Relative frequency as an estimate of probability
Relative frequency as an estimate of probability
When the outcomes are not equally likely, a probability is estimated by the relative frequency: in how many of many times the event happened.
– the number of times the event happened, – the number of times it was checked.
In one country, 180,000 babies were born in one year, and 92,340 of them were boys:
The probability that a newborn baby is a boy is about 0.513 – a little more than a half. A birth is not the toss of a coin: symmetry alone would get it wrong.
The more data there are, the closer the relative frequency comes to the probability.
A drawing pin thrown on a table lands on its back (point up) or on its side.
A computer simulated 1,000 throws of a pin whose probability of landing on its back is 0.6 – the dashed line:
- After a few throws, the relative frequency jumps up and down.
- After hundreds of throws, it settles near 0.6.
- A real pin does not come with its probability: it is thrown many times, and the probability is estimated by the relative frequency.
An insurance company estimates a probability from past data, and works out from it how many times the event is expected to happen:
Last year the company insured 40,000 young drivers, and 2,800 of them had an accident and made a claim.
Next year it will insure 50,000 young drivers.
A claim costs the company 12,000 ₪ on average.
- The estimated probability of an accident in a year: .
- The expected number of claims next year: .
- The expected cost: ₪.
- For each driver: ₪. That is the lowest price the company can charge without losing money.
“A 70% chance of rain tomorrow” means: of the days when the forecast said 70%, it rained on about 70%.
- It does not mean that rain will fall on 70% of the area, or for 70% of the hours.
- A single forecast cannot be checked: tomorrow it rains or it does not.
- Many forecasts are checked together: if it rained on 69 of 100 days with a 70% forecast, the forecast is good.
An estimate from relative frequency is good only when the data fit the question.
- Enough data: a player who scored 3 of 3 shots does not score with probability 1.
- The same population: data on young drivers do not fit an estimate of the accidents of 50-year-old drivers.
- The same conditions: the rain data of an especially wet year do not describe an ordinary year.
A computer simulates a basketball player taking free throws.
His probability of scoring is not written here: run the experiment to estimate it.
- Run 10 throws, and write the relative frequency of the baskets.
- Add 100 and then 1,000 throws, and write the relative frequency again.
- What is your estimate of the probability of scoring? Which of the estimates is better, and why?
3,600 babies were born in a hospital in one year, and 1,836 of them were boys.
Estimate the probability that a baby born in the hospital is a boy.
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – check and a new line · * – times · ^ – power, ↓ – end of the power
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – a new line · * – times · ^ – power, ↓ – end of the power
Next year the hospital expects 5,000 births. How many boys are expected to be born?
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – check and a new line · * – times · ^ – power, ↓ – end of the power
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – a new line · * – times · ^ – power, ↓ – end of the power
A company insures phones against a broken screen.
Last year it insured 25,000 phones, and 1,500 of them broke.
Next year it will insure 30,000 phones.
A screen repair costs the company 600 ₪.
Estimate the probability that an insured phone breaks in a year.
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – check and a new line · * – times · ^ – power, ↓ – end of the power
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – a new line · * – times · ^ – power, ↓ – end of the power
How many phones are expected to break next year?
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – check and a new line · * – times · ^ – power, ↓ – end of the power
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – a new line · * – times · ^ – power, ↓ – end of the power
What is the least the company must charge for each phone to cover the expected repairs? (in shekels)
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – check and a new line · * – times · ^ – power, ↓ – end of the power
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – a new line · * – times · ^ – power, ↓ – end of the power
A weather service forecast “30% rain” on 200 days.
It rained on 58 of those days.
What is the relative frequency of rainy days among the 200 days?
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – check and a new line · * – times · ^ – power, ↓ – end of the power
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – a new line · * – times · ^ – power, ↓ – end of the power
Are the “30% rain” forecasts good? Explain.
Which estimate is better? Explain each answer.
- Noa scored 7 of 10 shots, and Maya 140 of 200. Which of them can estimate her probability of scoring better?
- An insurance company wants to estimate the probability of an accident for 60-year-old drivers. Should it use the data of drivers aged 18–24?
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – check and a new line · * – times · ^ – power, ↓ – end of the power
/ – fraction bar · Tab or the arrows – between numerator and denominator, then a note above or below · Enter – a new line · * – times · ^ – power, ↓ – end of the power