Are those two events/teams comparable so that one could expect the same type of odds?

I'm not much into betting, but will not the odds change also based on how people are betting. So if many people started to bet on the unexpected result, due to for example that the unexpected Leicester team won last time, wouldn't odds tighten up automatically?

> Are those two events/teams comparable so that one could expect the same type of odds?

Yes IMO. It might be more rigorous to say: before 2016 there were always a handful of teams priced at 5000-1, after 2016 there were not. The league didn't objectively get more competitive.

> So if many people started to bet on the unexpected result, due to for example that the unexpected Leicester team won last time, wouldn't odds tighten up automatically

Definitely possible, and would explain the effect regardless of whether the punters were smart money Nicholas Taleb types who spotted that black swans were systematically underpriced, or gamblers dreaming of another Leicester payday.

Or, cynically, their original odds were right but they have taken the opportunity to reduce payouts because punters now consider that event more likely and thus would accept less.

It's less of an indication of bookies' predictions and more of an indicator of bettors' predictions.

A bookie isn't trying to gamble on the outcome; they're trying to get the bets to all balance out so that no matter what the outcome they take in enough money from the losers to pay the winners with some profit leftover.

As more people bet on a specific outcome, they will lower the odds to make sure that the pool of money taken in is still big enough to pay out if that outcome happens.