Let's actually bring that over here just to make it easy for the reader to see. It is very short, but it gets the major concepts across. We would need a library to go into detail.
Maybe this would be easier to give an example.
If you take a look at whales, you will see some of the pieces of evidence that lead towards evolution. A key thing to remember is that it is the whole of the data and not just any piece that is key.
Whales are mammals. They breathe air. They produce milk. They give birth to live offspring instead of laying eggs. During their development, whale embryoes have rear legs which disappear before birth. This is an example of ontogeny or developmental evidence for evolution. Sometimes the programmed cell death that should eliminate the legs fails to act and a whale will be born with rear legs. THis is another bit of evidence and are called atavisms.
Now, since the sea mammals have genes for making legs and most other mammals live on land, it was predicted that ancestors of whales should be found in the fossil record that are intermediate between whales and land dwelling animals. And this prediction came true. Some of these animals are Pakicetus, Rodhocetus, Dorudon, Ambulocetus, and Basilosaurus. So there is an example of the predictive power of evolution. It also opens up another line of evidence, the fossil record.
Now, once we found the fossil record for whales, we discovered that whales evolved from hooved animals. So we make another prediction. If you genetically test whales and various animals, you should find that whales are the closest to other hooved animals, specifically the even toed ungulates. When we did the actual genetic testing, whales did indeed test to be most closely related to animals such as pigs and camels. (Such a relationship would never be predicted by the "kinds" concept!) So this is another successful preddiction.
It is also an example of a couple of more kinds of evidence. The first is the relationships shown by geneticis. The second is extremely important. It is the twin nested heirarchy. This is how evolutionary trees produced by independent means point to the same conclusion.
You should see now how evolution makes successful predictions. Next let's talk about falsifiable.
I mentioned ontogeny and development. Falsification would come by finding development that does not make sense. Legs on a shark embryo, for example. Lactal nipples on a developing amphibian would be another.
Atavisms - This could be falsified by having animals born with atavistic feature that would not be expected. For example if a mammal were to be born with atavistic feathers, this could not be explained.
I'll stop here on falsification for succinctness. I can elaborate if needed.
The things such as I have mentioned are also tests for evolution. Beyond this, you can test and observe the mechanisms even if you cannot observe the long-term outcome in your life time. For example, we have observed many species developing new genes and new features through mutation. This and the other mechanisms can be observed and tested.