FI schedules and persistence at gambling in the U.K. betting office.
Adult gamblers bunch bets right before a fixed payout, proving FI schedule control works outside the lab.
01Research in Context
What this study did
Carr (1979) watched regular gamblers inside a London betting shop. The shop paid winners every 30 minutes. The study timed when people placed bets during each half-hour window.
No one was told to change behavior. The goal was to see if the fixed 30-minute payout shaped betting the way lab FI schedules shape pigeon or rat responses.
What they found
Most bets landed in the last two minutes before the payout clock reset. This matches the classic FI "scallop": long pause, then quick burst right before reinforcement.
The pattern looked like lab data from Dews (1978) with monkeys and Blackman (1970) with pigeons, but it happened with real money and adult humans in a noisy shop.
How this fits with other research
Taylor et al. (1993) later showed 3-month-old infants can pause on FI schedules and pick bigger delayed rewards. Together with Carr (1979), the line is clear: FI control shows up from babies to bookies.
Wanchisen et al. (1989) found rats with a VR history later act strange on FI. Gamblers also come with long VR slot-machine histories, yet they still show clean FI scallops in the betting office. Schedule history matters, but the FI clock can override it.
Hamm et al. (1978) saw FI schedules create varied response locations in pigeons. G saw the same schedule tighten response timing in people. Same schedule, different dimension—variability versus timing—both still FI-driven.
Why it matters
If you run token boards, classroom points, or any timed reward, expect the same last-minute rush. Space checkpoints evenly and plan for the pause-burst cycle. Watching a betting shop tells us our clients will likely wait, then sprint, so build extra tasks or choice time during the quiet early minutes to keep engagement steady.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Set a 5-minute visible countdown before the next point delivery and pre-load easy tasks for the quiet first three minutes.
02At a glance
03Original abstract
This study reports on the direct observations of customers in two U.K. betting offices gambling on horse and dog races. These observations revealed that bets were more frequently placed in the last minutes just prior to the start (the OFF), and that this was caused by high-frequency gamblers (customers who had eight or more bets in a session) consistently placing their bets in the last two minutes prior to the OFF. Low-frequency gamblers (three or fewer bets/session) avoided this time period placing their bets earlier, or after the OFF, i.e., on a later race. It was argued that the betting behavior of the "gamblers" could not be explained either in terms of "skillful betting" or solely in terms of variable ratio schedules but was more adequately accounted for in terms of an interval schedule. It was further suggested that time-based schedules might be of heuristic value in generally understanding persistence at gambling while losing.
Journal of applied behavior analysis, 1979 · doi:10.1901/jaba.1979.12-315