The theory behind constituency rigging is simple. Large parties have a great chance of winning a seat bonus in the small three-seat constituencies, while the larger five-seat constituencies make it easier for the smaller parties to win a seat.
For instance, a big party can win two in a three-seater with between 45 per cent and 50 per cent support, giving it two-thirds of the seats with less than half the vote. However, if the party vote drops into the low 40 per cent range, it will only win one seat and will see a lot of its votes wasted.
On the other hand, in four-seat constituencies a big party can win two out of four with 40 per cent or less, while in five-seaters it can take two with under 30 per cent. A good gerrymander depends on a fine anticipation of the party's likely share of the vote with the shape of the constituencies tailored to take maximum advantage.
Kevin Boland's 1969 revision was a classic example of a gerrymander that worked. Dublin, where Fianna Fáil rarely exceeded 40 per cent, was divided into eight four-seaters and two three-seaters. By contrast in the west, south and the Border counties, where the party vote approached 50 per cent, there were 20 three-seaters and just four four-seaters.
Jim Tully's 1974 revision was a gerrymander that went horribly wrong. He reversed the Boland tactic by introducing three-seaters across Dublin and four-seaters across rural Ireland. The theory was that Fianna Fáil would get just one out of three and two out of four.
The ploy backfired because Fianna Fáil won about 5 per cent more of the vote than Tully had expected. In the 1977 election it won two out of three, and three out of four across the country, giving the party its biggest ever Dáil majority.