Big Moves
Incremental games, where players make a series of small moves, are typically boring and predictable.
It's okay to have lots of small turns, but this should be punctuated by a few big turns that really change up the game.
On most turns, the player should have the option to make a big move, even if they actually do so far less often.
Big moves also allow players more control over their game experience.
In Reef, players build a three-dimensional matrix of coloured plastic "coral" bits. Cards will give a player points if there's a specific pattern of coral on their board. Players can make lots of little patterns, and score a few points each. Alternatively, they can build up to a mega-score, where they build multiple copies of the same pattern, and score the card multiple times at once.
If all moves are about the same magnitude, it also doesn't allow losing players to get back into the game.
Early in the development of my gangster game, you could see who was going to win, by the middle of the game. The player who had the most health and money tokens was in a commanding position, and little could stop them. The game was too incremental. I added numerous flashy effects, such as the Skull, which caused all players to lose half their health, and the Bomb, which caused opponents to lose a very large five health, but you had to successfully keep the token for a full two rounds before it exploded.
Big moves make the player feel free and powerful, like they're exploiting the game. These are the moves that make for memorable gameplay.
Create a class of "big things": expensive things with big effects that almost break the game.
In Radlands, there are people and effects that cost all your water (the game's resource) for one turn, and have effects that drastically alter the game.
I'm constantly looking for places to use the words "triple" and "double". Things with these words can also qualify as awesome things.
In my farm game, you spend all game building towards fulfilling your red "deed" cards. They give you lots of points. The Manor building is expensive, and makes your completed red deeds worth twice as many points.
Also in my farm game, most tokens aren't worth anything. Harbor makes all your tokens give you one point, each round.
Tokens like gold, diamonds, and pearls are worth points. The Treasury building says "tokens are worth triple points."
Variable actions
Try to make as many of your options variable as possible, so the players can construct all kinds of big moves of their own.
One way to create variable actions is by having variable costs, with the effect growing or repeating as the player pays more.
In my farming game, most actions can be repeated any number of times at once, limited only by your ability to pay.
Variable effects can also come from combos and setups, that make some turns awesome.
If I can line up ten space orcs, I should be able to zap all of them with one shot of my laser gun.
In A Feast For Odin, the Overseas Trading space upgrades any number of different green goods to blue. This is a big move, as players will set up for it by collecting green goods, and doing one big trade.
Try to use the word "all" (or each/every) on many of your objects. "Gain 1 fruit for each tree you have" is far more interesting than "gain 3 fruit."
And, if you can't make your actions variable, try combining them.
In Agricola, the "Bake Bread" action just turns some of your grain into food. That's a bit boring, so it's paired with the action that lets you plant things in your fields. This way, players will gather lots of grain, and things to plant, and then do a single big move, and accomplish a lot.
The size of the moves in your game should vary significantly, ideally as dictated by the players.
Stupid fun
You might say "all this potential for big moves will lead to outrageous and imbalanced stuff happening!" Yes, and it's great fun for the players.
Players should not feel penned in. They should feel like they can do anything, and break or exploit the game.
In a farming game I was working on, the players could farm a modest amount of crops and animals, and do some other things, within reason. Later, I gave players more freedom, and they started to abuse the game. One player created a giant herd of sheep for their wool, bought a sewing machine AND a loom, and began to mass-produce clothing. He didn't need that much clothing, but each piece was worth points. Another player ferried in an army of guests, and used his cooking pot to create an inordinate amount of soup to feed these happy visitors, which also earned him copious points.
I was very pleased to see this. These players were having fun. I made sure that these kinds of combos couldn't be set up easily, or be maintained with simple loops of actions. My job wasn't to stop the madness. It was to make sure the madness was fair.
Stupid or outrageous fun is still fun, and is often the best kind of fun. As a designer, it's your job to make it work.
In Radlands, the Gunner card hits all the opponent's unprotected people. (Notice the word "all" in there.) The card High Ground makes all the opponent's people unprotected. This is an obvious combo, but players love it.