The big story was the collapse of the Liberal Party in the geographical constituencies, but there were two other things that caught my attention.
The first was that the pan-democrats were incredibly fortunate with the results. For example, in NT East they got 57% of the vote and won 71% of the seats, and something similar happened on Hong Kong Island. The DAB piled up more votes than they needed, and independent (pro-Beijing) candidates such as Scarlett Pong Oi-Lan in NT East and Priscilla Leung Mei-fun in Kowloon West must have taken votes off the Liberal Party.
The second was that the one party who ought to be really satisfied with their performance were the League of Social Democrats (who are not as moderate as the name might imply). They won three seats, and who would have predicted that Leung Kwok Hung ("Long Hair") would get more votes than any of the other pan-democrat lists in NT East? Yes, by less than 600 votes, but at one stage the SCMP was predicting that he might lose.
Ah, opinion polls. What's this I see in the SCMP? Oh, yes - Exit polls largely accurate despite fears about response rate (subscription required):
The main exit polls appear to have gauged support for many of the candidates accurately despite pollsters' concerns that a poor response rate could compromise their reliability.
Questions about exit polls caused some of the biggest controversies of a largely uneventful campaign. Pan-democrats complained that their Beijing loyalist rivals might set up groups to conduct exit polls and use the results to adjust election strategy. They reminded voters of their right to refuse exit pollsters information. Many voters seemed to heed the reminder, and some even said they had lied to pollsters.
The value of the exit polls, including the one by the University of Hong Kong's public opinion programme to help broadcasters shape their coverage on election night, was called into question. Some candidates cited leaks of exit poll data when they issued "situation critical" calls to bring out more of their voters.
Initial results, released yesterday, from two exit polls - by the HKU programme and Hong Kong Research Association - indicated a response rate of about 50 per cent. The association said the 50 per cent level was barely acceptable to produce representative projections.
While its projections were largely reliable, its results indicated Audrey Eu Yuet-mee, who was second on the Civic Party ticket contesting the Hong Kong Island constituency, stood an "extremely slim chance" of winning. However, Ms Eu won, with her ticket taking 26.4 per cent of the votes cast by the 313,429 electors in the constituency.
In the New Territories West constituency, its poll gave Selina Chow Liang Shuk-yee of the Liberal Party and Fernando Cheung Chiu-hung of the Civic Party a "relatively bigger chance" of winning. But Mrs Chow won only 5.4 per cent of the vote and Dr Cheung just 7 per cent. Both lost. The association's data also gave James Tien Pei-chun, of the Liberal Party, a "relatively bigger chance" of winning in New Territories East. He also lost.
The HKU public opinion programme's exit poll gave five of the six eventual winners in Hong Kong Island a very high chance of victory. In New Territories East, it gave five of the seven eventual winners a very high chance and in New Territories West correctly predicted five of the eight winners.
This is obviously a new definition of "largely reliable". We got several of our predictions wrong, but some were correct, so that's OK then.
Independent pollster Li Pang-kwong, of Lingnan University, who did not conduct exit polling on Sunday, believed the discrepancies were partly attributable to interviewees giving false answers.
Yes, people sometimes lie when asked how they will vote or have voted, and it's something pollsters have to deal with.