Believe it or not, I'm still receiving a steady stream of messages and comments from people who are convinced they are world-leading experts on the Holyrood voting system, and who want to "explain" to me some mistakes I've supposedly made in my recent posts. The super-confident cluelessness of some of these people is quite entertaining in a way, but I'm becoming concerned that casual readers of the comments section may be misled by that confidence, so I'm going to set a few of the dafter misconceptions straight.
"Christ on a bicycle, what are you saying here? How can you dispute that a party with 3% of the list vote would take 4 seats? AMS is a proportional voting system and 4 seats is 3% of 129. Stop being an idiot, James."
I can dispute it very easily, because under the Holyrood system seats are not distributed proportionally on a national basis. Scotland is split into eight regions, each with between fifteen and seventeen seats. That means, unless something weird happens, that it's very unlikely a party would take a seat in any region with less than 5% of the list vote. Conceivably a party with 3% of the national vote might nick a seat or two if they do better in certain regions than in others - but the most likely outcome is that they would take zero seats.
"Christ on a bicycle, what are you saying here? How can you produce a hypothetical example of a Holyrood result where the list vote of the parties only adds up to 96%? All list votes count under AMS. Stop being an idiot, James."
I can do that because I'm taking into account the obvious fact that there will be a number of minor parties and independents on the list ballot, which in combination will take a non-trivial share of the vote. Indeed, any hypothetical example that doesn't take account of that point is flawed, although exactly what share of the vote minor parties will take can only be speculative. Incidentally, it's doubtful whether it can truly be said that "all list votes count under AMS" - to all intents and purposes votes for very small parties don't really count towards the seat allocation, and that means larger parties can sometimes win slightly more seats than they otherwise would on any given share of the vote.
"Christ on a bicycle, what are you saying here? How can you claim that the SNP can win significantly more seats than their share of the list ballot? AMS is a proportional system. Stop being an idiot, James."
I can say that because of a little something called constituency seats. If the largest party takes all or nearly all of the constituency seats in a region, there won't be enough list seats to fully compensate the other parties and make the overall result proportional. There are multiple examples of that scenario occurring in past Holyrood elections.
"Christ on a bicycle, what are you saying here? How can you produce a hypothetical example where the SNP take an average of 6.5 seats per region on 29% of the list vote? That list vote only entitles the SNP to two seats per region, which requires there to be 4.5 SNP constituency seats per region, and if they win that many they'd be over their 'quota' and wouldn't take any list seats anyway. Stop being an idiot, James."
A lot of that one is utter gibberish, but instead of trying to disentangle it, I'll just make this really simple. There is no 'cap' of 4.5 constituency seats per region - that would be an arithmetically impossible cap to enforce anyway. A party can win as many as 73 constituency seats nationwide, regardless of how high or low their share of the list vote is - and that's an average of 9.1 seats per region, not 6.5.