Video slot machines pull a lot of tricks to make it hard to tell how fair the game is; one of them is to ring up "wins" that are actually losses (you put in $1 and get $0.75 back, say), with a lot of fanfare and hoo-rah. These tricks are calculated to hook players into the game by stimulating their reward centers with intermittent stimulus, a powerfully addictive combination.
Researchers from the University of Waterloo's Gambling Research Lab created a series of short educational animations explaining how the scam worked, and showed them to inexperienced gamblers before exposing them to slot machines. The players in the experimental condition "not only remembered their actual number of wins more correctly, but they were also more capable of labelling losses disguised as wins during slot machine play."
In other words: the subjects who experienced false wins didn't feel like they'd won, and didn't experience the dopamine release that these false wins triggered in gamblers who hadn't seen the video.
Research has shown that LDWs are somatically, behaviourally and psychologically miscategorized as wins rather than the monetary losses that they truly are. What is not yet known is how LDWs are processed at a neural level. Gambling activates the limbic system and dopaminergic reward centres of the brain. These centres have been implicated for their roles in addictive behaviours. Interestingly, intracranial recording in monkeys have shown that phasic activation in dopamine (DA) neurons correspond to the predictability of reward outcome. Specifically, activation is greatest when cue signals predict that rewards will occur on 50% of outcomes (i.e. chance, or maximal unpredictability). Zack and Poulos remarked that spinning the reels on a slot machine, apart from their outcomes, could thus profoundly activate this DA system. Given that researchers previously showed that non-zero outcomes (i.e. wins and LDWs) can occur on close to 50% of spins on commercially available slot machines, LDWs could potentially activate the DA system if one miscategorizes these outcomes as wins rather than losses.
At the purely behavioural level one could speculate that LDWs could encourage prolonged slot machine play despite financial loss. Young, Wohl, Matheson, Baumann, and Anisman used extinction paradigms to show that players gamble for significantly longer during a losing streak if they previously played a game with many small interspersed wins (a high reinforcement rate). If LDWs cause players to (mis) perceive the reinforcement rate as high (even if the actual wins are relatively rare), one might predict similar effects. Studies in our lab are currently investigating this question.
Losses disguised as wins in multiline slots: using an educational animation to reduce erroneous win overestimates [Candice Graydon, Mike J. Dixon, Kevin A. Harrigan, Jonathan A. Fugelsang & Michelle Jarick/International Gambling Studies]