful.
EXAMPLES
For a vertical jump, higher is better. For sprint time, lower is better.
Better performance is
Enter at least six results, choose a direction, and compare.
Why compare two windows?
Early and recent averages reduce the influence of one unusually strong or weak attempt.
What does the spread show?
The sample SD describes the spread of the whole series. A training trend can increase it too, so it is not a separate estimate of testing error.
Why keep this descriptive?
These averages show what changed in the results you entered. They do not test significance, explain the cause, or account for time dependence.
READING BEHIND THE PAGES
Sources I used, with limits kept in view.
Research helps me test my assumptions. It does not turn this personal website into a controlled study or replace my coaches and medical professionals.
01
Hopkins · 2000
↗02Measures of reliability in sports medicine and science
Why repeated tests and normal variation matter before calling a change real.
Saw, Main & Gastin · 2016
↗03Monitoring the athlete training response
Why simple self-reports can add useful context to objective training data.
Dudley et al. · 2023
↗Training load monitoring in adolescent athletes
What the evidence can and cannot currently say about training load, performance, and injury in teenagers.
WHY THE MATH MATTERS TO ME
The most useful calculation is not the one that makes the story look perfect. It is the one that helps me ask a better next question.