1. Reviewers are God. The reviewers play God when they determine a paper's fate. But, some reviewers are less capable than others and make mistakes. Some reviewers are biased towards one trend of solution over another, and often project their bias. Most importantly, busy academics get their review done by their graduate students, many of whom may have a substandard expertise. The bottom line is, the reviews are often completely beside the point. Good conferences have 3/4 reviewers, but often one strong negative review (a veto) is more instrumental in rejecting a paper.
The problem is that paper writers have no way to defend their work. Luckily, some people have already acknowledged the flaw and have been trying to change things. Last year, my friend Shamsi submitted a paper to CHI (the most important HCI conference) only to get her paper rejected. Luckily, CHI had a rebuttal mechanism, where the author can appeal to the committee. She did that and got her paper accepted. Other conferences do not employ that because people think that it is too much work. CHI would be a good counter-example. It is a very big ACM conference (perhaps among the top 5 most attended conferences of ACM) and yet they can do it. The statistics suggest that only a few authors actually rebut (but those who rebut have a good reason for doing so). So, the organizers of other conferences should think about it.
2. Out of sight, out of mind. Researchers collaborate, but often fail to pay tributes to the collaborators. My wife Farhana did her undergraduate thesis on admission controller of MPLS networks in Bangladesh University of Engineering and Technology. Today she found out that her adviser published a paper on her work, a paper that she helped writing before coming to the US, but "forgot" to include her name in the author list.
Foul plays like this happen in academia all the time. There is no place where a victim can petition and claim due credits. Perhaps you would not want to include your name in a paper with other authors who chose not to mention your name in the first place. But in a competitive academic environment, where publication count is vital, you soon get past the disdain and want your fair share.
Note. For other flaws in the paper publication system, read
Parnas, D. L. 2007. Stop the numbers game. Commun. ACM 50, 11 (Nov. 2007), 19-21. DOI= http://doi.acm.org/10.1145/1297797.1297815