One of my main difficulties is...

[blazer-glory]

knowing when or even IF to attack my neighbors. How should a time when to attack my nearest rival. Too often I get embroiled in a drawn out war that completely distracts from the rest of the game and becomes the main focus. In the mean time the other civs are advancing their technologies or slowly conquering their corner of the world and eventually turning their attention to my battle worn cities which they capture very easily.

 

[troytheface]

if you do not get immediate success and make peace that civ will attack you later. it is a delima indeed.

 

[Iranon]

Civ4 is reasonably complex; any major decision should have a reason behind it.

If you want to war to gain an economic advantage (through the conquered cities) you want them short and decisive. This usually only works if you have an overwhelming advantage at least locally or when you can use the Apostolic Palace/UN to end the war on your terms.
Drawn-out wars will just create happiness problems for little gain.

At other times, you wish to cripple an opponent. Raze a few cities, pillage the hell out of them, steal their workers. If they relied on cottages, their ecnonmies might take centuries to recover. If you prepeared defensive positions where the AI will burn wave after wave of units, the war will cost you nothing while the opposing civ is kissing their progress goodbye.
If they are left weak, you raid their lands each time they are about to recover as a source of income and slave labour, without the maintenance and vulnerability issues of a large empire.

Full conquest is usually costly and will negatively effect your economy for some time - you are fighting in their territory, and you'll also be tackling fortified troops in walled cities on a hill. The AI is not particularly good at warfare, so you will often be able to conquer a superior civilisation if you put your mind to it, and the additional land will allow you to catch up to and surpass the peacemongers.
If it doesn't work quickly, however, you're stuck with a bunch of vulnerable, unhappy cities that revolt constantly and do very little apart from costing maintenance, draining your economy for a long time even after the war is over. Cripple them, make peace, rearm, return in 10 rounds.

 

[blazer-glory]

Thanks for the advice.

 

[OctavianFlu]

Good advice for sure; however I feel your underlying question is still not answered. When to attack to prevent a dragged out war? Check the power graph regularly. This will give you an indication if the potential war victim has their military might in consideration or not, and when you feel you have a powerful advantage - use it, for either of the reasons Iranon provided.

Also -- Alphabet, if you are running beyond the sword, gives you the ability for spies, or if you are using vanila or warlords, you can ask for open borders, and scout their land. When you feel that you can take on their defenses then go for it.

Or even better yet, ask another civilization to go to war with you. Sharing the same military struggle can bring you closure to your friends, but knowing when you have the advantage through either power, or technology should be the only time you should attack (unless there is another reason (key resource taken like oil via culture)) in order to secure a timely victory -- imho

 

[blazer-glory]

Actually another major issue I have is knowing what techs to research. Obviously I want everything, archers, gunpowder, strong navy etc but what should guide my decisions in what to research?

 

[Morgrad]

Another important part in making a war short - plan your route!

Know what city(ies) you're going to take in the first 1-3 turns of the war, where you're going to resupply, and what citied you're going to hit turns 4-10. If you have it mapped in your head first, you should be able to wage an effective 10-turn war that will win you the objectives you need won.

Once you have planned your 10-turn war (and keeping in mind that no plan survives contact with the enemy), you can decide if you're going to plan further;

"In 6 more turns I can take the rest of the cities on this landmass"
or
"In 3 more turns I can raze some border cities to ease cultural pressure"
or
"I need 12 more turns to completely wipe out that civ"

Check your options, plan for them, execute them.

As a general rule of thumb (very, very general), I want my stacks to contain 10 modern units + 6-8 modern or one-step back seige + ideally a Medic III from a GG. So if my battle plan is to take one city, resupply and take one more in 10 turns, one stack like that is fine - I would consider that my typical "axes, swords, and catapults" attack. But if I'm going to hit three cities on a broad front, I want 3 stacks just like it - which is obviously a lot more buildup.

All of that is, of course, entirely situational. If I'm hitting musketmen with tanks, the stacks can be much smaller. If I'm hitting Mechanized Infantry with Tanks, the stacks (esp. the seige) needs to be much, much larger.