There are 5M confirmed cases of SARS-CoV2 In the US. Let’s say the true # is 10X higher @ 50M out of a total US population of 328M which = 15% positive. There are a total of 12,659 CFB players and let’s assume they also get infected at the same rate despite some of the mitigation efforts mentioned (15% = 1928). The fatality rate/infection for their age group is 0.001% which means 0.02 players are likely at risk at most. The fatality rate is likely much lower for elite athletes vs. the general pop with whatever percentage of co-morbidities (obesity, cancer, diabetes etc).
I’m not saying the 2020 CFB season is necessarily worth risking a ~0.0001-0.0002% chance of death or any of the unknown long-term consequences, but I guarantee that >1 player will get CTE and/or long-term disabilities associated with traumatic brain injury. What is the difference between CTE and Covid liability for universities? If you can accept the risk associated with CTE you shouldn’t be worried about Covid. If you are super conservative and say 10% of CFB players will suffer significant CTE then Covid is somewhere around 60,000-fold less dangerous.
Stadiums filled with fans are not a good idea, but the kids should be able to play if they want to. Champions league is going fine.
In my opinion, it is very important to transition to thinking about acceptable risk explicitly in regards to COVID-19; and the comparison to CTE is a reasonable and potentially useful comparative tool. I do take issue with your method, namely no citation of your numbers and no apparent consideration of the uncertainty of those estimates.
You've done a back of the envelope calculation with numerous assumptions and estimated parameters that provide zero context about any uncertainty about those assumptions and estimates.
Even if the inputs (lacking citations) are the appropriate estimates based on current available evidence, your conclusion ignores the underlying uncertainty. You could be talking about a most likely 1/50 chance that one player dies, while ignoring a very appreciable chance that tens or hundreds of players could die.
If I'm able to find time, I may work up a more data driven counter argument, for now I can provide one anecdote that should illustrate a glaring limitation of you estimates.
You work from a 15% prevalence rate, in the MLB we have 2 teams that have already experienced outbreaks among their 30 player rosters just two weeks into the season. The Cardinals, so far, have
9 players who tested positive (30% of the roster), and the
Marlins had 18 players test positive (60%). There is clear evidence that at least in those cases, a group of young athletes sharing locker rooms, practicing and playing together, can experience an outbreak that is far more widespread than your stated estimate. This alone would multiply your estimate by at least a factor of 2 or 4; now imagine if we bump the extremely low stated fatality rate up to as still extremely low rate of 0.001%, in combination we're now multiplying your estimate by a factor of 20 or 40.
These are small adjustments to your assumptions and drastically change the outcome; any good model would include sensitivity analysis and a discussion of uncertainty of parameter estimates.