That said, I don't think I'm very good at it. I make a lot of rookie DM mistakes. So learn from my errors, and don't do the same to your players if you can avoid it.
I have discovered that I'm better at writing modules than campaigns. Most of my stories have a definite beginning and end. This showed a lot during the Praktas campaign, but was partly due to my gaming schedule. My group was full of people with busy lives, so we didn't get to play very often. If we'd had a continuous storyline, we'd have lost track of where we were and what we were doing. So it was just easier to make the sessions more module-like.
Besides, every time I tried to end a session halfway through the story, my players insisted we keep playing. Our longest session, "The Temple of Elemental Weevils", was meant to be divided across at least two sessions, but they wanted to see it through. And it's a good thing, since we didn't get to play again for four months. I'm afraid if the campaign was more freeform, we'd have had trouble finding a stopping place.
Now, there's nothing wrong with module-style play. Some people prefer it. But the problem is, it does mean I came pretty close to railroading my players. This is not entirely my fault - so far I've only DMed for newbies, who might not realize the sheer amount of freedom you have in this game. Every time I pointed them towards a dungeon, they were free to say, "Nah, I'd like to go explore the forest." But either the thought just didn't occur to them, or they knew full well that I put a lot of work into that dungeon. Plus, I think they preferred to stick to the "script" simply because they knew about my weakness when it comes to...
...Thinking On My Feet
I'm not very good at improv. When I put a session together, I like to write out every possible thing, in anticipation of everything the players might do. For the most part, that hasn't been difficult. Two of my players generally stuck to the "kill 'em and take their gold" philosophy, so I could usually predict what they were going to do. But they still managed to throw me for a loop a few times.
Example #1: In the "Pirate Queen" session, after the PCs defeated everyone on the pirate ship, they said, "Okay, we get her ready to sail and take her back to the mainland." Now why didn't I think of that? I handed my players an entire ship way too early in their careers. They could have sold it for an unbalancing amount of money for their level, or they could have used it as a mobile base of operations, or gone into the shipping business, or even become pirates themselves. Instead, I allowed the NPC Captain Jarran to tow the boat back to town, where the city claimed it and gave the PCs a reward for its capture.
What I wish I'd done: First off, Jarran's ship should have had a ship wizard like Robillard from Salvatore's books. While the PCs were exploring the pirate ship, I should have waited until they found the map, then told them that they notice the room was starting to fill with water. Having been badly damaged by the wizard's spells, the pirate ship was now sinking. The PCs would have had to grab what they could, then rush back to Jarran's ship before it sank.
Example #2: In "Small Sacrifices", Tirah was presented with a room she wanted to enter. The door was open, but she knew the floor tiles were enchanted to wake up the skeletal guards. She asked me, "Can I climb the walls?" I had to stop and think for a minute. I knew it was impossible. The walls were relatively smooth, and it would have impossible for her character to stick to them. But at the same time, it was the first D&D game she'd ever played, and I wanted to encourage her to think outside the box. So rather than just telling her, "No, the walls are too smooth," I wish I'd come up with some other suggestions for her. I wouldn't normally consider feeding a player ideas, but the session was mostly for practice anyway.
Example #3: In "The Princess and the Dragon" session, after Delmer revealed what he'd done with the Princess, the group demanded he accompany them to the dragon's den. I should have expected that one. I handled it okay, but I still wish I'd thought of it in advance. I could have had Delmer try to escape during the trip. Or maybe even had him try something really nasty, like lead them somewhere besides the dragon's den. Maybe he's less afraid of Trolls than dragons, or maybe he has some sort of allies in the mountains somewhere that would help him out. Hey, maybe that's how he got the Princess up the mountain in the first place, despite it being populated by hostile creatures. Just like Doctor Smith from Lost In Space, he could have really thrown a wrench in their plans.
Let The Dice Fall Where They May
I'm too nice. I just can't see myself killing off my PCs. In the very first session I ran, "Kobold Hall" from the DM book, there is a dragon's lair below the room with the kobold boss. We were running with just two level 1 characters... there's no way they should have been able to survive that encounter. I gave them every opportunity to flee, but they wanted to try it. And they beat it, barely. Mainly because I didn't have the dragon use any of his more vicious powers. Those two characters still died in the next session - twice - but if they'd wanted the characters to live the second time, I probably would have found a way to make it work.
As mentioned above, I don't always remember to use my special moves. I tend to forget to have the baddies work together, and I most of my monsters act like brutes: charge the nearest player, then make attack rolls until you die. There's so many interesting attacks out there, and battles can be more than simply counting hit points as you trade blows.
Balancing the Economy
Early on I designed Trasa's Bounty Hunter Guild. I made it sort of expensive to join, because I felt it was the sort of thing that should wait until higher levels. Then I got impatient. Rather than lower the price of admission, I raised the amount of gold I gave my players. Well, once they were in the BH guild, I couldn't just go back to giving them low rewards per bounty, or else it looks like joining the BH guild was a bad deal. I can see it now: "5000 to join, but I only get 50 per bounty? It'll be years before I break even!"
So, at level 5, my players were decked out in the best equipment they could get at their level (I did at least enforce restrictions on what they could buy, thank goodness), and they generally walked around with about 4000 gp spending money. So of course this meant that finding treasure in dungeons meant less. "Eh, I've already got some pretty good gloves. And boots. And necklace. And helmet..."
Lest you think I'm totally being unfair to myself, I admit I am learning. I am getting better at hiding my confusion when my players do something unexpected. If I ever run another campaign I'm going treat it more like LFR, with more structured XP and rewards. The Praktas campaign was good for me, but there is a reason I called it "Praktas" - it was always meant to be a learning experience for both the players and myself.