According to Occam's Razor, when two models fit observations equally well, which should be chosen?

• A. The most complex
• B. The simplest
• C. The model with the fewest variables
• D. The model that fits the data best and is simple

Answer: (B)

Occam's Razor says to prefer the simplest explanation when two models explain the observations equally well. The reason is that simpler models make fewer assumptions and have fewer moving parts, so they’re less likely to fit random noise in the data and are more likely to generalize to new situations. If two models describe the data with the same accuracy, the extra complexity of a more elaborate model doesn’t buy you any real explanatory power and can actually hurt predictive performance on new data. So the best choice in this tie is the simplest model.