The Loss-Chase Game (L-CG)
Loss-chasing describes the continual engagement in gambling behaviour to recover losses, and is associated with the transition from recreational to pathological gambling. We have recently developed a model of loss-chasing in rats. Testing takes place in five-hole operant chambers. Rats have a limited amount of time to earn food reward by making nosepoke responses in one of the response holes. Intermittently, these responses do not produce food reward but instead produce a light signal that the rat has lost time in which it might earn reward. Now, the rat must choose whether to chase this loss (and try to recover the time) or quit in order to move on. If the rat quits, it has to endure a 4s timeout period before being able to earn reward again. If the rat chooses to chase, it will either win or lose. If the rat wins, it can start earning reward again straight away, i.e. there will be no time penalty. However, if the rat loses, it will encounter a time penalty double the interval associated with a decision to quit. At the end of the chase time penalty, the animal chooses again whether to “chase” or “quit”. At this second decision point, the time penalties for both quitting and losing the chase are double what they were for chase 1 (i.e. 8s and 16s respectively) so that the time penalties mount up. The chase and quit options are signalled by illumination of other holes in the five-hole operant chamber. These lights flash on and off during the time-out “loss” periods associated with the choice of a particular hole. Loss periods are also signalled by distinct tones which makes the different stages of the task more discriminable to the rat.
Critically, the odds of winning any given chase are always 50:50, so the averaged time lost through chases is equal to the time lost through quits. In this way, chasing behaviour might reflect the loss-aversion often observed in human choice under uncertainty. Remarkably, we have found that, like human decision-makers, the majority of rats are very likely to chase losses following the first loss within a run of losing gambles (around 80%), and that this behaviour shows a steady reduction as the time penalties resulting from preceding unsuccessful decisions to chase mount up (around 60%). To further emphasise the similarities with the human behaviour, we have pilot data to indicate that serotonergic drugs can modulate a rats’ willingness to chase. This animal model could provide a novel technique for exploring the neurobiologial substrates of one of the central features of pathological gambling, and such research is currently ongoing.
For more information, email info@winstanleylab.com with “Loss-chasing” in the subject heading.
If you would like a copy of our L-CG Med PC program code, please email info@winstanleylab.com with [Code Request] in the subject heading.