Why is it to conduct a large number of trials when simulating genetic crosses?

If you were expecting a ratio of 9:3:3:1 you would have to make sure that you had enough offsrpring to count to give yourself a chance of getting that ratio. If you only had 8 offspring you would never get the ratio. Remember genetic ratios are probabilities of what to expect and you need a large enough sample to ensure accuracy.