Exact cause of hash code collisions in HashMap

Question

I know how hashmaps work and how they resolve collisions. But many answers related to collisions on SO point to two reasons:

• HashCode contract - unequal objects can have the same hash code
• Underlying array size - is fixed in size and hence bucket locations may be reused (using modulo operator) when it gets full.

My assumption is that reason 2 above is also correct, as am sure of reason 1.

Only 1 question:

• Javadoc says that the hashmap uses LF and Capacity to decide when to rehash. I understand this as partly resizing the underlying array. So how is reason 2 above even possible?????

NB: This question came closest to what I want to understand. However, the accepted answer seems to discard reason 2 altogether, so it still leaves me wondering.

Answer 1

Your reason 2 is flawed. A bucket may be reused even if there are plenty of empty buckets if adding two (unequal) objects that end up in the same bucket (hashcode % buckets.length). There's no "getting full" element here, except that as buckets get filled there's obviously a greater chance of ending up in a non-empty bucket.