There’s a bit of a debate now in the Labor Party over whether factions have any ideological ties, or are merely competing executive placement agencies. Frequently, factions have broken into fractions – sub-units which are even less likely to be ideologically driven than factions.
But if you start by assuming the ALP will always have factions, there’s still the question – how many factions is best? Here, economics may have some insights. Monopoly markets usually produce lower quality goods at higher prices than duopolies or perfect markets. This would suggest that states with a lot of ALP factions (or none at all – so everyone competes on their own terms) will produce better Labor politicians than states with one dominant faction.
Does this hold in practice? Not perfectly, but the best example of a state dominated by one faction is NSW, where the talent pool seems to be shallower than in most other states.