To start with, I'd like to boil down the idea of religion. Religion comes in many different forms and types; most commonly it is a set of beliefs about a deity, a supernatural being of some sort. By some definitions, everyone has a religion, so I will restrict it for now to just those who believe in an all powerful deity and received truths such as ancient texts. Almost every religion professes to be the truth, or at least a truth. First I will consider religions that profess to be the only truth. This would contain most religious people, after all, the main Abrahamic religions tend to fall under this category.
Faith
At the basis of many religions there is faith. The Oxford English Dictionary states that faith is:1. Belief, trust, confidence.
a. Confidence, reliance, trust (in the ability, goodness, etc., of a person; in the efficacy or worth of a thing; or in the truth of a statement or doctrine).
So faith is an acceptance of a truth, acknowledging that something is correct as far as we know and usually the implication is therefore to base ones actions upon it. It's the axioms you base your logic upon. For example I have faith that my mobile phone can send messages hundreds of miles reliably. I derive this partially from my faith in my senses to be roughly true. As a result I use it to communicate with my friends on a regular basis, and trust that what comes up on the screen is actually my friends who I meet in the flesh want to say. More basic is truths I find it hard not to accept, like that I exist and the world I see around me follows certain physical laws. I base this on observation, and what I have learnt fitting with observation. My observation, in turn are based on my faith in my senses, after all, what else can I believe in? I believe in gravity because I need to describe how things stay on the ground. A certain amount of faith combined with strong evidence (and lack of it to the contrary) allows me to hold the faith that gravity does work, that every mass near the surface of the earth is pulled towards it at an acceleration of 9.81 metres per second squared (unless acted upon by a non-negligible force).
It would not be unreasonable to say, by the scientific method that my beliefs are based on how reliably it fits the world I see around me.
I think we can agree that the scientific method is the best way to make a decision. Logically, it is the best way we can make predictions about future sensory input from current sensory input. I.e. the scientific method is the best way to evaluate and understand your reality. Very quickly, what is it? Make a theory, develop a prediction (hypothesis) from the theory, test that prediction. If it fits, try more predictions, test it forever (i.e. continually adjust to new evidence). If it doesn't, amend your theory and try again.
So up to now, religion is fine, lets say it's a theory. It's a statement about reality that can be used to make predictions. Let's evaluate it as a theory, what makes something a good/useful theory?
Reliability
For this subsection, lets say we come across several theories, each of which is a religion or belief set. The test of whether or not a theory is useful is its reliability: how often the predictions it makes fit with reality. Tied into this is how sure the theory was of the prediction (what margin of error did it concede) and how precisely it was defined. If a theory consistently produces correct results, it is reliable.Simplicity
There's something great about simplicity. As a mathematician it is what makes me tick. As Hardy said “A mathematician, like a painter or poet, is a maker of patterns. If his patterns are more permanent than theirs, it is because they are made with ideas.” The simplicity of the groups axioms which create amazing structures and interesting results when combined with intuitive constructions. Or the very basis, the Peano axioms, which alone can hold up the vastness of number theory. But that is just me rambling, more tangibly (and built on maths) are Newtons laws. Summed up with a simple F=ma, along with the law of gravitation which takes just the two masses and the distance between them, can explain the intricate and complex behaviour of our solar system to an amazing degree of accuracy. Instead of this simplicity, we could say that all celestial bodies are drawn to each other by magic pixies, and we are held on the earth by invisible ropes. Now you could say that's almost easier to understand, but it makes so many assumptions (the pixies have always been there, they are ultimately clever, they are invisible, impossible to interact with, they are reliable with mathematical accuracy). And these just aren't required, we could forget all about these sentient pixies, and have one assumption: one force gravity, interacts with mass with this relationship. Its the principle of Ockham's razor: "The [hypothesis] that makes the fewest assumptions should be selected." Especially if the two theories produce the same results, as assumptions should be cut down to a minimum.
And there's another problem with religion, it invokes a God: a sentient, complicated uncreated creator. When all that's needed is one law: things came into beginning at the start of time (as we can tell). This gets rid of messy questions and needless assumptions. We don't have to look far to find awkward questions because of the existence of God, it on the tip of everyone's tongue: Why won't God forgive everyone? Why doesn't God stop the pain? In the assumption department we have more problems. If we have this God who is a bit like us but more powerful, we have to put him in his own universe with its own rules (more assumptions). How does God interact with this reality we know and love? (more assumptions and laws)
It may be commonly argued that science is constantly changing and therefore cannot be claimed to be the truth. This is partially true, science attempts to constantly improve; better, more succinct and accurate theories replace old ones. To avoid the continuum fallacy, you should think of science getting closer and closer to the correct answer. If we get closer, this doesn't mean we were far away before. The strength of a theory is not based one whether or not a better one is produced, but how strong the evidence actually is. So it isn't correct to use theories which are more complex and less accurate. We use the best one we have at the time.
What's the point in a theory?
A theory gives us a framework for looking at the world, it can make predictions and statements about the reality we inhabit. To use gravity as an example, it allows us to understand nearly all of rocket science; space exploration is based on the ideas of Newton, it can for example accurately predict how much fuel is required to leave earth. To give you some perspective, NASA managed to get to the moon and back without considering quantum or relativistic effects, just plain old Newton. Gravity and the rest of Newton's laws provide us with definite, quantifiable predictions which are tested time and time again to be shown to be true (to a high degree of accuracy for low speeds up to around 1% of the speed of light).
This brings us to the second case, religions which do not claim to be only or total truth. These tell us nothing for sure, and are useless for prediction. A theory which states x is true sometimes, with no qualification of when is near useless. Again, like all unfalsifiable claims, unless it has contradictions it cannot be false but we can say it is useless. And if a religion is somewhere in between the two, the set can be split into parts and critiqued separately. Either a specific Religious claim can be falsified, or it is pointless in this reality to use as a theory.
The point of a theory is to make definite, well defined statements about reality. Usually this is with a view to use it for prediction, which also entails constant testing.
Now, I don't want to offend religious people by putting it on the same pedestal as horoscopes, it was an extreme example to demonstrate the idea. Religion is a much better developed theory than astrology. Most religions have a/some powerful god/s (compared to us), we can consider two distinct cases from this. The god/s do interact with the physical world or they don't. Firstly, if the gods did interact with things we could see and detect the interaction, I don't know of any conclusive studies which show such predictions from holy books being detected but I could be wrong. However, in this case, at least the theory is potentially useful. If the god doesn't interact then the theory is useless as well as completely unprovable, it would make no useful predictions or explain any phenomena and logically should be eliminated.
Well Defined
This brings me to my final point, the strength of a theory is also based on it's certainty and how interpretable it is. For example to say "something is going to happen today" is correct but vague in many ways. You must have precisely definable terms and not make general statements. A classic example is horoscopes, they nearly always make correct statement without actually telling you anything useful. One I randomly found on the internet:"Friends born under the Fixed signs of Taurus, Leo, Scorpio and Aquarius could offer the best kind of support over the next couple of days. You might need to access your own ability at detachment. A degree of confusion about a deadline is likely. Financially, however, all is likely to be clear - but not particularly pleasant. These issues may need to be addressed sooner rather than later."
Take the two highlighted phrases, they are near contradictory and actually describe the majority of financial situations we find ourselves in. Furthermore, there is no surety about these, there is no quantifiable way of measuring how right they were. The reading of horoscopes is therefore a poor theory, even discounted having no logical basis, they are pragmatically useless as a theory of how the universe works.
Similarly vague statements have little predictive power, particularly when they can be interpreted in so many ways. Being certain about something vague is analogous to being uncertain about something definite.
For example, saying
"Your house and your kingdom will endure forever before me; your throne will be established forever." 2 Samuel 7:16 (NIV)
is incredibly vague. At one point, this would have clearly referred to the physical line of David, however years later Paul was able to bend it to refer to the non-physical line including Jesus. This happens often today, if we define things in terms of some other more definite things, we could then make useful predictions. But we can't, every denomination disagrees. The more realistic denominations provide definitions so useless that it is a tautology or simply references something equally as badly defined.Thanks to the law of large numbers, someone is likely to make an interpretation that is correct, but this is no thanks to the poor theory. In this case its predictions are useful for only a small number of people, and only by chance.
For comparison, consider f=ma, where each term is defined relative to some exact object (for example a metre ruler) and their relationship mathematically. Newton's laws leave little room for interpretation and post-hoc reasoning. Admittedly, there is some amount of interpretability with everything, but it is clear to see that some things allow more than others. Almost everyone around the world with A level equivalent physics knowledge knows how to interpret f=ma. It's accuracy is high for a large percentage of people, it is a useful theory.
Summary
We use theories all the time, some of them we test and some of them we don't. The scientific method (almost by definition) is the best way to decide, whether or not we should a certain theory. Religion constitutes a terrible theory, that, when scrutinised offers very few helpful and accurate predictions.In summary, you can believe in an all powerful God who made everything (it is possible and it is a theory), however, it is not only more complex and less accurate, but riddled with inconsistencies. Furthermore, it is useless for solid predictions as it is open to so many wildly different interpretations. We can explain the universe accurately, reliably and in a well defined sense without a God.
"it is superfluous to suppose that what can be accounted for by a few principles has been produced by many" Thomas Aquinas (Ironic I know, I'll come back to this)
PS: I apologise that this article is somewhat confused, it has been written over a long period of time and has become a little unwieldy and haphazard.