I'm normally a very good GM, but on a couple of my recent sessions I really dropped the ball. Sometimes it happens to even the best of us.
Basically, I wrote a set of Zombie RPG rules, and they didn't work. I still think they would work reasonably well for what they're intended to do, but that wasn't what my players were interested in. They wanted to get the heck out of dodge, whereas my rules naively assumed they'd stay in densely populated areas that would turn into an undead wasteland over time. I had chapters devoted to sieges and supply hunts, two scenarios my players did everything in their power to avoid. (And seriously, can you blame them?) They stocked up early before things got really bad, and headed out of town.
As it turns out, there's no such thing as a supply shortage during a zombie outbreak. Supply shortages like you see after a hurricane or other disaster are a function of a break down in supply lines without a corresponding significant decrease in consumption. Unless your zombies are eating canned goods, there's no shortages. Though malnutrition would eventually play a factor with a diet so restricted, any of us could survive for a very long time on a couple of cans of beans a day with no immediate ill effects. Next time you're in a grocery store, attempt to count just the canned beans in one aisle of the store. Thousands of cans. Either survivors are rare and thus any given grocery store is several years of food for the party, or the infection rate is low and the PCs can readily avoid the zombies. Neither results in particularly tense and exciting sessions, and while everyone else seemed to be having fun, I felt my pacing was dreadful.
There was a point where I probably should have stymied their efforts to leave town, for the sake of the narrative. If I'd kept them in town, where the zombies were, I could have kept the game sharp. However, the rules I was using were intentionally very light, and assumed a strong level of PC competency, so there wasn't a particularly good way to provide an obstacle to the players that they wouldn't have a very high likelihood of getting past... at least not that early in the timeline. And since I wrote the rules, I didn't want to just hand-wave the obstacle, lest the it be perceived that my rules were intentionally light specifically to allow for unfair GM fiat.
Some time after the PCs got out of town, after many more sessions of pacing that I felt was wretched, I essentially decided to end the campaign... but I didn't decide to do so in a reasonable way, or at a decent time. I just kinda got frustrated during the final scene of a random session where they'd once again dodged a particularly slow-moving offscreen bullet. In my frustration, I had an NPC go psycho and sabotage their car. No build-up, no foreshadowing, not even a die-roll to blame it on, just pretty much "so, that psycho guy cut your break line and now you're hurtling down the hill."
I told myself I wasn't ending the campaign, I was just creating a good cliff-hanger and making the game exciting for once. What a crock. It was abrupt and unfair, and I'm lucky I didn't lose my GMing badge over it.
My players looked at the situation, and politely said "well, I think that's about enough of this campaign." And what could I do but agree. Your GM's a dick, and a hack. The very sense of engineered unfairness that I'd carefully steered clear of 5 or 6 sessions earlier, now returned with a vengeance.
I hope I haven't burned any bridges or good will there, because that bizarre turn of events was very abnormal for me. I like a good tragic ending, but this wasn't a good one, and it wasn't planned or prepared for in any way. Nobody left angry (that I'm aware of), but it was a rather pathetic thing for me to do as GM. Sorry I let you down, folks.