One armed bandit

The central question in reinforcement theory is whether animals allocate their responses optimally between the two activities, that is, whether they act so as to maximize their rewards. It turns out that the answer depends on the reward mechanisms associated with the two activities. If both reward mechanisms are of the one-armed-bandit type, animals often do the rational thing and allocate all their attention to the activity with the highest probability of reward. Sometimes, however, they commit the "gambler s fallacy" of distributing the stakes in proportion to the odds. If both mechanisms are of the other kind, findings are also ambiguous. With one mechanism of each kind, as in Fig. IX. 1, animals usually do not optimize. 

The animal can allocate between 0 and 100 percent of its time to the two activities. One is rewarded by a VR (variable-ratio) mechanism this is the one-armed bandit. The other is rewarded by a VI (variable-interval) mechanism this is the mechanism... 

Reno is almost proud to play second fiddle to its excessive cousin to the south. By not re-creating itself every decade with brand new hotels, Reno has been able to hold onto its title as the biggest little city in the world. With Yosemite National Park and over a dozen nearby ski resorts, there s more to do in Reno than gamble. Residents spend more of their time mountain biking and skiing then yanking on the one-armed-bandit. 

---