I asked ChatGPT to arbitrate. It said I was right but being a troll
JonE has the stronger objection to the evidence and the safety comparison, but overstates what the original poster actually said. Kankama is right about exercise, but doesn’t answer the methodological criticism.
There are three separate questions here.
1. Does the yoga anecdote establish what is claimed?
It supports the modest observation that some older people have excellent balance. It does not, by itself, establish that yoga caused their balance, quantify improvement, or overturn the general relationship between age and balance.
Recruiting an older group from a yoga class potentially mixes age with training, underlying health and selection: people with good balance may be more likely to attend and continue yoga. JonE’s analogy identifies that problem reasonably well.
However, we haven’t seen the study. That recruitment could be perfectly appropriate for some research questions. The questionable part is presenting the anecdote as having “blown up” the age–balance correlation.
2. Can older people improve balance through training?
Yes. The underlying proposition is supported by much better evidence than this anecdote. Randomised trials of yoga show balance benefits compared with inactive controls; a large Cochrane review found exercise programmes reduced the rate of falls among older people by about 23%. Neither finding establishes a comparable reduction in falling overboard or drowning.
So Kankama is right about the general proposition, but a true conclusion does not make every argument offered for it sound.
JonE’s observation that the good participants have practised for years doesn’t refute trainability. It is compatible with training helping, although it still doesn’t prove causation.
The masters cycling examples have the same limitation: even accepting the figures, exceptional athletes demonstrate what is possible, rather than what is typical or achievable for everyone.
3. Does that support the argument about lifejackets?
No comparative evidence has been supplied. Better balance could plausibly reduce some ways of entering the water; a lifejacket addresses the consequences once someone is in it. These are complementary precautions.
There is direct evidence relevant to lifejackets: one matched observational study found wearing a PFD was associated with approximately half the risk of drowning death, although the authors acknowledged possible bias from excluded records.
The original closing sentence compares the benefits of working on fitness with “worrying about” safety gear. That wording sidesteps the useful comparison: taking fitness measures versus actually using appropriate equipment.
But JonE’s claim that the poster said yoga was more likely to prevent drowning than reasonable precautions goes beyond the supplied wording. The insinuation is understandable; the explicit comparative claim isn’t there. “Triggered” and “trolling” add no evidence.
Finally, Kankama’s definition of assertion is wrong: an assertion is a statement put forward as true, whether supported or unsupported. And an unspecified second-hand account of a study is not the same as supplying its data.
My ruling: JonE’s substantive criticism largely stands. It would be stronger with less personal language and a more precise distinction between “older people can retain excellent balance” and “this anecdote proves training causes it or justifies dispensing with safety equipment.”
