TFT economy: gold, interest and streaks

Gold in TFT is not a score and it is not a safety net. It is a machine for converting time into board strength, and the machine has a fixed maximum rate. Understanding that rate — and the fact that it caps — is most of what separates an economy that works from one that feels like hoarding.

The income table

Your gold at the end of a round is the sum of three things. They stack, always, and none of them is optional.

Add the best of each and you get 14 gold a round: 5 base, 5 interest, 3 streak, 1 for winning. The floor, for a player at zero gold alternating wins and losses, is 5. Nearly three to one. Across the twenty-odd rounds of a real game that gap is well over a hundred gold — roughly fifty shop refreshes, or two and a half levels.

That is the whole economy game in one sentence. There is no clever trick underneath it; there is a table, and there is the discipline to stay on the right side of it while the round timer is running and your board is losing.

Interest

+1 gold per full 10 held, at the end of every round, capped at +5. Three things follow that most players know and don't act on.

50 is the number. Not "save as much as possible" — 50 exactly. At 50 gold you are earning the maximum the game will ever pay you for holding. The 51st gold earns nothing. Gold above 50 is not economy, it's just gold you haven't spent yet, and it is only correct to hold it if you have a specific plan for it within a round or two.

Spending down to a breakpoint is free. If you're at 54 gold, the first four you spend cost you nothing at all. If you're at 50 and you spend two on a refresh, you've paid 2 gold and a gold of interest next round, and the round after, and every round until you're back at 50. A 2-gold roll below a breakpoint is not a 2-gold roll.

The ramp is slow at the bottom. Going from 0 to 50 gold on base income alone takes many rounds, and each 10 you cross speeds up the next. This is why the early stages matter more to your economy than the late ones: gold saved at 2-3 compounds for fifteen rounds, gold saved at 4-5 compounds for four.

Streaks

Win streaks and loss streaks pay exactly the same: +1 at two in a row, +2 at five, +3 at six or more. The bonus caps at +3; the streak itself never does, so a tenth consecutive loss still pays +3.

The strategic content of that symmetry is simple and under-appreciated: the worst record in TFT is a mixed one. Winning four and losing four in alternation earns you nothing while the player who lost all eight banked +3 a round for most of them and picked first at every carousel. If you cannot win, losing properly is a real plan. What isn't a plan is drifting.

Win streaking

The stronger of the two when it's available, because you collect the streak bonus and the +1 for winning and you keep your health. Health is what buys you the right to roll late. The trap is spending to maintain a streak: buying a 3-cost you don't want at 2-5 to win one more round is usually a loss, because you paid 3 gold plus the interest to gain 4.

Win streaking works when your board is naturally ahead — good early units, an early item, a strong augment. Chasing it when you're not ahead is how people end stage 3 at 20 gold with a board of units they don't want.

Loss streaking

You deliberately field a weak board, collect the streak gold, and take the carousel priority that comes with low health. Done properly you arrive at stage 4 with 50 gold, a level advantage and first pick on every carousel since 2-4. Done badly you arrive at stage 4 with 50 gold and 12 health, and die before you can spend it.

The discipline is that loss streaking has an exit and you must know it before you start. Pick the round you will stabilise on — it's usually 4-1 or 4-2 — and work backwards to check you'll have both the gold and the health to survive the attempt. Losing on purpose without a stabilise round chosen is not a strategy, it's a story you tell yourself while you lose.

The three econ plans

Almost every game resolves into one of three shapes. You don't choose at 1-1; you notice which one the game is offering you by about 2-3.

The middle row is not a failure state. Standard econ is correct most of the time and it is what you should default to when the game hasn't told you anything. The two extremes are what you do when it has.

When to break econ

Economy discipline is a default, not a religion. There are four situations where spending below 50 is correct, and recognising them is as much a skill as saving is.

• You are about to die. Interest on gold you never spend is worth nothing. Below roughly 25 health in stage 4, the calculation inverts: the strongest board you can field right now beats a stronger one two rounds from now, because you may not get two rounds.
• A one-gold purchase completes a 2★. Buying the third copy of a unit you're already playing is almost never wrong. It is the highest return per gold in the game and it doesn't commit you to anything.
• You've reached your roll level with 50+ gold. This is the plan working. Spend it. Sitting at 70 gold at level 8 at 4-3 because "saving is good" is the most expensive way to misunderstand this chapter.
• Levelling wins the round outright. Four gold for an extra board slot that turns a loss into a win pays for itself immediately in health and streak.

Notice that three of the four are about spending. The failure mode of learning economy is over-correcting into hoarding, and a player at 90 gold in stage 5 with a level 7 board has made a bigger mistake than the one who rolled too early.

Common gold leaks

In rough order of how much they cost the average player:

1. Rolling without a target. Refreshing to "see what's there" costs 2 gold plus interest plus the attention you didn't spend on positioning. If you can't name the unit you're rolling for, close the shop.
2. Buying units you won't play. A 3-cost bought on impulse at 3-2 is 3 gold and a bench slot, and you'll sell it for 3 gold two rounds later having gained nothing but a broken interest breakpoint.
3. Levelling off-curve. XP costs 4 gold for 4, so buying XP when you need 2 more wastes 2. Small, but it happens five times a game. See chapter 3.
4. Holding past 50 with no plan. Every round at 60 gold is a round where 10 gold did nothing at all.
5. Breaking a streak for one gold. Selling a unit to squeeze out a win that ends a 5-loss streak trades +2 a round for +1 once.

Economy tells you how much you can spend. The next chapter is about the single largest thing to spend it on — levels, and when to buy them.