Every Game Has 5-6 Decisions That Matter

Most TDS players autopilot through the first 20 waves — place a tower here, upgrade a Farm there, following the same build order every game. Then wave 25 hits, something goes wrong, and they can’t figure out why.

The reason: TDS isn’t a tower placement game. It’s a timing game disguised as a tower defense game. You have a limited number of coins and a limited number of waves before each difficulty spike. Every purchase is an opportunity cost — buying X means you can’t buy Y for 2-3 more waves. The players who win consistently don’t place better towers. They make better timing decisions.

This guide covers the 5 critical decision points in a standard 35-wave game, and the framework for making them right.


Decision 1: Wave 5 — First Farm or First DPS?

You have 700 coins. The shop offers Farm (400 coins) and Militant (500 coins).

What most players do: Buy a Militant. “I need defense! Enemies are getting stronger!”

Why it’s wrong: Wave 5 enemies have 10 HP. Your starting Scout/Freezer handles them fine through wave 8. The Militant sits there doing nothing useful while the Farm would already be generating passive income.

The right choice: Buy Farm and upgrade it to level 2 (700 coins total). By wave 10, this Farm has generated 1,750 coins — paying for itself and earning you an extra 1,050 coins that a Militant-first player doesn’t have.

The exception: If you’re playing Hardcore mode or a mode where wave 8 has a dangerous enemy, buy minimal defense FIRST (one Militant at level 1, 500 coins), then Farm. But in Normal, Fallen, and Golden modes, Farm first is always correct.


Decision 2: Wave 12 — Second Farm or First Commander?

You have 1,200 coins. One level 2 Farm. One Militant for early defense. Enemies are starting to get tankier.

What most players do: Buy a second Militant or a Minigunner. “I’m worried about wave 15.”

Why it’s wrong: Wave 15 enemies have 30-40 HP. One level 2 Militant with decent placement handles them. What you actually need is scaling — the ability to afford multiple high-level DPS towers by wave 25.

The right choice: Second Farm, upgrade to level 2 (700 coins). Then buy Commander (800 coins) on wave 15. The second Farm finances the Commander. The Commander’s fire rate buff (+25% at level 1) effectively makes all your towers 25% stronger — more value than a second DPS tower for the same cost.

The framework: Before buying a second DPS tower, ask: “Can I survive the next 3 waves with my current defense?” If yes: Farm or support tower. If no: minimal DPS investment (one level, not max).


Decision 3: Wave 20 — Upgrade DPS or Place DJ?

You have a Commander, 2 Farms at level 2, and a Militant at level 2. You have 1,800 coins. Wave 20 has the first real boss.

What most players do: Max out the Militant (level 3 costs 1,800). “Boss is coming, need damage!”

Why it’s wrong: DJ costs 1,200 coins and gives a 15% range buff + 10% discount on all upgrades within its radius. A level 3 Militant with DJ’s range buff covers 30% more track. A level 2 Militant + DJ costs the same (1,700 total vs 1,800 for level 3 Militant) and is more effective.

The right choice: Place DJ first (1,200 coins). Then upgrade Militant to level 3 (1,800 × 0.9 = 1,620 after DJ discount). You now have a level 3 Militant with +15% range, and future upgrades in DJ’s radius cost 10% less. The DJ pays for itself by wave 30 through upgrade discounts alone.


Decision 4: Wave 27 — Sell Early Tower or Keep It?

Your level 2 Militant is barely scratching wave 27 enemies. It’s dealing 8% of their HP per shot. You need 2,500 coins for Minigunner but only have 1,200.

What most players do: Keep the Militant and save for 3 more waves. “It’s still doing SOME damage. Selling it wastes coins.”

Why it’s wrong: A tower dealing 8% damage per shot on wave 27 is essentially doing 0 damage — enemies will walk past it before it kills them. It’s taking up a placement slot that a better tower could use. And the 500 coins from selling it gets you to Minigunner 2 waves earlier than saving.

The right choice: Sell the Militant (returns ~1,260 coins), buy Minigunner immediately (1,500 coins). The Minigunner at level 1 does more DPS than a level 2 Militant against wave 27+ enemies. You traded a dead-weight tower for a meta tower, and the 2-wave head start on Minigunner upgrades saves your run.

The framework: If a tower is dealing less than 15% of an enemy’s HP per shot, sell it. The sell penalty (30%) is less than the cost of keeping a useless tower — the placement slot it occupies has value.


Decision 5: Wave 32 — Max One Tower or Place Multiple?

You have Minigunner level 2, Commander level 2, DJ level 2, 2 Farms level 3. You have 4,500 coins. Wave 35 is the final boss.

What most players do: Buy a second DPS tower (Accelerator, Turret, or another Minigunner). “More towers = more damage.”

Why it’s wrong: A second DPS tower at level 1 deals ~25% of the damage of a fully maxed first DPS tower. You’re spending 2,500+ coins for a tower that won’t reach its potential before the game ends.

The right choice: Max your existing Minigunner to level 4 (3,200 coins). A level 4 Minigunner with Commander and DJ buffs deals 4x the DPS of a level 2 Minigunner. The remaining 1,300 coins go to Commander level 3 for another fire rate buff. One maxed tower with full support beats two half-built towers.

The framework: In the final 5 waves, max 2 towers with full support rather than placing 4 towers at half-strength. Support buffs (Commander fire rate, DJ range + discount) multiply each other. A level-4 DPS tower with level-3 Commander and level-2 DJ is worth more than three level-2 DPS towers with no support.


The 5-Decision Summary Card

WaveCoinsRight ChoiceWrong ChoiceWhy
5~700Farm → level 2Buy MilitantFarm pays for itself by wave 9
12~1,200Second Farm + save for CommanderSecond DPS towerScaling > immediate defense
20~1,800DJ first, then upgrade DPSMax DPS, skip DJDJ discount pays for DJ by wave 30
27~1,200Sell dead tower, buy meta replacementKeep underperforming towerPlacement slot value > sell penalty
32~4,500Max 1 DPS + max 1 supportPlace multiple half-built towersBuff stacking > tower count