The Cauchy distribution is a continuous distribution in Intro to Probability with a peak at its center and very heavy tails. Its mean and variance do not exist, so it behaves very differently from the normal distribution.
The Cauchy distribution is a continuous probability distribution in Intro to Probability that has a strong peak in the middle and very long tails on both sides. It is usually written with a location parameter, which sets the center, and a scale parameter, which controls how spread out the curve is.
Its probability density function is where is the location and is the scale. If and , you get the standard Cauchy distribution. The curve is symmetric, so it looks centered, but that symmetry can be misleading.
The big surprise is that the Cauchy distribution has no finite mean and no finite variance. That means the usual summaries you use for many distributions, like expected value and spread, are undefined. In other words, you cannot say, “the average outcome is here” in the same way you would for a normal or uniform distribution.
This is also why the moment-generating function does not exist for the Cauchy distribution. MGF methods rely on moments like the mean and variance, and the Cauchy does not have those moments in a usable form. You may still see its characteristic function in more advanced work, but for a basic probability course, the main idea is that many standard shortcuts break down here.
A common misconception is that a graph with a clear center must have a usable mean. The Cauchy distribution shows that a distribution can look nicely balanced and still refuse to settle down numerically. If you draw repeated samples, the sample mean can jump around wildly instead of stabilizing the way it does for many other distributions.
The Cauchy distribution shows you where standard probability tools stop working cleanly. In Intro to Probability, a lot of distributions are introduced through expected value, variance, and moment-generating functions, so Cauchy is a useful counterexample that makes those ideas feel less automatic.
It matters most in the chapter on moment-generating functions and moments because it is one of the clearest examples of a distribution with no MGF. That helps you see that MGFs are powerful, but not universal. If a problem asks for the mean or variance of a Cauchy random variable, the correct move is not to force a formula, but to recognize that those quantities are undefined.
It also builds intuition for heavy tails. Compared with a normal distribution, a Cauchy random variable has a much higher chance of producing extreme values. That makes it a useful model in some physics and engineering settings, and it also explains why averages of Cauchy samples are unreliable.
In a probability class, Cauchy is often the example that tests whether you are using a formula mechanically or thinking about what the distribution actually allows. If you can spot why the mean, variance, and MGF fail here, you are thinking like a probability student instead of just plugging into formulas.
Keep studying Intro to Probability Unit 13
Visual cheatsheet
view galleryHeavy-Tailed Distribution
The Cauchy distribution is a classic heavy-tailed distribution, which means extreme values happen more often than they do in thin-tailed distributions like the normal. That tail behavior is exactly why sample averages can behave badly. When you see a probability model with frequent outliers or rare but huge jumps, heavy tails are the first feature to check.
Moment-Generating Function (MGF)
MGFs are often used to find moments and study sums of independent random variables, but the Cauchy distribution is a counterexample because its MGF does not exist. That makes it a good reminder that MGF techniques have limits. If a problem asks for an MGF, you should first ask whether the integral actually converges.
Characteristic Function
The characteristic function exists for the Cauchy distribution even though the MGF does not. In more advanced probability, characteristic functions can sometimes do work that MGFs cannot. For Intro to Probability, the main takeaway is that existence of one transform does not automatically mean the distribution has ordinary moments.
Convergence
Cauchy is a famous example when talking about convergence of sample means. For many distributions, the sample mean settles toward the expected value as the sample size grows, but that idea fails here because there is no finite mean to converge to. It is a strong reminder that convergence results depend on the distribution you start with.
A quiz or problem set might ask you to identify whether a Cauchy random variable has a mean, variance, or MGF, and the right answer is that it does not. You may also be asked to compare it with a distribution like the normal or uniform and explain why averages from a Cauchy sample are unstable. If the question gives the density, you should read the location parameter as the center and the scale parameter as the width, then recognize the heavy tails. A short response that says “undefined moments, heavy tails, no MGF” usually earns the main credit when the prompt is about behavior rather than calculation.
The Cauchy distribution is often confused with the normal distribution because both are symmetric and centered, but they behave very differently in the tails. The normal distribution has a finite mean and variance, and its sample means stabilize. The Cauchy distribution has much heavier tails, so outliers matter more and the usual summary statistics do not exist.
The Cauchy distribution is a continuous distribution with a clear center, but its mean and variance are undefined.
Its heavy tails make extreme values much more likely than in a normal distribution.
The moment-generating function does not exist for the Cauchy distribution, so MGF-based methods do not work here.
Sample means from a Cauchy distribution do not settle down the way they do for many other distributions.
In Intro to Probability, Cauchy is a useful example of why you cannot assume every distribution has usable moments.
The Cauchy distribution is a continuous probability distribution with a central peak and very heavy tails. In Intro to Probability, it is mainly used as an example of a distribution whose mean, variance, and MGF do not exist. That makes it stand out from the better-behaved distributions you usually compute with.
Its tails are too heavy for the integrals that define the mean and variance to converge. Even though the graph looks symmetric, the far-out values contribute enough mass that the usual averages do not settle to finite numbers. That is why the expected value and variance are undefined, not just hard to compute.
No. Both are symmetric, but the normal distribution has finite moments and thinner tails, while the Cauchy distribution has heavy tails and undefined mean and variance. If a problem involves stable averages or standard deviation, you are probably not dealing with Cauchy behavior.
Most of the time, you use it to recognize what does not work. If a question asks for an MGF, mean, variance, or convergence of the sample mean, Cauchy is the distribution that signals those tools fail. You may also be asked to interpret its density parameters as the center and scale of the curve.