How is a PAO image better than a single one-image peg?

A peg gives one image per item; PAO gives one image per three items — triple the data density.

A simple peg system (like a plain Major-System word) is fine, but stores one item per picture. PAO's fixed Person/Action/Object roles let one picture legitimately carry three items without confusion, because each slot is decoded separately. You trade a bigger setup (three attributes per item, well-drilled) for a threefold reduction in how many images and loci you need. For large volumes — decks, long numbers — that trade is decisive.

Go deeper:

• Mnemonic peg system (Wikipedia) — contrasts one-item pegs with the three-item PAO.