Video game industry practicesVideo game industry practices are similar to those of other entertainment industries. There is a history of two things: treating its development talent poorly, and promoting independent development. The first leads to the second, and, amazingly enough, it repeats cyclically as the new company grows and becomes more like the old company in its business practices.
The most famous case of this is probably that of the "original" independent developer, Activision, which was made up of a small group of disgruntled Atari developers. Activision succeeded and grew, and is now the world's second largest game publisher. The people that started it, on the other hand, have long since faded into the shadows, either to do other work or to start over again with another new company. For example, one of the original founders of Activision, Alan Miller, left Activision to found game developer Accolade (now Atari née Infogrames).
The problems developers faced then were somewhat different from those of today. The Activision group, for example, made it a priority to let the developers get credit in the packaging and title screens for their work, which Atari had totally disallowed. Through the early-to-mid 80s, many developers faced the more distressing problem of working with fly-by-night or unscrupulous publishers that would either fold unexpectedly or run off with the game profits entirely.
The largest problem today is still economic and has to do with the publisher-developer contracts. For their work, developers usually receive up to around 20% in royalties, while the rest goes to the publisher. Alternately, studios are bought and absorbed into the holdings of the publisher, where very often, after having lost their original leadership around the time of the buyout, they stumble along, falter and are closed after a few years. In both cases, the result is one-sided. This vicious cycle has two other negative consequences: A big push for projects to finish in time to make the holiday purchasing window, and creative control going to the publisher.
Making ship dates results in one of the most stressful aspects of development: Crunch mode. Software development in general is prone to having heavy loads of overtime work at the end of projects, when deadlines draw near and all the loose ends need tying; games are no different in that regard, and have the additional burden of being tied to game content (art, music, levels, scripting, etc.). The resulting weeks (or months) of chaos take their toll on the whole team, and probably are one of the leading reasons why people quit.
Creative control is the more idealistic angle of this. Not surprisingly, people who make games want to have control over the types of games they're making. But since the publisher has the ultimate control over what gets on the shelf, it's all about what the publisher thinks is best to make. Publishers being large, profit-seeking entities, are predisposed to follow whatever market trends they see. The upshot: if one game is a megahit, a series of passionless clones will shuffle along, following its footsteps.
One promising effort at breaking this cycle come from the smallest group, interestingly enough: small companies of one person or a handful of people that develop and self-publish using the shareware model, over the Internet.