In Gilad, Hemo, Micali, Vlachos, Zeldovich’s paper “Algorand: Scaling Byzantine Agreements for Cryptocurrencies”, in Appendix C.1, in the proof of how safety is assured under weak synchrony, could someone explain:

- p. 21, column 1, para 4: How the number 4,100 is arrived at in the expression i + 3j > 4,100?
- The expected number of malicious members on the committee is 0.2 * 10,000 = 2000. This is OK. But why do they further multiply this by 0.2 to get the expected value for j as 0.2 * 2000 = 400? This is unclear.
- Is the multiplication of j by 3 due to 3 steps within each while loop iteration of BinaryBA*?