In creating an ancestral lang for a hard-fantasy world that is based on the language of your home, you first have to understand that language. And for an impatient worldbuilder like me, the Vulgar Language Generator seems the best place to go. Click the "English" choice, and you automatically get the following consonant clusters:
Word initial: b bɹ d dj ʤ f fj fɹ h j k kj kw kɹ l m n nj p pl pɹ s sm sp st t tɹ ʧ v w ð g gɹ ɹ ʃ ʍ θ θɹ
Mid-word: b bl d f ft fɹ h k l ldɹ lm ls lw lɹ m mb ms mθ n nd ndɹ nl ns nt ntɹ nɹ p pl s st sʧ t ts tw ʧ v w z zn ð ŋgl g ɹ θ
Word final: b d dz ʤ f ft k ks kst kt l ld lf lp lt lvz m md mz n nd ndz nst nt nʧ nz p ps pt s skt st t ts ʧ v vd z ð ŋ ŋk ŋz g ɹ ʃ θ
Vowels: aɪ aʊ eɪ i iə iː oː u uː æ ɑː ɒ ɔ ɔɪ ɔː ə əʊ ɛ ɛː ɜː ɪ ɪə ʊə ʌ ʊ
And there are some things in English that came naturally before knowing about it, like affixes, sound-verb-object and adjective before noun.
However, there are some things about English that I know nothing about, like its dropoff rate. Vulgar defined dropoff rate as follows:
Phonemes are ranked by frequency from left (most frequent) to right (least frequent). Medium makes the frequencies slightly more even than fast. When using equiprobable, phonemes can be custom weighted by writing =multiplier, eg: p=10 makes p ten times more common than a phoneme without a weighting.
Vulgar gives us choices for our phonemes: "fast", "medium" or "equiprobable" for dropoff rate and "naturalistic ranking", "natural with randomness", "alphabetical" or "previous" for how to arrange the phonemes. Which of these choices matches English in real life?