Some are celebrating, others are sad, but the result is there for all to see.

Now that the result is confirmed, although it has been clear for some time, it is possible to begin the analysis of what happened. Many people will have theories about why Zed Seselja was defeated and why David Pocock won.

As a first cut at a statistical analysis of the results I have looked at the fall in the Liberal vote by electorate and across some of the range of polling booths across Canberra.

The first reflection on the result is that Zed lost everywhere.

In the southern seat of Bean, where he might have expected to do better, the Liberal vote overall was 27.17 per cent compared to 34.84 per cent in 2019. A fall of 7.67 per cent.

In the central seat of Canberra, where the Liberals traditionally do badly, the Seselja vote was a mere 20.25 per cent, a fall of 7.48 per cent from the 2019 level of 27.73 per cent.

The northern seat of Fenner saw a similar fall. In 2019 the combined Liberal vote in this seat was 34.37 per cent. In 2022 it fell to 26.92 per cent, a decline of 7.45 per cent.

This shows a remarkable similarity across Canberra at the large scale of electorate-by electorate.

David Pocock's vote showed a pattern not far removed from this. His vote ranged from 19.02 per cent in Fenner to 21.33 per cent in Bean and 23.11 per cent in Canberra. This shows he was taking votes from the Greens and Labor as well as from Zed Seselja directly.

At the more local level, on a booth-by booth analysis, the variation is greater, although the pattern of decline in support for former senator Seselja is common across all booths and regions.

What the statistics appear to show is that some of the biggest swings against the former senator occurred in some of the Liberal party's strongest areas.

In Conder, where the vote for the Liberal Senate ticket was over 40 per cent in 2019 the result in 2022 was a massive 10 per cent drop in support and in Gordon the decline in support was from 44.7 in 2019 to 35.2 per cent in 2022.

These were not the biggest swings against the Liberals in Bean in 2022, that title goes to Mawson with a more than 11 per cent swing (unless you count Norfolk Island where for some reason the swing was 22 per cent).

The apparent pattern becomes clearer when the booth figures in Canberra and Fenner are examined.

In the leafy suburb of Deakin, which bears some resemblance to the "teal" suburbs in Sydney and Melbourne, the drop in support for the Liberal Senate ticket was 12.5 per cent!

READ MORE:

In Fenner, the usually strong Liberal polling booths at Forde and Nicholls showed swings of more than 10 per cent.

The swings against Zed Seselja in the strongest Liberal areas do not tell the whole story. There was a substantial swing everywhere.

In the traditionally strong Labor booths in the inner north there were not many votes for the Liberals to lose. But lose votes they did. The lowest vote for the former senator was in Lyneham, where his ticket polled a mere 12.3 per cent, a drop of 4.6 per cent from a pretty dismal result in 2019.

Of course, some of the credit must go to David Pocock. It appears that he did not fall for the trap of focusing his campaign in the anti-Liberal areas like the inner north. He did well there and won votes from all the other parties in a way which helped him stay ahead of the Greens.

However, the figures reinforce the anecdotal evidence during the campaign that he campaigned effectively all across the ACT, and he appears to have done particularly well in the Weston Creek area.

Major electoral upsets always have a number of causes. In this case there was a national mood for change and there was a very appealing independent candidate. Without these factors there would not have been a change.

However, the magnitude and location of the big swings in the ACT Senate election in 2022 reinforce the view that Zed Seselja was the main cause of the Liberals failure to win a seat in Canberra for the first time.