March Math: The 40/30/30 Doctrine
Guards eat damage, Gunners grind the midline, Marksmen burst from the back. The proven competitive split — and when to break it on purpose.
Troops win fights before heroes ever cast a skill. Tiles Survive! runs a rock-paper-scissors counter system between Guards, Gunners and Marksmen that compounds multiplicatively — which means a balanced, lower-power march routinely deletes a single-type stack with a much bigger power score.
This is the math behind that: the doctrine split, the counter triangle, the tier gating that makes Power Plant your real Barracks, and the three situations where you break the baseline on purpose.
The doctrine: 40/30/30
The proven competitive baseline is 40-50% Guards, 30-40% Gunners, 20-30% Marksmen. The do-nothing-wrong midpoint is 40/30/30 — strong frontline, consistent midline damage, backline burst, with no single type overexposed to its counter.
The one hard floor: never drop below 40% Guards. Below that line your frontline collapses in the opening damage exchange, and Gunners and Marksmen die before their ultimates resolve. A march that never gets to cast is a march that was dead on arrival regardless of its backline on paper.
| March type | Guards | Gunners | Marksmen | When |
|---|---|---|---|---|
| Baseline doctrine | 40% | 30% | 30% | Default for rallies and unknown matchups |
| Frontline-heavy | 50% | 30% | 20% | PvP sieges and long defensive stands |
| Sustained PvE | 40% | 40% | 20% | Monster hunting, infected waves, boss brackets |
| Burst punish | 40% | 25% | 35% | Scouted a Gunner-heavy enemy (Marksmen counter Gunners) |
The counter triangle
Troops counter each other in a fixed loop, and the damage modifiers stack multiplicatively with everything else — hero faction counters, gear special stats and research buffs all ride on top of it.
- Guards counter Marksmen — the frontline eats the backline's burst window.
- Marksmen counter Gunners — which is why every march keeps a 20-30% Marksman floor even in Guard-heavy setups.
- Gunners counter Guards — the workhorse damage that grinds through frontlines.
Tier gating: the Power Plant is your real Barracks
Every troop tier from T6 to T10 requires a matching Power Plant level before your training queues can produce it at full output. Rush Barracks research ahead of your Power Plant and you train lower-tier troops at max cost — every Waypoint converted into training resources produces less than it should.
The order is fixed: Power Plant first, Barracks second, troop-tier research third. A T9 Guard produces roughly double the rally power of a T8 Guard at the same training cost, so every Power Plant level compounds the value of every training queue you run for the rest of the account.
Heroes ride on top of troops
The troop layer compounds with the hero layer. A Guard-class hero leading Guard troops in the frontline slot produces measurably more output than the same hero leading a generic mix — hero class, troop type and formation slot all want to agree.
The current meta march is the living example: Rosie, Nikola and Tarzan anchoring the front while Tara and Layla support from behind. It's also a faction play — four Stalwart heroes plus one Rover aligns the squad for faction bonuses on top of class matching.
When to break the doctrine on purpose
The baseline is for unknown matchups. The moment you have scout data, adjust: against Gunner-heavy attackers, shift Marksman share up and turn their midline into a liability. On defense — gold mines, garrisons, resource tiles — lean Guard-heavy so the stand lasts until reinforcements land.
In PvE, Gunner share carries sustained damage through long waves, which is exactly what boss brackets like Ghoulion Pursuit score. The doctrine is a floor plan, not a cage: scout, adjust, win.
Field notes // tl;dr
Burn this into the map table
- 40-50 / 30-40 / 20-30 is the competitive split; 40/30/30 is the safe midpoint.
- Never below 40% Guards — the opening exchange decides whether your ultimates ever cast.
- Counters compound: Guards > Marksmen > Gunners > Guards.
- Power Plant → Barracks → research. T9 ≈ 2× T8 power per unit at the same cost.
- Match hero class to troop type to formation slot; scout before you specialize.
Put it to work
Field questions
What is the best troop composition in Tiles Survive?
The proven competitive split is 40-50% Guards, 30-40% Gunners and 20-30% Marksmen. It covers all three damage types simultaneously and doesn't overcommit to any single type's counter.
What counters what between Guards, Gunners and Marksmen?
Guards counter Marksmen, Marksmen counter Gunners, and Gunners counter Guards. These modifiers compound with faction counters, gear stats and research.
Why do I lose to marches with lower power?
Power score is a sum, not a matchup. Counter multipliers, faction alignment, hero-skill timing and tier synchronization all outweigh raw power — a T9-heavy balanced march beats a mixed-tier stack with a bigger number.