Theoretical probability and observed frequency
A model gives a theoretical probability, while an experiment gives a proportion based on observed outcomes. The two need not agree exactly in a short run.
Probability & statistics / English
The Law of Large Numbers | Animation
EarthPen · YouTube · 3:09
YouTube · EarthPenOriginal video
READ & KEEP
The explanation, unpacked.
Reviewed learning material · Video analysis · English01 02 03
Read the full overview
The law of large numbers explains why repeated trials can make an observed proportion approach its theoretical probability. Simple probability examples distinguish a long-run tendency from the noisy outcomes of a small experiment.
Use the learning inspector for key ideas and moments, or open the reading tabs for the complete notes.
Chapters
0:28Theoretical probability and observed frequency1:38State the long-run principle carefully2:18More trials do not compensate for past outcomesLearning script
Generated from the video's visuals and explanation; not verbatim speech.
A model gives a theoretical probability, while an experiment gives a proportion based on observed outcomes. The two need not agree exactly in a short run.
For independent repeated trials with a common success probability , the observed success proportion approaches as the number of trials grows, in the probabilistic sense specified by the law. It need not get closer at every single step.
The comparison of 10, 100 and 1,000 trials illustrates why larger samples often stabilize proportions. A fair independent coin still has heads probability on the next toss, regardless of an earlier streak.
Knowledge cards
State the long-run principle carefully
For independent repeated trials with a common success probability , the observed success proportion approaches as the number of trials grows, in the probabilistic sense specified by the law. It need not get closer at every single step.
More trials do not compensate for past outcomes
The comparison of 10, 100 and 1,000 trials illustrates why larger samples often stabilize proportions. A fair independent coin still has heads probability on the next toss, regardless of an earlier streak.
Explore the knowledge in this video
- Law of large numbers ExplanationAt 0:28
Why this connection?
A model gives a theoretical probability, while an experiment gives a proportion based on observed outcomes. The two need not agree exactly in a short run. For independent repeated trials with a common success probability , the observed success proportion approaches as the number of trials grows, in the probabilistic sense specified by the law. It need not get closer at every single step. The comparison of 10, 100 and 1,000 trials illustrates why larger samples often stabilize proportions. A fair independent coin still has heads probability on the next toss, regardless of an earlier streak.
- Law of large numbers ExplanationAt 0:28
Why this connection?
A model gives a theoretical probability, while an experiment gives a proportion based on observed outcomes. The two need not agree exactly in a short run. For independent repeated trials with a common success probability , the observed success proportion approaches as the number of trials grows, in the probabilistic sense specified by the law. It need not get closer at every single step. The comparison of 10, 100 and 1,000 trials illustrates why larger samples often stabilize proportions. A fair independent coin still has heads probability on the next toss, regardless of an earlier streak.
Questions this video answers
Meet the concept↗
No, the observed success proportion does not necessarily get closer to the theoretical probability at every single step. The law of large numbers describes a long-run tendency where the proportion approaches as the number of trials grows, but it explicitly states that it need not get closer at every single step.
Conditions: The trials are independent.; Each trial has a common success probability .; Analyzing the behavior of the proportion step-by-step.
1:38 – 2:18Watch this moment ↗
Find a method↗
Comparing these different sample sizes illustrates that larger samples tend to produce observed proportions that are more stable and closer to the theoretical probability. With fewer trials (like 10), random noise can cause significant deviations from the expected proportion.
Conditions: The comparison involves sample sizes of 10, 100, and 1,000 trials.; The goal is to understand the stabilization of observed proportions.
2:18 – 3:09Watch this moment ↗
Meet the concept↗
No, a fair independent coin still has a heads probability of on the next toss, regardless of an earlier streak. The trials are independent, meaning past outcomes do not influence future probabilities.
Conditions: The coin is fair.; The tosses are independent.; There has been an earlier streak of outcomes.
2:18 – 3:09Watch this moment ↗
Understand why↗
The law of large numbers states that for independent repeated trials with a common success probability , the observed success proportion approaches as the number of trials grows. This is a long-run tendency specified by the law, meaning that while individual short-term outcomes may vary, the cumulative average stabilizes around the theoretical value over many repetitions.
Conditions: The trials are independent.; Each trial shares a common success probability .; The number of trials grows large.
1:38 – 2:18Watch this moment ↗
Meet the concept↗
A model provides a theoretical probability, which is the expected likelihood of an outcome based on the underlying rules. An experiment provides an observed proportion, which is calculated from the actual outcomes recorded.
Conditions: The context is a short run of experiments.; Comparing theoretical probability from a model with observed proportion from data.
0:28 – 1:38Watch this moment ↗
Similar explanations
YouTube4:01Another explanation of Law of large numbers.
Law of large numbers
YouTube8:59Probability & statistics·English
Law of large numbers | Probability and Statistics | Khan AcademyFoundationalReviewed notes
Another explanation of Law of large numbers.
Law of large numbers
Connect the next step
YouTube4:22Prerequisite in Probability and statistics: Independence → Law of large numbers.
Independence · View learning sequence
YouTube2:27Probability & statistics·English
Die rolling probability with independent events | Precalculus | Khan AcademyFoundationalReviewed notes
Prerequisite in Probability and statistics: Independence → Law of large numbers.
Independence · View learning sequence
YouTube6:59Prerequisite in Probability and statistics: Expectation → Law of large numbers.
Expectation · View learning sequence
A different approach
YouTube8:59Probability & statistics·English
Law of large numbers | Probability and Statistics | Khan AcademyFoundationalReviewed notes
Whiteboard for Law of large numbers, compared with the current visual explanation.
Law of large numbers