The results for the WSFS Mark protection Committee election are out!
Given that the Business Meeting yesterday voted to conduct future elections (from 2027, if it is ratified next year) as single multi-member counts, rather than repeated single-seat counts as at present, I was asked to produce my analysis of how that would have worked with this year’s vote.
(Disclosure: I voted for Warren Buff, so my vote started with him and was not transferred in any of the counts, either in real life or in my simulation.)
We know the number of first preferences counted in each round, and we can work out the transfers from Alana Vincent and Andrew January as follows:
| Round 1 | Round 2 | Round 3 | |||
| Alana Vincent | 27 | -27 | 0 | 0 | |
| Andrew January | 25 | +23 | 48 | -48 | 0 |
| Warren Buff | 12 | +3 | 15 | +18 | 33 |
| Joni Dashoff | 11 | 11 | +8 | 19 | |
| Todd Dashoff | 3 | 3 | +5 | 8 | |
| Judy Bemis | 3 | 3 | 3 | ||
| Arturo Serrano | 2 | 2 | +6 | 8 | |
| Gary Blog | 1 | 1 | +4 | 5 | |
| Linda Ross-Mansfield | 1 | 1 | 1 | ||
| Roseanna Pendlebury | 0 | 0 | +7 | 7 | |
| not transferred | +1 | 1 | 1 |
With 85 votes in total and three places to elect, the quota is 85/(3+1)+.001 = 21.251
Both Alana Vincent and Andrew January exceed the quota on the first count, and are elected, Alana Vincent first with a surplus of 5.749 and Andrew January second with a surplus of 3.749.
Where do Alana’s votes go?
1 is not transferred
3 go to Warren Buff
23 go to Andrew January, and are then transferred along with the rest of his votes in the real-life Round 3.
For purposes of the thought experiment, I’m going to assume that Andrew’s original votes, and the ones he inherited from Alana, behaved similarly to each other, and split ties in Andrew’s favour.
So, let’s assume that the number of voites transferred from Alana to the next candidate after Andrew was as follows:
| Warren Buff | 12 |
| Joni Dashoff | 4 |
| Todd Dashoff | 2 |
| Judy Bemis | 0 |
| Arturo Serrano | 3 |
| Gary Blog | 2 |
| Linda Ross-Mansfield | 0 |
| Roseanna Pendlebury | 3 |
| not transferred | 1 |
The surplus value of these votes is Alana’s surplus (5.749) divided by the number of transferable votes (26) and rounded down: 0.221
So the second count looks like this:
| Count 1 | Count 2 | ||
| Alana Vincent | 27 | -5.749 | 21.251 |
| Andrew January | 25 | – | 25 |
| Warren Buff | 12 | +2.652 | 14.652 |
| Joni Dashoff | 11 | +0.884 | 11.884 |
| Todd Dashoff | 3 | +0.442 | 3.442 |
| Judy Bemis | 3 | 3 | |
| Arturo Serrano | 2 | +0.663 | 2.663 |
| Gary Blog | 1 | +0.442 | 1.442 |
| Linda Ross-Mansfield | 1 | 1 | |
| Roseanna Pendebury | 0 | +0.663 | 0.663 |
| not transferred | +0.003 | 0.003 |
(The extra .003 in the Non-Transferable row is an accounting artefact to take account of rounding.)
Next we transfer Andrew January’s first preference votes. Again I’m assuming that they split much the same way as the votes he got from Alana Vincent, as follows:
| Warren Buff | 9 |
| Joni Dashoff | 4 |
| Todd Dashoff | 3 |
| Judy Bemis | 0 |
| Arturo Serrano | 3 |
| Gary Blog | 2 |
| Linda Ross-Mansfield | 0 |
| Roseanna Pendlebury | 4 |
| not transferred | 0 |
The surplus value of these votes is Andrew’s surplus (3.749) divided by the number of transferable votes (25) and rounded down: 0.149
So the third count looks like this:
| Count 1 | Count 2 | Count 3 | ||
| Alana Vincent | 27 | 21.251 | 21.251 | |
| Andrew January | 25 | 25 | -3.749 | 21.251 |
| Warren Buff | 12 | 14.652 | +1.341 | 15.993 |
| Joni Dashoff | 11 | 11.884 | +0.596 | 12.480 |
| Todd Dashoff | 3 | 3.442 | +0.447 | 3.889 |
| Judy Bemis | 3 | 3 | 3 | |
| Arturo Serrano | 2 | 2.663 | +0.447 | 3.110 |
| Gary Blog | 1 | 1.442 | +0.298 | 1.740 |
| Roseanna Pendlebury | 0 | 0.663 | +0.596 | 1.259 |
| Linda Ross-Mansfield | 1 | 1 | 1 | |
| not transferred | 0.003 | +0.024 | 0.027 |
Linda Ross-Mansfield has the fewest votes at this stage. We know from the real-life election that her vote transferred to Warren Buff, so the fourth count looks like this:
| Count 1 | Count 2 | Count 3 | Count 4 | ||
| Alana Vincent | 27 | 21.251 | 21.251 | 21.251 | |
| Andrew January | 25 | 25 | 21.251 | 21.251 | |
| Warren Buff | 12 | 14.652 | 15.993 | +1 | 16.993 |
| Joni Dashoff | 11 | 11.884 | 12.480 | 12.480 | |
| Todd Dashoff | 3 | 3.442 | 3.889 | 3.889 | |
| Arturo Serrano | 2 | 2.663 | 3.110 | 3.110 | |
| Judy Bemis | 3 | 3 | 3 | 3 | |
| Gary Blog | 1 | 1.442 | 1.740 | 1.740 | |
| Roseanna Pendlebury | 0 | 0.663 | 1.259 | 1.259 | |
| Linda Ross-Mansfield | 1 | 1 | 1 | -1 | |
| not transferred | 0.003 | 0.027 | 0.027 |
Roseanna Pendlebury now has the fewest votes – the 7 that she got from Alana Vincent and Andrew January.
We know that in real life, 1 of these was not transferred, 3 went to Warren Buff and 1 each to Joni Dashoff, Todd Dashoff and Artur Serrano.
We don’t know which votes came from Alana and which from Andrew, but let’s imagine that the 3 from Alana went to Warren and the other 4 split evenly.
So the fifth round looks like this.
| Count 1 | Count 2 | Count 3 | Count 4 | Count 5 | ||
| Alana Vincent | 27 | 21.251 | 21.251 | 21.251 | 21.251 | |
| Andrew January | 25 | 25 | 21.251 | 21.251 | 21.251 | |
| Warren Buff | 12 | 14.652 | 15.993 | 16.993 | +0.663 | 17.656 |
| Joni Dashoff | 11 | 11.884 | 12.480 | 12.480 | +0.149 | 12.629 |
| Todd Dashoff | 3 | 3.442 | 3.889 | 3.889 | +0.149 | 4.038 |
| Arturo Serrano | 2 | 2.663 | 3.110 | 3.110 | +0.149 | 3.259 |
| Judy Bemis | 3 | 3 | 3 | 3 | 3 | |
| Gary Blog | 1 | 1.442 | 1.740 | 1.740 | 1.740 | |
| Roseanna Pendlebury | 0 | 0.663 | 1.259 | 1.259 | -1.259 | |
| Linda Ross-Mansfield | 1 | 1 | 1 | |||
| not transferred | 0.003 | 0.027 | 0.027 | +0.149 | 0.176 |
Gary Blog now has the fewest votes – his own first preference, and 2each from Andrew January and Alana Vincent.
In real life, his first preference went to Warren Buff and the other 4evenly between Warren Buff, Joni Dashoff, Todd Dashoff and Arturo Serrano.
I’m going to imagine that the 2 votes that started with Alana went to Warren and Joni, and the 2 votes that started with Andrew went to Todd and Arturo.
So the sixth round looks like this:
| Count 1 | Count 2 | Count 3 | Count 4 | Count 5 | Count 6 | ||
| Alana Vincent | 27 | 21.251 | 21.251 | 21.251 | 21.251 | 21.251 | |
| Andrew January | 25 | 25 | 21.251 | 21.251 | 21.251 | 21.251 | |
| Warren Buff | 12 | 14.652 | 15.993 | 16.993 | 17.656 | +1.221 | 18.877 |
| Joni Dashoff | 11 | 11.884 | 12.480 | 12.480 | 12.629 | +0.221 | 12.850 |
| Todd Dashoff | 3 | 3.442 | 3.889 | 3.889 | 4.038 | +0.149 | 4.187 |
| Arturo Serrano | 2 | 2.663 | 3.110 | 3.110 | 3.259 | +0.149 | 3.408 |
| Judy Bemis | 3 | 3 | 3 | 3 | 3 | 3 | |
| Gary Blog | 1 | 1.442 | 1.740 | 1.740 | 1.740 | -1.740 | |
| Roseanna Pendlebury | 0 | 0.663 | 1.259 | 1.259 | |||
| Linda Ross-Mansfield | 1 | 1 | 1 | ||||
| not transferred | 0.003 | 0.027 | 0.027 | 0.176 | 0.176 |
Judy Bemis now has the fewest votes – her 3 first preferences.
We know that 2 of these went to Joni Dashoff and 1 to Warren Buff.
So the seventh round looks like this:
| Count 1 | Count 2 | Count 3 | Count 4 | Count 5 | Count 6 | Count 7 | ||
| Alana Vincent | 27 | 21.251 | 21.251 | 21.251 | 21.251 | 21.251 | 21.251 | |
| Andrew January | 25 | 25 | 21.251 | 21.251 | 21.251 | 21.251 | 21.251 | |
| Warren Buff | 12 | 14.652 | 15.993 | 16.993 | 17.656 | 18.877 | +1 | 19.877 |
| Joni Dashoff | 11 | 11.884 | 12.480 | 12.480 | 12.629 | 12.850 | +2 | 14.850 |
| Todd Dashoff | 3 | 3.442 | 3.889 | 3.889 | 4.038 | 4.187 | 4.187 | |
| Arturo Serrano | 2 | 2.663 | 3.110 | 3.110 | 3.259 | 3.408 | 3.408 | |
| Judy Bemis | 3 | 3 | 3 | 3 | 3 | 3 | -3 | |
| Gary Blog | 1 | 1.442 | 1.740 | 1.740 | 1.740 | |||
| R Pendlebury | 0 | 0.663 | 1.259 | 1.259 | ||||
| L Ross-Mansfield | 1 | 1 | 1 | |||||
| not transferred | 0.003 | 0.027 | 0.027 | 0.176 | 0.176 | 0.176 |
Arturo Serrano now has the fewest votes – his 2 first preferences, and also 3 votes from each of Alana and Arthur.
We know that 1 of his first preferences transferred to Warren, but the other went to Alana, and we have no good information about the rest.
For the sake of illustration let’s split the first preference votes evenly between Warren and Joni, and the others evenly between the three remaining candidate.
So the eighth round looks like this:
Arturo Serrano now has the fewest votes – his 2 first preferences, 3 votes from Alana and 5 votes from Andrew.
We know that one of his first preferences transferred to Warren, but the other went to Alana in the first instance, and we have no good information about the rest.
For the sake of illustration let’s split the 2 first preference votes evenly between Warren and Joni, the 3 votes from Alana between all three remaining candidates, and the 5 votes from Andrew 2 to Warren, 2 to Joni and 1 to Todd.
So the eighth round looks like this:
| C1 | C2 | C3 | C4 | C5 | C6 | C7 | C8 | ||
| Alana Vincent | 27 | 21.251 | 21.251 | 21.251 | 21.251 | 21.251 | 21.251 | 21.251 | |
| Andrew January | 25 | 25 | 21.251 | 21.251 | 21.251 | 21.251 | 21.251 | 21.251 | |
| Warren Buff | 12 | 14.652 | 15.993 | 16.993 | 17.656 | 18.877 | 19.877 | +1.519 | 21.396 |
| Joni Dashoff | 11 | 11.884 | 12.480 | 12.480 | 12.629 | 12.850 | 14.850 | +1.519 | 16.369 |
| Todd Dashoff | 3 | 3.442 | 3.889 | 3.889 | 4.038 | 4.187 | 4.187 | +0.370 | 4.557 |
| Arturo Serrano | 2 | 2.663 | 3.110 | 3.110 | 3.259 | 3.408 | 3.408 | -3.408 | |
| Judy Bemis | 3 | 3 | 3 | 3 | 3 | 3 | |||
| Gary Blog | 1 | 1.442 | 1.740 | 1.740 | 1.740 | ||||
| R Pendlebury | 0 | 0.663 | 1.259 | 1.259 | |||||
| L Ross-Mansfield | 1 | 1 | 1 | ||||||
| not transferred | 0.003 | 0.027 | 0.027 | 0.176 | 0.176 | 0.176 | 0.176 |
Warren Buff now exceeds the quota and is elected third, after Alana Vincent and Andrew January.
There are various improvements that could be made to the presentation of the above, but I think this captures the likely flow of voting, without too many heroic assumptions.
The process would be completed in 8 rounds of counting, rather than the 14 that it took in real life.
There were 76 transfers of votes here, compared to 84 in real life.
The result was actually the same as in real life; but that was because the opposition to the three elected candidates was less than a quarter of the total vote.
Hopefully we can ratify this change for next year.