The English language was in period of rapid change when printing became popular. A lot of regional and historical variations became "frozen" in print. The explanation of crazy spellings, like "night," is "that's really how the word was pronounced when they began printing it." There are no reliable pronunciation rules. Our spelling is not phonetic. The spelling of a word doesn't tell us how to pronounce a word. The spelling can be a hint. It helps us remember the pronunciation. The "a" in "angel" has the "long a" sound, the same vowel sound as "brain" or "pain" or "cane" or "Jane." But in the adjectival form, "angelic," it has the "short a" sound, the same as the "a" in "animal" or "can"--or the word "an." This makes no logical sense at all. There is no reason why adding the -ic should change the pronunciation. The "a" also has the "short a" sound in the name of the big city, Los Angeles. And it also has the "short a" sound in the word "angle." When we see a word we haven't seen before, sometimes we guess, based on it resembling a word we know "strange" or "danger." Sometimes we guess right, sometimes we guess wrong. I think every native English speaker is embarrassed from time to time by mispronouncing words they have seen often in print, but have never heard said aloud. We guessed the pronunciation without checking a dictionary, and guessed wrong. We use dictionaries for pronunciation, not just for word meanings.

When you have vowel consonant vowel (A N I mal), the first vowel is usually short. When you have vowel consonant consonant (A N G el), the first vowel is usually long. Of course, there are a ton of exceptions. English just isn't a phonetic language.

Welcome to the English language and its vast array of irregularities! Perhaps there are historians of the language who can say for sure. My theory: "ange" in "angel" is like "range" "strange" "danger" Notice that the "g" in these words is /dg/ which may force the "a" to be long /ei/ Compare with "anger, "banger" "animal", where the "a" is short /ae/ Possibly there is a pattern related to the etymological origin of the words. The examples I gave like "angel" are perhaps from Latin via French but the other group are not. I have no idea if that theory is reliable!