One serious adjustment that long-time players find themselves making in my world is the action point system. There are a lot of different things that cost different amounts, and from the outset these seem arbitrary and difficult to incorporate into standard play. Most campaigns are not tactical in nature ~ the old way is how I played for decades before altering my play 12 years ago.
Being forced to accept that characters don't just attack monsters in neat lines, that they have to account for every action as a measured deal, chafes at some people's conception of the game and simply acts as a confusion for others. D&D is not a war game, I have been told; but I hasten to disagree.
D&D is a game ~ and combat is a game mechanic within that game. It needs rules to make it interesting, to provide a payoff and make the characters feel threatened and fearful of dying. Forcing players to account for their actions produces a limitation on how many actions they can take; this makes them feel that adequacy to the event is in question, which creates stress and concurrently risk that they must take in order to overcome the game's obstacle, the monsters trying to kill them.
But the action point system has one more element that makes this difficult, and it is the hardest element to initially accept.
The tendency is to suppose that dividing the round into points creates a group of 'mini-rounds,' which of course can bleed over from round to round. If it takes four action points to load a bow, it is assumed automatically that the character can spend 2 action points this round and then 2 action points next round.
I don't allow this and I want to explain why.
By limiting the expenditure of action points to that which can be performed in a given round, and not letting them bleed over (with a few specific exceptions, for actions which cannot be performed in the space of 12 seconds), this forces the player to see each round as an exercise in efficiency. They can't rely on events being the same after the enemy attacks, so each time the enemy attacks resets the system. If, then, they can't make use of all their action points in a given round (or they don't have enough action points), they must adapt to that circumstance either by choosing a different action or wasting time doing nothing.
The tendency of players to view time in a role-playing game as something that can be easily allotted like sorting out the number of cookies between themselves and their companions is a weakness of most games. Rushing about in a stress situation is full of stopping and starting, of being confused, of having to wait for someone in front to move so that we can go ahead, of planning to do something only to find that it is now too late and so on. By forcing players to think in one-round limitations instead of making plans for how they'll spend their next forty action points over the next eight rounds, they must reassess and reassess, constantly.
Because the combat is frenetic and changes the balance of what's happening, players CAN'T have an organized plan of what they'll do for the next bunch of rounds. They have to work to be efficient one round at a time with their movement, hoping they can use their points in the best possible way. This means they will intentionally go one hex further just to avoid wasting the point, or deciding to pull out something this round (or forego it) because they have extra time and they're thinking "what can I do with this point?"
This pattern produces a much more confused pattern of behaviour, making the combat overall more random, less controlled, harder to predict and, therefore, a more difficult sub-game to play, since the learning curve is steeper than it would be if action points were allowed to bleed from round to round.
Overall, making the game harder makes it better. Those players who have adapted are able to sense this, though perhaps not all may realize that it is the lack of bleed that improves play, rather than subverting it.