Correlation, Causation, and You

Written by: Dr. Kris Camelio PT, DPT, MS, CSCS, CEAS-I

On this International Workers' Day, we celebrate, recognize, and remember the workers, the members and organizers of labor unions, and all the selfless people who have held-- and continue to hold humanity dearer than business. People who cherish the lives and well-beings of their families, their coworkers, their coworkers' families, and everyone else in the community. People who help achieve critical victories that move public health policy towards an ideal that protects fundamental human rights, protects' peoples' health, and protects peoples' dignity. We salute the people working and organizing to prevent work-related tragedies from happening to the next generations of people just trying to live. We honor everyone operating in good faith to help make work more safe, more enjoyable, and more enriching for many generations ahead.

If I can offer one lesson I've learned throughout the years, it’s this: 'listen to your body'. If you’re not sure whether you should do an exercise, add weight on the bar, play a recreational sport, go for a jog-- listen to your body. Does your body, your joint, your muscle/tendon/ligament, the biological structure in question- does it seem to want to move more, or does it seem to want to move less? Will you feel better or worse after you do that particular activity, exercise, movement, or recreational sport that you're considering doing? This can help inform the answers to some of the questions you may have.

If I can share two lessons I've learned throughout the years, it's 1) 'listen to your body', and 2) correlation does not necessarily imply causation. Correlation generally means that, as one variable increases, the other variable increases. Correlations usually aren't perfect; they don't explain why something happens. Correlations simply imply that two things [aka independent variables, or disparate values] change at a similar rate. Consider these truths: all cats can be biscuit-makers [you know, cats making biscuits? The way cats seem to knead dough as they make a comfy surface upon which to nap? That's the phenomenon to which I'm referring, it's adorable], but, not all biscuit-makers can be cats! Furthermore, if one person in a kitchen can produce one biscuit, it would, then, make sense, that more people in a kitchen could probably produce more biscuits in that kitchen. However, just because multiple biscuits exist in a kitchen, it does not necessarily mean, or imply (nor can a resonable person conclude, based on this evidence alone), that those biscuits were produced by multiple people. One person can still produce multiple biscuits. Many kitchens have multiple people with no biscuits! We need to ask more questions, we need more context, to further determine the relationship between people in kitchens, and biscuits in kitchens (if that was, of course, something a person hungry for further information--and/or biscuits-- would want to know).

Here's a sports example of why correlation does not necessarily imply causation: as one team’s score increases, we'd expect that team's likelihood of winning increases, right? Therefore, there's probably a correlation between how much a team scores, and how likely they are to win the game (in fact, there is-- but it may be weaker than you think; more on that shortly).

However, just because a team scores a certain number of points, it doesn’t necessarily mean those points will, ultimately, cause them to win more games. While I'm assuming that points scored has a correlation to games won, it's certainly not a perfect correlation to games won. Why not? The Denver Nuggets scored the most points this 2025-2026 season of any NBA team-- over 10,000. However, this season, The Nuggets' record was 54-28, whereas, the Detroit Pistons this season had a record of 60-22. Thus, the Pistons won more games than the Nuggets throughout the course of the season, while scoring fewer points than the Nuggets did. The team that scored the most points this season did not, in fact, win the most games. Thus, the correlation between overall points scored and overall games won is, at least, not perfect.

Just because a team scores a large number of points doesn't even mean they will win *a* game! On one winter night in Denver Colorado in 1983, the Denver Nuggets scored 184 points in a single game. I'm pretty sure that's more than any of the winning postseason teams have scored on a single night this year. Of course, if one team doesn't score anything, all the other team needs, is to make one shot in order to win. It makes sense that, since the Denver Nuggets scored 184 that game, one could expect the Nuggets won that game. However, by the end of this (yes, triple-overtime) game, the Detroit Pistons had scored 186 points. Yup, 186 is more than 184. Thus, even though scoring 184 points in any game may, conceivably, be a strong indicator that team will have won the game-- clearly there were other factors-- other forces at play-- including the legendary Isaiah Thomas, who scored 52 points that night, and the DETROIT BASKETBALL that contributed to the victory for the Pistons over the Nuggets that night.

The cause of the victory is ultimately scoring more points than one's opponent, which also involves ensuring that opponent *doesn’t score* sometimes-- y'know, playing defense-- and a summative Katamari of other key factors in each game (including rebounding percentage) [nods in the general direction of Ausar Thompson's 15-rebound performance in game 5 of this playoff's Pistons-Magic series] that ultimately add up to a win. In fact, one study estimated that the number of points scored in an NBA game only predicts the winner about 13% of the time!

This relates to health. Just because a person 'works out' doesn't necessarily imply that person is doing so in a way that is conducive to their physical health (regardless of whatever operational definition of 'health' you'd like to employ here). Just because a person can deadlift x amount of weight, or never deadlifts ever, doesn’t necessarily mean that person will or won’t experience back pain at some point this year (there are many, many more factors to consider). Just because a person's issue feels better, or worse, with, yoga, for example, doesn’t mean that yoga, in itself, is necessarily the thing making the issue feel better or worse. Just because a person does a particular stretch, or exercise, or mobility drill, or eats a particular thing, or takes a particular dietary supplement-- each of those behaviors don't necessarily imply that it's the behavior, itself that is causing a particular health outcome.

Furthermore, just because there is correlation between x injury and y sport, it doesn’t even mean a person will inevitably get injured by participating in, or rehearsing-- if not playing-- some version of the sport. Just because a person's body was moving at the time it experienced an injury, or an unpleasant sensation, that doesn’t necessarily mean that moving at all, or moving in other ways is going to cause a person to experience that injury or unpleasant sensation again. It’s generally the type of movement, the magnitude of the forces associated that movement, and a combination of many other factors, that differentiates movement that is within a physiological bandwidth of health that allows one's body to maintain its homeostasis, from movement that causes one's body to have to repair itself to the extent the person might need to change their behaviors, and/or enlist the help of a professional, to allow one's body to properly heal, and enable one's body to do what one intends it to do. In conclusion: 1) listen to your body, and 2) remember that correlation doesn’t always imply causation.

And, if I can offer you 3 things, it’s the first two, and that, 3) by the way, I offer virtual services to help you: feel and move better, prevent injuries before they happen, and keep injuries from happening again.

See Also:

Cabarkapa D, Deane MA, Fry AC, Jones GT, Cabarkapa DV, Philipp NM, Yu D. Game statistics that discriminate winning and losing at the NBA level of basketball competition. PLoS One. 2022 Aug 19;17(8):e0273427. doi: 10.1371/journal.pone.0273427. PMID: 35984813; PMCID: PMC9390892.

Brown, D. G., Lamb, C. E., & Byl, L. R. (2025). Don’t train your model on my novel: AI refusal statements.

Copyright 2026 Valid Physical Therapy Services, LLC, All Rights Reserved.

No AI has been used in the creation of this article, nor may it be used to train AI. This article is the original work and Intellectual Property of Dr. Kris Camelio, and may not be modified from its original version, nor entered into Artificial Intelligence Archives and/or Language Learning Models without the express written consent of Dr. Camelio and/or representatives operating on his behalf.