6.3 Normal Distribution—Lap Times

Normal Distribution of Lap Times

more resources to help you study

practice questions

Probability calculations with normal distributions

A normal distribution is a continuous probability distribution that's symmetric and bell-shaped. Two parameters define it completely: the mean ($\mu$), which represents the average lap time, and the standard deviation ($\sigma$), which measures how much lap times vary from that average.

To find the probability of a specific lap time occurring, you first convert the raw lap time to a z-score using this formula:

$z = \frac{x - \mu}{\sigma}$

Here, $x$ is the lap time you're interested in. The z-score tells you how many standard deviations that time sits above or below the mean. A z-score of $-1.5$, for instance, means the lap was 1.5 standard deviations faster than average.

Once you have the z-score, use a standard normal table or calculator to find the corresponding probability. That probability represents the likelihood of a lap time being faster (or slower) than your chosen value.

Finding probability between two lap times:

1. Calculate the z-score for each lap time.
2. Look up the cumulative probability for each z-score.
3. Subtract the smaller probability from the larger one.

For example, if lap times are normally distributed with $\mu = 1{:}38$ and $\sigma = 3$ seconds, and you want the probability of a lap between 1:35 and 1:40, you'd compute both z-scores, find each cumulative area, and subtract.

Interpretation of normal distribution parameters

Normal distributions for racing data are unimodal (one peak, representing the most common lap time) and symmetric (the left and right halves are mirror images of each other).

• The mean ($\mu$) is the average or expected lap time. A lower mean indicates faster average performance. Comparing Driver A's mean of 1:40 to Driver B's mean of 1:42 tells you Driver A is, on average, two seconds quicker per lap.
• The standard deviation ($\sigma$) measures consistency. A smaller $\sigma$ means lap times cluster tightly around the mean, while a larger $\sigma$ means more variability. Driver A with $\sigma = 1.5$ sec is far more consistent than Driver B with $\sigma = 3$ sec.

Comparing both parameters together gives you the full picture. Driver A (mean 1:40, SD 1.5 sec) is both faster and more consistent than Driver B (mean 1:42, SD 3 sec). In racing strategy, consistency can matter just as much as raw speed.

A note on skewness: real lap time data isn't always perfectly symmetric. If a distribution is right-skewed, it means occasional slow laps (due to traffic, mistakes, or pit issues) pull the tail out to the right. A truly normal distribution has zero skewness, so noticeable skew suggests the normal model may not be a perfect fit.

Empirical rule for lap time variability

The empirical rule (also called the 68-95-99.7 rule) gives you quick benchmarks for how data spreads in any normal distribution:

1. About 68% of lap times fall within $\mu \pm 1\sigma$
2. About 95% of lap times fall within $\mu \pm 2\sigma$
3. About 99.7% of lap times fall within $\mu \pm 3\sigma$

Suppose a driver has a mean lap time of 1:40 with $\sigma = 2$ seconds:

• 68% of laps fall between 1:38 and 1:42
• 95% of laps fall between 1:36 and 1:44
• 99.7% of laps fall between 1:34 and 1:46

Identifying outliers: Any lap time beyond $\mu \pm 3\sigma$ is extremely rare (only 0.3% of the time). A lap of 1:32 from this driver would be 4 standard deviations below the mean, which signals something genuinely exceptional or unusual worth investigating.

You can also use the empirical rule to check whether a dataset actually follows a normal distribution. If you observe 75% of lap times within one SD (instead of the expected 68%) or only 92% within two SDs (instead of 95%), the data may deviate from normality. Kurtosis helps quantify this: it measures how peaked or flat the distribution is compared to a standard normal curve. High kurtosis means lap times are more concentrated around the mean with heavier tails; low kurtosis means a flatter, more spread-out shape.

Statistical inference for lap times

These tools let you move from describing a single dataset to drawing conclusions about a larger population of lap times.

• The central limit theorem states that the sampling distribution of mean lap times approaches a normal distribution as sample size increases, regardless of the original distribution's shape. This is why normal distribution methods work so well even when individual lap times aren't perfectly normal.
• Confidence intervals give you a range of plausible values for the true population mean lap time. A 95% confidence interval built from a sample of 30 laps tells you where the driver's true average likely falls.
• Hypothesis testing lets you formally compare performance. For example, you could test whether Driver A's mean lap time is significantly different from Driver B's, rather than relying on eyeballing the difference.