So, looking at the final scores, my Sunday prediction was wrong in 19 out of 39 constituencies – just under half. Two of these were simple error on my part – I mixed up Labour and the Social Democrats in Wicklow, and FF and FG in Mayo.

But in general I did not appreciate how transfer-toxic FF (and to a lesser extent FG) have become. The best illustration here is Dublin South-Central, where neither of the traditional parties won a seat, despite starting in second and third place on first preferences.

The older parties did not even transfer well internally. The best example is Tipperary where Fianna Fail had over a quota of first preferences, and Fine Gael had more than 0.8 of a quota, normally enough to win a seat, but FF only just scraped into the last seat and FG were runners-up. The FG front-runner got less than 60% of his running-mate's transfers; the FF front-runner got only 35% from their second candidate when she was eliminated!

By contrast, Greens, Social Democrats and Independents were often more transfer friendly than I expected. Independent candidate Dean Mulligan started in ninth place in Dublin Fingal, but came with 150 votes of winning the fifth and final seat.

All this meant that my Sunday prediction was way out from reality, except, oddly enough, for SF who I correctly called at 37 seats. FF and FG underperformed my Sunday prediction, on 38 and 35 respectively rather than 47 and 37. Labour got the 6 that I expected once I sorted out my Wicklow mistake. The others all did better: 21 Independents (including Aontu) rather than 17, 12 Greens rather than 9, 6 Social Democrats rather then 3, 5 SPBP rather than 3.

All in all, a useful object lesson in humility for me!