Momentary maximizing in concurrent schedules with a minimum interchangeover interval.

Taylor et al. (1993) tested pigeons on two side-by-side keys. Each key paid off on its own VI schedule. The birds could hop left or right at any time, but after each hop they had to wait a set number of seconds before the next hop.

The team slowly raised this forced-wait time from 0 s up to 12 s. They watched how the birds split their pecks between the keys.

With no wait time the birds matched: they pecked in the same ratio as the payoffs. As the wait grew, the split moved toward 50-50 even though the payoff rates stayed the same.

Longer change-over delays flattened the matching curve, giving a win to momentary-maximizing theory over simple molar matching.

Fantino (1969) saw matching break down when the schedules were very unequal; C et al. show it also breaks when you add hop delays. Both tell the same story: matching is fragile when local costs rise.

Avellaneda et al. (2025) later built a new matching formula that lets sensitivity drop as overall reinforcement falls, echoing the flattening C et al. saw.

White (1979) found that hop-by-hop choices follow a simple Markov chain with no memory; C et al. add that forcing a pause between hops is enough to push birds away from strict matching.

If you run concurrent programs in the lab or in practice, keep change-over delays short. Long delays wash out the sensitivity you want to study or to use for treatment decisions. When you see flat choice, first check the switch cost before you tweak the payoff rates.