Guessing Arabic Present Tense from Past Tense

I have a question about predictability of the present tense forms of Arabic verbs only from their past tense forms. In Arabic, most verbs are made up of 3-consonant verbal roots—roots with 4 consonants are a lot less common and not of interest here. Adding vowels to 3-consonant roots, we get the default past tense forms, i.e. the 3rd person singular. In the active voice, the first and the third vowels are always fatha (a) and the second vowel is one of all three possible vowels of Arabic, i.e. ـَ (a, fatha), ـِ (i, kasra), or ـُ (u, damma) depending on the verb. The prototypical 3-consonant root is فعل (f-ʕ-l), related to ‘doing’, which has the past tense form فَعَلَ (faʕala), meaning ‘he did’. The default present tense form of a 3-consonant root is constructed by adding the prefix يَ (ya-) to the root, having a ـۡ (no vowel, sukūn) on the first consonant of the root, adding the suffix ـُ (u, damma) to the last consonant of the root, and adding one of the three possible vowels to the second consonant of the root.

This is where it gets interesting. The vowel of the second consonant of the root in the past tense form and the vowel of the second consonant of the root in the present tense form are not fixed and can be one of the three possible vowels. I had read of a rule that if the second vowel in the past form is damma (u), then the second vowel of the present form is always damma (u). For example, the 3rd person singular past tense form of the root بعد (b-ʕ-d) is بَعُدَ (baʕuda), meaning ‘he was far’, and the 3rd person singular present form is يَبْعُدُ (yabʕudu), meaning ‘he is far’. Then I wondered if the intuition of native speakers can be put into statistics and predict the present form of a verb only from its past form, specifically predicting the second vowel of the present form from all the letters of the past form.

… (to be continued)