Warhammer Underworlds - Analysis Series - Guard

By Jon Grant 1st May 2026

A few months ago, I started writing my Warhammer Underworlds Analysis Series. Here, I sought to uncover some of the maths buried in the game as a mechanism to influence decision‑making. The first article looked at durability in Warhammer Underworlds. This I defined as the total number of successful 2‑damage attacks required to remove all fighters in a warband. There are two key ways to think about this:

Base Durability

Assumes all attacks hit and all saves fail. This focuses purely on the Health and fighter count of a warband.

Save‑Adjusted Durability

Accounts for each fighter’s saves, which can cause attacks to fail and force your opponent to spend more activations to remove your fighters.

Another idea you need to understand for this article is: 

Save Breakpoint

The Save Breakpoint is the point at which an attack reaches two successful hits, causing defensive Saves to become unreliable and durability to collapse back toward its base value.

If you would like a more detailed explanation of these concepts, you may wish to read the original articles on durability and attacks before continuing.

When looking at durability, I considered it as an entirely passive component of the game. Fighters are attacked, saves are rolled, and durability emerges from how many activations it takes to remove models from the board. Now, I want to consider how durability can be actively influenced by taking a look at Guard.

Buying Time 

Guard is an active choice. However Guard is applied to a fighter, it represents an upfront investment (possibly a gamble) of resources in exchange for survivability later. This investment can come from an activation, a warscroll ability, or a ploy card.

Ploy cards that provide access to Guard

Among those costs, time is the most important. Spending an activation to put a fighter on Guard means you are not attacking, moving, or scoring that activation. In that sense, Guard is a gamble. You give something up now in the hope that your opponent will be forced to spend more activations later trying to achieve the same result. More on this later.

Guard lasts for a full round. Once applied, the token remains until the end of the current round, covering four activations for each player. If Guard forces even one additional failed attack during that window, it has begun to pay for itself. If it forces multiple failures, Guard becomes one of the most efficient durability tools in the game.

To understand why this happens, we need to look at what Guard actually changes.

What Does Guard Do?

Guard does not give a fighter extra save dice. A fighter still rolls the number of dice on their Save characteristic and nothing more. What Guard does is change which results count as successful saves.

Normally, a fighter only succeeds on saves that match their Save characteristic, with Crits always counting as successes. While a fighter is on Guard, Crits, Blocks, and Dodges all count as defensive successes, regardless of whether the fighter has a Block or Dodge save characteristic.

Guard, therefore, increases the reliability of each save die by expanding the range of faces that count as successes.

This has two immediate consequences:

• Fighters with Dodge saves benefit enormously while on Guard.
• Fighters with multiple save dice become significantly harder to damage for the time they remain on Guard.

Guard and Save‑Adjusted Durability

In my previous article, I showed that Save‑Adjusted Durability increases when attacks fail because of successful Save rolls, forcing opponents to spend additional activations to damage and eventually slay fighters. Guard pushes Save‑Adjusted Durability much further.

Against attacks which contain only a single successful hit, the effect is particularly pronounced. A fighter with a single Dodge save normally cancels roughly one attack in three. While on Guard, that same fighter cancels around two attacks in three. Functionally, Guard nearly doubles the number of attack activations required to deal the same amount of damage to that model.

The impact grows sharply with additional save dice. Fighters with two defence dice already perform well against single‑hit attacks. While on Guard, a fighter with a two‑dice Save can frequently defend multiple single‑hit attacks in a row. In practical terms, a Guarded fighter with two save dice can feel almost untouchable for a round, forcing opponents to either commit heavily or attack elsewhere.

This is the key durability insight: Guard causes opponents to spend their limited turns on attacks that rarely translate into progress.

When Is Guard Most Effective?

Guard does not provide equal protection against all attack profiles. Its effectiveness depends on how reliably an opponent can reach the Save Breakpoint. As shown in the attack analysis, most common weapon profiles in Warhammer Underworlds struggle to consistently generate more than a single successful hit. This is particularly true for Sword‑based attacks and for 2 Hammer profiles.

Two Hammer attacks sit in an awkward middle ground. They are reliable at producing a single hit, but unreliable at producing two. Roughly three‑quarters of the time they will land at least one success, but only around a quarter of the time will they reach the Save Breakpoint of two successes (you can see the probability in my Attack analysis article). Most of the time, therefore, a 2 Hammer attack presents the defender with a familiar situation, a successful hit that can still be cancelled by a Save.

This is exactly the environment in which Guard is strongest. By improving the reliability of defensive dice, Guard increases the chance that these single‑hit attacks fail. The opponent is still attacking and using their activations, but often without inflicting damage. Each failed attempt consumes one of their valuable turns, reducing their chances of scoring.

Sword attacks exaggerate this effect even further. Regardless of how many Sword dice are rolled, Sword‑based profiles struggle to reach two successes reliably without additional support. They frequently produce one hit, but rarely more. Against these attacks, Guard consistently increases Save‑Adjusted Durability, causing opponents to repeatedly invest activations in attempts to force through damage.

In contrast, Guard rapidly loses its value once attacks reliably reach two or more successes. As established earlier, two hits represent the Save Breakpoint. Beyond this point, single save dice become irrelevant (even when on Guard) and even multiple save dice have a low probability of causing an attack to fail. When facing highly accurate attacks, such as three or more Hammers with rerolls or support, Guard no longer result in sufficient numbers of failed activations and instead becomes a poor use of resources (other than to prevent Drive Back).

This explains why Guard feels amazing in some match‑ups and underwhelming in others. Guard works at its best below the Save Breakpoint, where attacks have a single hit. Two Hammer attacks and Sword‑based profiles exist almost entirely within this range. Against them, Guard converts defensive reliability directly into wasted opponent activations. Once attacks consistently exceed the Save Breakpoint, that advantage disappears.

This explains a key pattern in play. Guard feels very strong against 2–3 Sword attacks or 2 Hammer attacks, but quickly feels ineffective once facing 3 or more Hammers, with or without support and rerolls.

Why Guarded Fighters Often Fail Saves When Staggered

At first glance, Guard and Stagger appear to pull in opposite directions. Guard improves defensive reliability, while Stagger improves offensive reliability. When these two mechanics interact, however, Stagger tends to win, and the reason lies in how attacks exceed the Save Breakpoint.

Guard increases the likelihood that a defender will roll at least one successful save. This is highly effective when attacks are producing only a single hit. In that situation, a single success is enough to cancel the attack entirely, and Guard dramatically increases the chance that this happens. The opponent spends an activation, rolls their dice, and nothing changes on the board.

Stagger changes the nature of the attack. By granting rerolls, Stagger does not simply make it more likely that an attack will hit; it makes it more likely that the attack will hit multiple times. As shown in the attack analysis, rerolls significantly increase the chance of rolling two or more successes. Once an attack reaches two hits, the defender is no longer looking for a single successful save, but for multiple successes in the same roll.

This is where Guard begins to fail. Although Guard improves the reliability of each individual save die, it does not increase the number of dice rolled. A Guarded fighter still rolls the same number of defence dice as before. When hit with two successes in an attack, that small defensive dice pool is often insufficient. Single save dice become irrelevant, and even two save dice struggle to cancel the attack consistently.

In practical terms, Stagger pushes attacks to exceed the Save Breakpoint. Once that happens, the defensive advantage provided by Guard ends. The opponent is no longer wasting activations.

As such, Guard is most effective in a games state of unreliable, single‑hit attacks. Stagger shifts the game into an environment of reliable, multi‑hit attacks.

Guard is Multiplied by Health

Health adds durability over the long term. Guard produce a temporary spike in durability that disappears once the round ends.

This is why Guard often feels stronger on fighters who already require multiple hits to remove. A 5‑Health fighter that needs three successful attacks to be taken out has more opportunity for Guard to operate effectively. If even one of those required attacks fails, the opponent’s plan is significantly disrupted and your fighter may remain in place. If two attacks fail, the fighter on Guard may survive the entire round, and your opponent has likely wasted opportunities to score Objectives.

By contrast, low‑Health fighters benefit less from Guard. Guard still makes removing them less reliable, but once a single successful 2‑damage attack lands, a 2‑Health fighter is gone.

In effect, Guard is strongest on fighters who both:

• Have multiple save dice, and
• Can survive more than one successful hit.

These fighters act as durability anchors, drawing attacks and tanking activations disproportionately while they remain alive.

Guard, Warband Size, and Activation Pressure

Guard also scales inversely with warband size. Small elite warbands can create more situations where opponents have few options other than to attack fighters on Guard. In larger warbands, it is easier for opponents to ignore Guarded fighters, reducing its overall value.

Guard can therefore be used to control where an opponent allocates their limited number of attacks each round and, again, to slow their Objective scoring.

Guard Has a Shelf Life

Despite its strengths, Guard is not a permanent boost to durability.

Once the round ends and Guard tokens are removed, the fighter immediately reverts to their normal defensive profile.

This is why timing matters. Guard is most powerful when applied:

• Early in a round
• On fighters likely to be attacked multiple times
• When opponents cannot easily generate multiple hits

Used correctly, Guard can buy an entire round of scoring and positioning. Used poorly, it is simply a waste of resources.

Guard as an Activation Economy Tool

When Guard is viewed through the lens of durability, it is best understood not as defence, but as a form of activation control.

You spend an activation now so that your opponent must spend more later to achieve the same goal. If they refuse, they concede board positioning and scoring opportunities. If they commit, they risk wasting attacks that achieve nothing.

This framing explains why Guard is often strongest in Take and Hold playstyles rather than Strike. The value of Guard is not measured in wounds prevented, but in opponent activations that achieve nothing.

As durability is effectively a measure of time, time is the most important resource in Warhammer Underworlds.