Trying to predict the outcome of the Saturday week’s South Australian election is a mug’s game. After an early burst of polling optimism for Nick Xenophon’s SA-BEST that had the party on 32% and Xenophon as preferred premier, more recent polling suggests a significant softening. The Australian claimed on the weekend that the party’s vote was in “free fall”, having fallen to 21% since December, but admitted this was partly methodological, given Xenophon wasn’t running candidates in every seat. That 21% is actually 27% in the 36 seats the party is running in.

A number of seat-based polls suggest SA-BEST will struggle to get the kind of primary vote that will enable it to pick up seats, which it will have to do so off major party preferences. But 27% for a party that didn’t exist several months ago is a remarkable achievement for Xenophon and testament to the deep level of disaffection toward the major parties in South Australia. For Labor, the disaffection is understandable — it has clung to power now since the dawn of the century and is long past its use-by date. For Liberals, you can only wonder how many different ways a party can invent to not win elections.

If SA-BEST maintains something like the level of support indicated in the polls, it will match or better the achievement of One Nation in 1998 in Queensland, when the racist party secured just under 23% of the vote. The difference will likely be that One Nation managed to secure 11 seats, whereas SA-BEST will have to get lucky with preference flows to pick up even a handful. Instead, its preferences — who knows which way they’ll go; Xenophon is running an open ticket — will likely determine which of the major parties governs.

One Nation had an advantage that Xenophon and co don’t have: its vote was concentrated in regional seats, where it achieved primary vote levels there that saw it easily defeat Nationals MPs and edge out Labor MPs with Coalition preferences. SA-BEST isn’t a party of regional grievance (except in the sense that it is a party of grievance for the whole state), and it wouldn’t matter in a state like South Australia anyway, which lacks the large regional population that characterises the Queensland electorate.

SA-BEST is thus in a democratic limbo — not quite a major party but much larger than a Greens-style minor-party. The Greens in South Australia are on just 7%, barely ahead of Cory Bernardi’s extreme right Australian Conservatives on 6%; significantly, both parties are preferencing against Xenophon’s party. SA-BEST is as great a threat to them as it is to the major parties, perhaps more so: the more political disaffection Xenophon can vacuum up, the less there is for bit-players like the Greens and Bernardi to feed on.

The limbo is created by the fact that, usually, there just isn’t enough room for more than two major parties; if both are getting over 30%, even the strongest third party will struggle to compete. One major party needs to collapse dramatically to allow a new player in: that’s what happened in Queensland in 1998, with the Coalition vote falling 18 points to 31%. Even so, One Nation still needed the concentration effect of regional electorates to secure its seats. Even if SA-Best picked up 30% of the vote in its seats — an implausible achievement, and an historic one — it may well emerge with no seats at all if preferences don’t flow for it.

That sort of result would signal that Nick Xenophon’s big political gamble had failed. But it would also signal that something’s badly wrong in the South Australian political system if more than a quarter of disaffected voters end up unrepresented.