Video summary

What is probability | Expected Values, Frequency Distribution, Complement

Main summary

Key takeaways

Educational

Main Ideas and Concepts Covered

Uncertainty & Decision-Making

Many real-world decisions (e.g., CEOs investing in R&D, mergers/buyouts) require estimating which outcomes are more likely. Probability and statistics help predict the likelihood of each outcome to support better choices.

What Probability Means

  • Probability = the chance that something happens.
  • Academic framing: likelihood of an event occurring.
  • An event can be:
    • a specific outcome, or
    • a combination of outcomes (e.g., Heads in a coin flip; divisible-by-3 on a die).

Probabilities are often written as real numbers between 0 and 1:

  • Example: 20% = 0.2

Interpretation:

  • P = 1 → event is absolutely certain
  • P = 0 → event is absolutely certain to NOT happen
  • Typical probabilities fall between 0 and 1 (e.g., 0.2, 0.5, 0.66)

Computing Theoretical Probability (Counting Favorable Outcomes)

General rule: [ P(A) = \frac{\text{number of favorable outcomes}}{\text{total number of possible outcomes}} ]

Terms:

  • Sample space = all possible outcomes (also referred to as the set of outcomes)

Examples of Theoretical Probabilities

  • Coin flip

    • Event: Heads
    • Favorable = 1, total = 2 [ P(\text{Heads}) = \frac{1}{2} = 0.5 ]

  • Die roll

    • Event: rolling a 4
    • Favorable = 1, total = 6 [ P(4) = \frac{1}{6} \approx 0.167 ]

  • Die divisibility condition

    • Event: number divisible by 3 → outcomes are {3, 6}
    • Favorable = 2, total = 6 [ P(\text{divisible by 3}) = \frac{2}{6} \approx 0.33 ]

  • Two independent events

    • Probability both occur = product of individual probabilities
    • Example mentioned: getting an Ace and a spade [ P(\text{Ace}) \times P(\text{spade}) ]

Expected Values (What You “Expect” Over Many Trials)

What an Experiment Is

If you repeat trials to estimate/observe outcomes, that whole process is an experiment. Example: tossing a coin multiple times and recording outcomes across flips is one experiment with many trials.

Theoretical vs. Experimental Probability

  • Theoretical (true) probability: derived from reasoning/counting.
  • Experimental probability: estimated from repeated trials.
  • Experimental probabilities are often close approximations, but not guaranteed to match exactly.

Formula for Experimental Probability

[ \text{Experimental probability}=\frac{\text{number of successful trials}}{\text{total number of trials}} ]

Expected Value for Categorical Outcomes

Expected value of event A (E(A)) is the number of times you expect a categorical event to occur when repeating the experiment.

Formula: [ E(A) = P(A) \times n ] where n = number of trials.

Example: drawing a spade

  • (P(\text{spade}) = \frac{1}{4} = 0.25)
  • Repeat n = 20 draws [ E = 0.25 \times 20 = 5 ]

Important clarification: you won’t always get exactly 5; outcomes could be 4, 6, or even 20. Expected value is an average/long-run expectation.

Expected Value for Numerical Outcomes

When outcomes have numeric values: [ \text{Expected value} = \sum (\text{outcome value} \times \text{probability of that outcome}) ]

Example: bow and arrow scoring

  • Outer layer = 10 (prob. 0.5)
  • Second layer = 20 (prob. 0.4)
  • Bullseye = 100 (prob. 0.1)

Expected value:

  • (0.5 \times 10 + 0.4 \times 20 + 0.1 \times 100)
  • (= 5 + 8 + 10 = \mathbf{23})

Key interpretation: 23 is not necessarily a possible single-shot score, but it’s useful for long-run prediction/comparison.

How Expected Value Is Used

Expected value helps make predictions about the future based on past data. Often predictions are given as intervals due to uncertainty (example used: meteorology forecasting likely ranges).

Dice Example → Frequency Distribution and Probability Frequency Distribution

Setup: Sum of Two Dice

Rolling two standard 6-sided dice:

  • Total possible outcomes = (6 \times 6 = 36)
  • A conceptual 6×6 table is used to count how often each sum occurs.
  • The sum 7 occurs 6 times.

Therefore: [ P(\text{sum} = 7) = \frac{6}{36} = \frac{1}{6} ]

Expected value of the sum is computed similarly, and the explanation states it ends up being 7.

Why “Most Probable” vs. “Expected” Can Differ

Even if 7 is the expected value (and likely), the probability of getting exactly 7 is only 1/6. So you can’t reasonably “bet” on one exact outcome.

Probability Frequency Distribution (Method)

Probability frequency distribution = the collection of probabilities for each possible outcome, represented via a table or graph.

Construction steps:

  1. Build a frequency table
    • For each unique outcome (e.g., each possible sum), count how many times it appears.
    • That count is the frequency.

  2. Convert frequency to probability

    • For each outcome: [ \text{probability}=\frac{\text{frequency of the outcome}}{\text{size of the sample space}} ]

    • The resulting set of probabilities forms the probability frequency distribution.

Using the Distribution for Predictions/Intervals

On the graph:

  • X-axis = possible outcome values
  • Y-axis = probability of each outcome

To form an interval with the highest probability, choose values around the expected value, where the probability bars are highest.

Complement of an Event (Opposite Event)

Definition

The complement of an event is everything the event is not.

Notation:

  • Complement of A is written as A’ (apostrophe)

Complement Properties

  • Complement of the complement returns the original event: [ (A’)’ = A ]

Key Probability Rule

Probabilities across all possible outcomes sum to 1, so: [ P(\text{complement of } A) = 1 - P(A) ]

Why Sums of Probabilities Might Be > 1 or < 1

  • Sum > 1 can happen if outcomes are counted simultaneously and effectively double-counted.
  • Sum < 1 means not all possible outcomes were accounted for.

Examples of Complement Reasoning

  • Die example

    • Event: “roll an even number”
    • Complement: “NOT rolling even” = rolling an odd number

  • Multi-outcome example

    • Event: rolling {1, 2, 4, 5, 6}
    • Complement: “rolling NOT in that set” = rolling 3
    • Calculation:
      • (P(\text{not getting 3}) = 1 - P(\text{getting 3}))
      • (P(\text{getting 3}) = \frac{1}{6})
      • So: [ P(\text{not getting 3}) = 1 - \frac{1}{6} = \frac{5}{6} ]

Speakers / Sources Featured

  • No specific named speakers are mentioned (only generic narration and “Hey, everyone!”).

Original video