What type of distribution is used for the number of bidders?

• A. Normal distribution
• B. Discrete integer uniform distribution
• C. Poisson distribution
• D. Exponential distribution

Answer: (B)

The number of bidders is a discrete, bounded count outcome for a single auction, so the best modeling choice is a discrete uniform distribution when you have no information favoring any particular count within a plausible range. This means every integer value in that range is equally likely, which is a neutral, noninformative assumption that keeps the model simple and avoids introducing bias toward small or large bidder counts.

Why the other options fit poorly here: a normal distribution is a continuous, unbounded model, which isn’t natural for a count that must be a whole number and is typically limited to a realistic range. A Poisson distribution models the number of events occurring in a fixed interval with a rate parameter and tends to produce a skew toward smaller counts with probability mass spreading to larger values, which isn’t appropriate for a bounded bidding scenario without evidence of such a pattern. An exponential distribution describes time between events, not a count of discrete outcomes, and is also continuous.

So, using a discrete uniform distribution captures the idea of equal likelihood for each possible bidder count within the intended range when there’s no data to suggest a particular count is more probable.