Since none of us lived in the days of the dinosaur, scientists are trying to put together a puzzle with not just one but MANY pieces missing. Evolution - as well as creation - is a belief system. Choose which one you want to believe.It seems a valid enough point. We don't have all the evidence, so how can we be sure its one way or the other? Are the missing pieces damning to evolution, and is supporting either just a matter of belief?
An emerging pattern
To once again use a detective analogy, imagine we are a detective at the scene of a crime. Before coming to the crime scene, we know nothing of the situation beyond there's been a murder. Upon arriving at the crime scene, a few pieces of evidence are gathered. There's a bit of blood on the victim that is not her own, marks on her neck and rashes on her ankles. A bullet is lodged in the wall. Cocaine on the ground.
Further digging around reveals relationships the young woman was involved in, people she was friends with and activities she was involved in. It turned out that there were a few phone numbers listed on her phone for known drug dealers. From there, a check with phone companies and subsequent interviews, it places one man with the motive and means to kill her.
In this completely hypothetical situation, the point is not to show that yes that man killed her, rather that patterns can emerge from just a few pieces of information. And as the investigation goes on, more and more information fits the same pattern.
Would one conclude that it's equally likely that this man killed her as her dying through natural causes? This sounds absurd for a good reason, but it's the gambit that steve made above. The police don't have the exact information showing every moment of the man's actions during the day, let alone when the murder took place. Yet the evidence points so strongly towards that not only was she murdered but she was murdered by him.
The fact of the matter is that we don't need every piece of evidence before one can say anything. Just by looking at a few key pieces of evidence a pattern will emerge. The evidence has to of course be relevant, there's no point in interviewing everyone on the planet just to find the murderer. So how does this apply to whether evolution happened?
Making key predictions
Like the detective arriving at the crime scene, we find ourselves on this planet after the fact. We have several facts about the state of life today, such as the geographic distribution of animals, anatomical features, life-cycles, genetic code, etc. There are also observations of the life-cycle through time, of artificial selection, of heredity of traits, of mutations, etc. Then there's also relics of aeons past, the fossil record being the best example among others such as vestigial traits and dead genes.
Through observations independent of the truth of evolution, the age of the planet is determined to be around 4.6 billion years old, with the first life appearing somewhere around 3.5 billion years ago. It's important to understand that the time frames are independent of the fact of evolution, that geologists, astronomers and nuclear physicists will still confirm the earth to being old even if life is separately created.
Evolution is a predictive hypothesis, it predicts that certain patterns should be seen. For instance, apes should have DNA closer to each other than they do with monkeys. And this is exactly what is found, humans and chimpanzees are more closely related to each other than each animal is with any other species. The DNA shows a tree of life that what can be derived from cladistics and even the fossil record.
But it's the fossil record that's under contention here. And the fossil record is incomplete, that is to say only a fraction of all the species that have ever lived have been fossilised, and only a fraction of those have been found. Yet if evolution were true, then there should be particular kinds of fossils found that fit the pattern.
Not only is there a general fitting of the pattern in that there are no mammals before fish or reptiles before amphibians, or even humans and tyrannosaurs in the same strata. The general pattern of evolutionary radiance does indeed fit the fossil record, but it goes further than that. Fossils showing key evolutionary sequences and the emergence of novel traits litter the fossil record.
That now there's a sequence of fossils showing the adaptation from land mammal to cetaceans. Archaeopteryx is well known to show the link between dinosaurs and birds, but in recent years many non-avian dinosaurs have found to be feathered as well as many more fossils of birds with saurian features. A plethora of fossils showing hominid evolution over the last 7 million years have been found, showing clearly a progression through time.
The best example of the predictive power of evolution is the story of the fishibian Tiktaalik. There are no amphibians before 365MYA in the fossil record, and lobe finned fish with amphibian characteristics around 385MYA. So by looking in rocks ~375MYA, and there was found a fossil exhibiting the very traits that were predicted. Nothing speaks better for the validity of evolution than predicting the kind of fossil to find at an exact period of time.
The creation model
There are two problems for creationism in regard to the fossil record. If all species were created at the same time, then why is there a pattern showing change through time in the geological record. It must be that if creation were to be accepted then only particular models of creation could possibly be correct. Young earth creation is ruled out by means other than by evolution, geology shows that life has been on this earth for over 3 billion years. Creation would be restricted to species coming into being in the pattern that evolution predicts.
The second problem is the problem of redundancy. That sure, a creator could make feathered dinosaurs, at each point in the way making a slightly more complex and refined feather. Then the creator could create feathered dinobirds that show primitive flight. Then fully fledged flying birds that have since lost many saurian characteristics. It's arguing a point of redundancy for the creator, that each slight successive change in the fossil record is a new and unique creation.
That any fossils showing change at all is evidence of evolution and a problem for creation. That the sequence showing the evolution of early mammals from synapsids shouldn't exist if mammals were created ex nihilo. That there are fishibians, dinobirds, land whales, few legged snakes - just to name a few of the vast numbers of transitional forms. The only pattern that even remotely explains these forms is evolution, creation doesn't cut it. Missing pieces be damned, there are enough of the right pieces to show the shape of the puzzle.