I greatly enjoyed this class. I found the material very interesting and left the class with a strong understanding of probability. McMillan is an excellent lecturer and clearly wants students to succeed. He gives generous partial credit if he can understand your reasoning, and even invited us to speak with him in office hours about our midterms and was willing to give additional points back if you explained your process. He was also very receptive to student feedback, making changes throughout the semester to improve the course. His course website contained typed notes that were incredibly useful, and I found them to be much easier to understand than the textbook.

The structure of the class was lectures on M/W, and discussions on Friday with 4-5 student presenters per week: each presenting student would be given a homework problem to explain on the board and were given 25 minutes to prepare.

In terms of workload, there were two homework sets per week, taking 4-6 hours combined. There were also two group programming projects that were 1-2 weeks long and were 15% of your total grade. They were relatively simple, but they did take up a lot of time during the weeks when they were active, and a bad group member could ruin your grade.

Some grading could feel a little arbitrary at times. Homework and exam grades were clear, but grades such as "Participation" seemed pretty much random. Same with peer evals: in every other class, it seems like you typically get a 100% on peer evals unless you did especially badly, but in this class my teammates told me I did a great job and yet I lost quite a few points. It is possible my teammates lied to me and were very negative on the evals, but I suspect McMillan is using some strange formula that results in lower grades than most other classes.

This was all especially annoying since most grades (peer evals for both projects, participation, attendance, any extra credit assignments) weren't given back until well after the final exam, so you had no way to ask questions about how the grades were calculated or to figure out where you went wrong to improve your future performance.

Finally, exams were relatively tough but graded on a generous square-root-esque curve, and I did not find them to feel unfairly difficult. If you understand all of the homework problems, I would expect you to score in the A range for all exams (including the final) post-curve.

Overall: take it. Excellent material, lectures, and notes, and a professor who definitely cares. Just know you won't have visibility into a meaningful chunk of your grade until well after the class is over.

Professor McMillan teaches probability very differently than the other APMA 3100 sections. Instead of frequent mastery checkpoints, McMillan has a more traditional grading system with a few midterms throughout the semester, which are graded on an approximately square root curve. However, his students do still take the same final as other sections. To account for his more challenging exams, the required averages to achieve each letter grade are lower than normal. Several extra credit opportunities are also presented throughout the semester.

In terms of assignments, he has problem sets due twice each week with 5-10 problems each in lieu of worksheets. These problems are generally more challenging than the examples in the notes and require a deeper understanding of the course material to complete. No lecture time is dedicated to completing them. Students are also required to complete two coding projects (just like other APMA 3100 sections) with preassigned groups, which McMillan tweaks after the first project based on results from a peer-assessment survey.

In terms of teaching style, he has a more theoretical approach. During lecture, McMillan does walk through solving a few examples and practice problems, but a lot of the class is spent providing conceptual explanations and deriving formulas and theorems. Attendance is not required except on Fridays for presentations, which consist of several students presenting their solutions to homework problems. His lecture notes for each week, which present the course material in a very compact manner with sections for theory, examples, and practice, are available online before lecture and an annotated versions of these notes are posted, though not after every class.