After the voting proves unreliable, the instructor should try the same decay by using the 10 or 100 sample size button. Students could suggest a constructive re-phrasing of the polling question. By putting ranges on the possible outcome, students should see the power in statistical observations. In other words, their voting should become more predictable and successful.
Going through the questions on the last page, the class will see that more precise information is revealed as the number of data grows.
A challenging extension would be to pick a particle with approximately three decay modes and note the probability of various decays within samples of 10, then 100, and then 1000. Students could divide into groups to study a particular particle. They could accumulate as much data as your lesson plan allows. Additionally, groups from different classes could pool their data in an effort to get more precise results for a particular particle.
In the end, each of your experimental groups could compare their final tallies with the accepted values as published by the Particle Data Group at http://pdg.lbl.gov. If there is a big discrepancy, you should probably attribute it to the program's random number routine.
Depending on the speed of your computer, the time it takes for the "1000" button to finish sampling could be anything from 20 seconds to a few minutes. Students should have worked on both of the other decay pages before trying this one.
Decays are generated by a random number routine that produces samples from data accepted and published by the Particle Data Group. Many of the antiparticle results are not based on experimental evidence, however. For the sake of verisimilitude, they are programmed to mirror the better known decay behavior of their respective particles.