OK, i feel there is something that i should do before i finally convert to civ4 
This is going to be a Sid game with completely random settings (except for the map size which will be standard). But i'll not play a start if it seems too easy, i.e. a start with a food bonus or an archipelago map with me statring on an isolated island.
I am going to play aggressively right from the start. There were quite a few Sid games done before where the human player expanded peacefully and traded/stole/bought/got via GL slingshot the techs necessary for space or diplo win (or artillery&factories for easy military victory over the AI). I don't want to play such a game because it is too uncertain. The neighbouring AI civ may decide to attack you and then you are toast. I want a win that does not depend on luck (or at least not that much). This means that i'll have to keep a decent army right from the start. Given the fact that the start will be without a food bonus this means that i'll have to go for an ultra-early archer rush. Here you may say that it also depends on luck greatly, but it is not true. Compared to expanding peacefully the human player has much more control over the situation and knowing AI you can almost certainly tell how many units are needed to guarantee capturing a town. In the beginning the AI keeps 2 spears in each town except maybe the capital. And a stack of 10 archers will capture such a town with almost 100% propability. Actually i would often use 7 or even 5 archers, but in this game i'll not take any chances, at least not in the very beginning.
I'll play the game by the GOTM rules, which means that i might use evil things avoided in SGs like RoP rape, breaking deals and some other dirty tricks. There is one thing that i'll not do though - I will not use the GL elevator which means that if i capture the GL city before education i have to raze or gift it on the same turn. This just seems too cheesy to me. On Sid it is very difficult to conquer new territory and in the same time keep up in tech: you can either expand and fall behind in tech or keep up, but without expanding. GL elevator simplifies things too much.

This is going to be a Sid game with completely random settings (except for the map size which will be standard). But i'll not play a start if it seems too easy, i.e. a start with a food bonus or an archipelago map with me statring on an isolated island.
I am going to play aggressively right from the start. There were quite a few Sid games done before where the human player expanded peacefully and traded/stole/bought/got via GL slingshot the techs necessary for space or diplo win (or artillery&factories for easy military victory over the AI). I don't want to play such a game because it is too uncertain. The neighbouring AI civ may decide to attack you and then you are toast. I want a win that does not depend on luck (or at least not that much). This means that i'll have to keep a decent army right from the start. Given the fact that the start will be without a food bonus this means that i'll have to go for an ultra-early archer rush. Here you may say that it also depends on luck greatly, but it is not true. Compared to expanding peacefully the human player has much more control over the situation and knowing AI you can almost certainly tell how many units are needed to guarantee capturing a town. In the beginning the AI keeps 2 spears in each town except maybe the capital. And a stack of 10 archers will capture such a town with almost 100% propability. Actually i would often use 7 or even 5 archers, but in this game i'll not take any chances, at least not in the very beginning.
I'll play the game by the GOTM rules, which means that i might use evil things avoided in SGs like RoP rape, breaking deals and some other dirty tricks. There is one thing that i'll not do though - I will not use the GL elevator which means that if i capture the GL city before education i have to raze or gift it on the same turn. This just seems too cheesy to me. On Sid it is very difficult to conquer new territory and in the same time keep up in tech: you can either expand and fall behind in tech or keep up, but without expanding. GL elevator simplifies things too much.