With four enemies converging on one objective, you cannot reach all of them. This guide is a method for choosing which to answer, and - just as important - for deciding which to deliberately give up on.
Time to objective, not distance to objective
The naive rule is to fight whatever is closest to the objective. The better rule is to fight whatever will reach it soonest, and those differ whenever enemy speeds differ - which is most of the time.
A rusher at 200 pixels moving at 118 pixels per second arrives in 1.7 seconds. A brute at 120 pixels moving at 58 arrives in 2.1. The brute is closer and the rusher is the problem.
You do not need to compute this. The heuristic is that fast enemies outrank slow ones at any similar distance, and the enemy types are visually distinct precisely so this judgement is quick.
Ranged enemies outrank everything
A caster does not need to reach the objective at all. It stops at range and shells it, which means distance-based priority does not apply to it in any form.
It is also the only enemy whose threat does not decrease while you deal with something else. Everything with legs has to travel; a caster is already in position.
In Blade Defender, casters target the core rather than you, which makes them invisible as a personal threat and easy to leave alive. That is the single most common reason a core falls in that game.
Ad space reserved. See README.md to connect an ad unit.
Using your own body
Standing between an enemy and the objective works. Melee enemies will engage you rather than walking past, which converts a threat to the objective into a threat to you - and your health is not the fail state.
This is the most reliable emergency answer when you cannot kill something in time. It costs health, it costs the combo, and it is frequently correct.
It does not work on casters or on anything flying past you, which is another reason those rank higher.
The order, summarised
In rough order of priority for any defence game on this site.
- Casters and anything ranged, always first, regardless of position.
- Fast enemies closest to the objective by arrival time.
- Anything you can kill on the way to something else - free value, no detour.
- Slow, tough enemies far from the objective. They can wait.
- Anything attacking you personally, unless it is preventing you from moving.
- Spikers, which cannot be attacked and should simply be routed around.
Re-checking the order
Priority is not decided once per wave. The ordering changes as enemies move, take damage and change state, and a player who commits to a plan at the start of a wave is usually executing an out-of-date one by the middle of it.
The cheap version of re-checking is a single glance at the objective every few seconds. If nothing is close to it, whatever you are currently fighting is fine. If something is, the ordering has changed and the enemy in front of you has just stopped being the priority.
That works because the objective is one place rather than several. Scanning it costs a fraction of the attention that tracking five enemies would, and it answers the only question the priority list exists to serve.
The cost of switching
Correct priority still has a price: every switch costs the distance between two targets, and switching constantly is its own failure mode.
A player who re-evaluates every half second spends the wave travelling rather than fighting, arriving late to each threat and committing to none. The priority list is meant to be consulted at decision points - when a target dies, when something new appears, when the objective comes under threat - rather than continuously.
The balance is to choose deliberately and then commit until something changes. A slightly suboptimal target fought to completion beats an optimal one abandoned halfway, and the difference between the two is usually the run.
Deciding what to concede
With four incoming and time for two, two are getting through. Choosing which is better than discovering it.
Concede the cheapest. A single rusher does one damage to the objective; a brute does two. If one of each is getting through, let the rusher go.
Concede the recoverable. If letting one through costs a repair you would have earned anyway - because something else already got through this wave - the marginal cost of the second is lower than the first.
Never concede a caster. It will keep dealing damage after the wave rather than once.
Key takeaways
- Prioritise by time to objective, not distance - a fast enemy further away often arrives first.
- Ranged enemies outrank everything because they never need to arrive at all.
- With more threats than time, choose what to concede rather than discovering it.
- Concede the cheapest damage, never a caster.
- Standing between an enemy and the objective converts its threat into a threat to you, which is not the fail state.
- Anything attacking you personally is near the bottom of the priority list.
- Re-check the ordering at decision points, using a glance at the objective rather than a scan of every enemy.
- Commit to a chosen target until something changes - an optimal target abandoned halfway loses to a suboptimal one finished.
- A caster that ignores you personally is the single most commonly left-alive threat in the category.
- An enemy fighting you is spending time it would otherwise spend advancing, which is why it ranks low.
Frequently asked questions
Which enemy should I attack first in a defence game?
Anything ranged, always. After that, whichever will reach the objective soonest by arrival time rather than by distance.
Why do casters rank so high?
Because they never need to arrive. They stop at range and damage the objective from there, so distance-based priority does not apply and walking away does not reduce their threat.
What should I do when I cannot reach everything?
Choose what to concede deliberately. Give up the cheapest damage - a single rusher over a brute - and never give up a caster.
Does standing in the way work?
Yes, against melee enemies. They will engage you rather than walking past, which converts a threat to the objective into a threat to you - and your health is not the fail state.
Where does the enemy attacking me rank?
Near the bottom, unless it is physically preventing you from reaching something more important. An enemy fighting you is doing no damage to the objective.
How often should I re-check my priority?
At decision points rather than continuously - when a target dies, when something new appears, or when the objective comes under threat. A glance at the objective every few seconds answers the question more cheaply than tracking every enemy.
Is it bad to keep changing targets?
Yes. Every switch costs the distance between the two, so a player who re-evaluates constantly spends the wave travelling and arrives late to everything. A slightly suboptimal target fought to completion beats an optimal one abandoned halfway.
What is the single most common priority mistake?
Leaving a caster alive because it is not attacking you. In Blade Defender casters target the core rather than the player, which makes them invisible as a personal threat and is the usual reason a core falls.