How Many Fish Are in the Lake?
You can't count every fish in a lake — so ecologists tag some, let them mix, and catch again. Run the whole method with a bag of sweets, then find out how wrong it can be.
Where this idea comes from
Start here — these are the sources that inspired this exploration.
Introduction
Ecologists estimating a fish population catch a netful, tag them, release them, and later catch again: the fraction of tagged fish in the second catch tells you how big the whole population is. The same trick counts homeless populations, festival crowds and census undercounts. You can run it honestly on your desk: a big bag of pasta or sweets is the lake, a marker pen does the tagging — and because YOU can count the true total afterwards, you can do what ecologists never can: measure exactly how wrong the method was, and then investigate what makes it more or less reliable.
Guiding Questions
- Run the experiment: 'catch' a handful from the bag, mark them, mix them back, catch again, and count the marked ones. What's your reasoning to turn those counts into an estimate of the total?
- Now count the bag for real. How far off were you? Repeat the whole process several times — how spread out are your estimates?
- Something about your experiment controls how trustworthy the estimate is. Form a hypothesis about what it is, vary it, and measure whether the spread of estimates behaves as you predicted.
- Real fish aren't pasta: tagged animals might die, hide, or get caught more easily. Choose one such failure, build it into your experiment deliberately, and see which way it biases the answer.
- Find a real use of this method — fish stocks, census checking, festival attendance — and judge: knowing what you now know, how much would you trust the published number?