Game economies typically inflate over time. It's quite a simple way of doing things that simultaneously confers benefits to more involved players in the same way fair, non-inflationary economies do. When constructing the economic landscape, it's important to insure inflation provides no hindrance to new players. A mature, inflated economy necessarily bars new entrants and should be avoided lest a prohibiting in-crowd is fostered. Rather than talk about different kinds of economies, I'd like to look at deflationary practices as seen in other games.
Many games sloppily tie in one-time purchases for in-game money and expect it to depress the quantity of money held by players. The raw value injection into the economy is done by the game world to support player activity. If you loot a bandit and get silver pieces or loot a squirrel for its acorn you're creating money - and it's not out of thin air as people say, you're creating it out of time. In any game where players pay per-month, it's fair to expect that time put into the game acts as a function for money generation. Below are a few styles of temporal gold sinks.
1) Reagent Spells
In D&D, sometimes you literally have to use silver or gold pieces directly to power spells. Or, burn diamond dust (
https://roll20.net/compendium/dnd5e/Stoneskin). If approached with a more flavorful approach, I think this is a very good way to remove gold from the economy.
2) Goblinization
In MMOs, there's a rare subset of players that just like to play markets and sit on unending amounts of gold. This goblin archetype is difficult for me to understand the appeal of but I know quite a bit about it because my father hit gold cap on 4-5 characters at the same time in the original World of Warcraft: The Burning Crusade expansion and there is a lot to it. Economies of scale often filter through with the pareto principle - the top 20% tend to own 80% of the capital, for example. As long as this practice isn't discouraged, there will be some rare players who do the servers deflationary work for you.
3) Task Assistance
For all tasks, players self-select the tasks they're willing to do. In many cases due to friends or guilds, there are tasks people don't want to do but will participate nonetheless. Much of the time, people will take a less profitable course to completion including purchasing sub-optimal (profitability wise) assistance. If the task is combative: Purchasable buffs. Administrative: A subscription to efficiency. Geographic: Faster movement. Whether the task is a one time thing or repeating, a temporary desirable purchase can be offered to temporarily help the person or the person's group.
4) Procedural Power / Escalating Rarity
In D&D, it's easy to create procedural power, so I'll use an anecdote to explain what it is and it's deflationary purpose. Imagine a potion that costs 1g and makes you do 1% more damage. At times, players can opt to combine those for a potion that makes you do 2% more damage. Then, they can combine 3 of those for a potion that makes you do 4% more damage. Then, 3 of those for 8% more damage, and so on. The math here is that a cubing of the cost correlates with a squaring of the power received, which on the surface looks like it gets out of hand quickly but truthfully the cost gets out of hands way, way faster. This kind of curve is easy to write up and play out in paper RPGs, and programmatically speaking I imagine there is a way to replicate this with a simple system, but I am curious if a plain exponential gold sink would be easily spotted as a wolf unto your gold in sheeps clothing. The ease of understanding in D&D tends to play nice when 'saving up power', as players tend to hoard the activatable until there are special occasions. They may have an assortment tier 1, 2, and 3 potion all in their inventory and the time they spend ecstatically awaiting their additional tier 1 potion acquisitions, because they haven't had a tier 4 yet and it'll be *twice as strong* (wow!). It feels like watching someone who has a lottery ticket, despite it not being up to chance - they're sitting on their stash and waiting provides excitement.
Escalating rarity works the same way - something that is available to 1%, 0.5%, 0.25%, etc of the playerbase tends to take exponentially more money due to the typical Pareto Distribution of funds throughout the playerbase. Looking at Classic WoW, a cosmetic mount and title "Scarab Lord" which only about 1-5 players per server get ended up costing upwards of 50,000g in some places (The most expensive normal in-game purchase is 1000g, which takes 25 hours of farming to attain from raw creature farm, Scarab Lord traded in 625 hours of players farming creatures for one player to attain).
5) Pay-To-Explore Hill Climbing
From the perspective of game design, Hill Climbing has a specific meaning:
https://en.wikipedia.org/wiki/Hill_climbing. Any time players may be literally exploring the map, or exploring a more abstract space, payment to continue traversing in one direction which will then reveal if it's higher or lower rewards after the fact.
Conclusion
In all opportunities, avoid one-time costs. Deflating fantasy economies has been an exercise in creating as many subscription or recurring-type fees as players will tolerate. The amount of inflationary or deflationary pressure in Ashes of Creation that will world best will be up to time to tell.
