... Or the guy was just like so many average humans and assumed that the machine was doing something more than randomly picking cards to hold. If the machine s

... Or the guy was just like so many average humans and assumed that the machine was doing something more than randomly picking cards to hold. If the machine said it was auto-picking the worst cards, we'd all be mad. If the machine said it was randomly picking the cards, we'd all be surprised: why even do that? Who does it help? Therefore, from an implicatures pov, the only reason for the machine to pick cards (or "suggest" cards) is to recommend a good hand. Since the computer can often pick the best hand, why would we not assume that it's doing so, to give us mere humans "a fighting chance", so to speak? Why should we assume that it's not picking the computably best cards to hold, other than our belief that the machine "cheats"? And that's what the article examines. I don't think it's ignorance on our gambler's part; I think it's the standard "Black Hat UI" issue mentioned on HN so often: We assume that sw does somethign to maximize utility for the user, based on the design or experience, when it's instead doing something to maximize profit for some other entity. Yes, it's a gambling machine; we all understand that it's designed to take our money. But offering a feature as part of the experience that appears to help when it doesn't (or to a lesser degree than would be expected via presentation), would be deemed misleading by any reasonable person, small print and disclaimers aside.