I normally hate it when people put their names as part of the title of deck posts, but there are so many of these decks that I figured doing so would make it easier to find/distinguish.
I made it because I saw slight faults in most other decks of this type. I tend to get decks as close to efficient as possible, so I thought I'd give my version of the ideal "basic" form of this deck.
Hover over cards for details, click for permalink
6rn 6rn 7jp 7jp 7jp 7jp 7jp 7jp 7jq 7jq 7jq 7k3 7k3 7k4 7k4 7k4 808 808 808 808 808 808 808 808 80i 80i 80i 80i 80i 80i 8pu
The amount of towers seems to be nearly perfect. With the aether mark, you only need two or three to use fractals often enough, and with eight towers, you usually start with two to three. If you tweak this, I recommend staying between seven and nine towers
Rays of Light and Fractals are the key cards, so there are six of both. If you tweak these, I recommend a minimum of five for both.
The light dragons are typically only necessary late game, and a single fractal usually means you'll win in two to three turns. Again, I recommend staying within a range of one from the amount of cards in this deck. You can also use golden dragons as well. They would probably be better, to tell the truth, but I wasn't paying attention while I was upgrading the cards. It still works great with light dragons though.
Having three hopes means you have a pretty good chance of getting one early. If you're worried about it though, go ahead and add one. I don't recommend taking one away though, as you usually want one out by turn four or five.
The luciferaces are really only for late game use. If your opponent has lots of creature control, giving light production to a couple dragons can be of use. They really aren't necessary though, so you can remove them if you'd like. I advise keeping them in if you add a miracle to the deck.
The SoG's ensure that early game pressure doesn't hurt you. You can usually get out a hope before a speed deck can kill you, but these are just in case. I'd say three is all you'd need at most.
If you want to do some things differently, a miracle or two certainly wouldn't hurt. Taking several upgraded sundials would also be useful (it doesn't matter how many, but if you want them, they would probably be most efficient if you took more than three). Since they let you draw a card, it's basically like not having them at all. If you want creature control, two or three lightnings would do the trick and help against growth decks and some speed decks. You don't need many because you'd hardly ever need them in the first place.
Whatever changes you decide to make, I advise keeping the deck at a maximum of thirty-three cards. For the most part you can use anything I suggested and keep the deck just or nearly as efficient. Some changes may make it do a little better while others may make it do a little worse. It also depends on what decks you are fighting though, so don't abandon your changes until you've tested several games.
I've seen some decks with electrocuters in them. You really don't need those, since you want your quanta for fractals. But if you disagree, I can't stop you.
I played a lot of games against AI3 and don't recall losing any. I'm not one to keep stats though, plus I'm sick, so I'd rather not be bothered with such trifles.
Enjoy, criticize, make babies.