Haha, I love how your emoticons pretty much sum things up.
Chiming in to second Kaplods' answer of trial and error. In general, women have lower calorie needs than men, older adults need less than younger adults, and less active people need fewer calories than more active ones. But that changes SO wildly based on your metabolism, your past diet history, your current state of health, the types of calories you eat, any food sensitivities you may have that those generalities have almost as many exceptions as rules.
One of the harder lessons to learn is how to take time to listen to your body--the only thing that can really answer the question, "how many calories?" It may take a week (or two or three or more) to find out if your initial calorie deficit is the one you should use.
Once you find out the calorie level you're comfortable with, you can't just leave it on cruise control, but will likely adjust it over time as you lose more weight or become more active. Are you currently keeping track of your calories? It can be really helpful to do that even before you've created any kind of calorie deficit so that you know where you currently stand. If your weight's stable at your current intake, you may not need to decrease your calories much to start the weight-loss ball rolling merrily downhill.
As for water and exercise, I consider those separate things from counting calories. It's good for me to drink plenty of water and to exercise, so I do it. All three are definitely linked physiologically, but as I can't effectively measure how many calories I burn off in exercise (don't believe the counts on pedometers and stationary bikes; they're no better than a guess), I don't look at exercise as a means to lose weight. Instead, it's a means to look better naked once I lose the weight.
Your mileage may vary, though. Some people like to "eat the workout" and consume back calories they burn while others raise their calorie count by a couple hundred on heavy-exercise days because all that working out makes them hungry. Those methods can work, too.
Trial and error is the key--as long as you remember that the "errors" don't give you a reason to stop, only to keep experimenting.
