How do these 1st, 2nd, 3rd choices work on moderator elections?

Question

With four current candidates ("M", "G", "Y" and "W") and three choices (1st, 2nd and 3rd) there are 24 ways to vote all three, 12 ways to vote 1st and 2nd and skip 3rd, and four ways to vote for only 1st.

I'll ignore voting only for 2nd and/or 3rd and skipping first and/or second.

What are the guidelines for the voting strategy here?

1. Are there any benefits for voting for any less than all three choices (1st, 2nd and 3rd)?
2. Does the voting robot let us get away with checking any less than all three?

Python

from itertools import permutations
for n in range(3, 0, -1):
    options = list(permutations('MGYW', n))
    options = [''.join(option) for option in options]
    print(n, len(options), options)
    print('')

results in:

(3, 24, ['MGY', 'MGW', 'MYG', 'MYW', 'MWG', 'MWY', 'GMY', 'GMW', 'GYM', 'GYW', 'GWM', 'GWY', 'YMG', 'YMW', 'YGM', 'YGW', 'YWM', 'YWG', 'WMG', 'WMY', 'WGM', 'WGY', 'WYM', 'WYG'])

(2, 12, ['MG', 'MY', 'MW', 'GM', 'GY', 'GW', 'YM', 'YG', 'YW', 'WM', 'WG', 'WY'])

(1, 4, ['M', 'G', 'Y', 'W'])

Answer 1

TL;DR

Are there any benefits for voting for any less than all three choices (1st, 2nd and 3rd)?

No.

Does the voting robot let us get away with checking any less than all three?

Yes.

---

The simplest way to understand it is it's more or less working like ranked choice voting.

Let's say we have 100 voters and candidates A, B and C. Our threshold for winning is 51 votes. The first round breaks down like so

• A - 35 votes
• B - 32 votes
• C - 33 votes

Nobody hits the threshold so we eliminate the lowest candidate, which is B. Now we look at the second choice for people who voted for B (assuming all B voters had a second pick)

• A - 10 votes
• C - 22 votes

That means our second round ends up as

• A - 45 votes
• C - 55 votes

C wins, crossing the 51 vote threshold.

Now, if you didn't pick a second choice and voted for B, your vote falls into the exhausted pool (doesn't directly benefit any candidate).

You can see how rounds play out with more candidates by looking at the Stack Overflow 2019 vote results. I should note that I oversimplified the process a bit so it's understandable. In larger elections, where all votes are exhausted and there's still no winner, SE uses Meek STV to pick a winner (if you thought orbital mechanics were complicated...). It's unlikely STV will pick a winner for Space.SE (1 seat, fewer candidates and smaller vote pool)