Measured
Does body armour protect the head?
- Sample
- n=87 head hits
- Tested on
- 6.37
- Tests
- T-0001, T-0003 – T-0007
- Written
Question If a unit wears a helm and a body armour, which item decides its protection against a hit to the head?
verified 6.37updated 2026-09-23 · Jake Courtney
On this page · 7 sections
Question
If a unit wears a helm and a body armour, which item decides its protection against a hit to the head?
A unit’s armour comes in pieces. A helm and a suit of body armour are separate items, and each one lists how much protection it gives to each part of the body it covers. A chain hauberk lists protection for the body, arms and legs, and nothing for the head. So when a blow lands on the head of a soldier wearing both, it is not obvious what stands in its way: the helm alone, the helm with some share of the hauberk added, or nothing at all.
Hypothesis
Before testing, we wrote: “We think the head uses the helm's own head value, 21, because the body armour lists no head protection at all.”
Before the game was created, we read the two items’ values from the armour data and wrote down what we expected. The hauberk lists nothing for the head, so we expected it to play no part there, and the head to use the helm’s own value alone.
Method
One battle, fought live on turn 1 of a small test game with the debug log switched on. A test dummy with no natural protection wore a chain hauberk (no head value) and a full helmet (head value 21). Three test attackers struck it with a slashing, a blunt and a piercing weapon, and the province's own troops joined in. Every damage roll against the dummy was read from the log, checked against its own printed arithmetic, and sorted by hit location and damage type. The head value was predicted from the armour data before the game was created.
| Test | Map | Units | Turns | Harness |
|---|---|---|---|---|
| T-0001 | Small test map, Early Ages; the battle in one nation's capital | Leve, Accensus, Rorarus · 4 custom test units | 1 | 78c5b9d5 · Headless host with the debug log (-d -d -g), parsed by strikeparse parser-01.v1 |
| T-0003 | Test map made from a random map, Early Ages; four placed battles per host | 12 custom test units | 1 | f220e06a · Headless host with the debug log (-d -d -g), four hosts, parsed by strikeparse parser-01.v1 |
| T-0004 | Three generated test maps with 8, 16 or 19 arenas; Early Ages | 21 custom test units | 1 | 08bb5c60 · Headless hosts with the debug log (-d -d -g), parsed by strikeparse parser-01.v1 |
| T-0005 | A rewritten test map with 20, 38 or 45 arenas around 8 hubs; Early Ages | 49 custom test units | 1 | de3fdaec · Headless hosts with the debug log (-d -d -g), parsed by strikeparse parser-01.v1 |
| T-0006 | 45-arena test map, Early Ages; each arena one battle per host | 87 units · 44 custom test units | 1 | e4bc0ceb · Ten headless hosts with the debug log (-d -d -g), 45 arenas each, parsed by strikeparse parser-01.v1 |
| T-0007 | 45-arena test map, Early Ages; each arena one battle per host | 63 custom test units | 1 | dde901a4 · Ten headless hosts with the debug log (-d -d -g), 45 arenas each, parsed by strikeparse parser-01.v1 |
The test was one battle on the first turn of a small test game, with the game’s debug log switched on. The log prints the sum behind every damage roll, including the protectionprotection · What stands between a unit and a blow, from its armour plus any natural protection of its own. The defender adds a DRN to it, and a hit wounds only when the damage total is higher, so a tie does nothing. Full article the defender used, so the value the game applied can be read straight from it.
The defender was a test dummy with no natural protection of its own, so any protection it had came from its two pieces of armour. Three test attackers struck it with a slashing weapon, a blunt weapon and a piercing weapon, and the troops already in the province joined the fight. We read every damage roll made against the dummy, checked that each one added up, and sorted the rolls by where they landed and by the kind of damage.
Hits that defeat armour were left out of the head count. They take a share off the protection before the roll, which would hide the value we were looking for.
Sample
6 tests · 1,501 runs · 594,600 observation records
| Test | What it tested | Status | Runs | Observations | Tested on |
|---|---|---|---|---|---|
| T-0001 | Head protection with a helm worn over body armour | Done | 1 | 5,953 | 6.37 |
| T-0003 | Piercing hits, helms and shield hits against armoured test dummies | Done | 16 | 10,158 | 6.37 |
| T-0004 | Body armour and helm protection across 19 armour designs | Done | 172 | 67,651 | 6.37 |
| T-0005 | Armour rules re-tested on new armours on a 45-arena map | Done | 412 | 161,557 | 6.37 |
| T-0006 | Real units' weapons against a helmed test dummy, and a giant with no helm | Done | 450 | 151,660 | 6.37 |
| T-0007 | A helm over body armour, the control arena of a helm-and-barding test | Done | 450 | 197,621 | 6.37 |
The question was answered by one battle, the first test listed here. Its head results are the damage rolls against the dummy that landed on its head and did not defeat its armour.
The five later tests were built mainly for other questions, and each one repeated the head check on its own test defenders, with other helms over body armours that list no head value. Their sample sizes count every record they parsed, most of them about something else; the head hits they add are counted in the claim on this rule.
Result
Every head hit that was not armour-defeating was resolved against the helm's head value of 21: at 21 for slashing and blunt hits, and at 18 for piercing hits, which reduce protection by a share of it rather than a fixed amount. The body armour added nothing to the head.
Every head hit we counted met the helm’s value and nothing more. Slashing and blunt hits met the helm’s full value. Piercing hits met a lower one, because a piercing attack counts only part of the protection in front of it, and the lower value was exactly the helm’s value after that cut. None of the head hits met a value with anything from the hauberk added.
Later tests repeated the check with other helms, over other body armours, and found the same rule each time.
What it changes
The rule now has a measured answer: when the body armour lists nothing for the head, the head’s protection is the helm’s. The combat page states it with this test as its evidence.
Open questions
This test used a defender with no natural protection and a body armour with no head value, so it says nothing about either of the cases below. Both are still being tested, and neither answer is settled enough to state here.
- What protects the head of a unit with natural protection and a helm?
- Does a helm worn over a barding add its head value to the barding's?
Sources
- Jake's test harness (headless battles, parsed debug logs) · our tests (T-0001, T-0003 – T-0007, 6.37)
Mechanics this updates
Terms on this page
Generated from the terms used above. This block is what prints.
- protection
- What stands between a unit and a blow, from its armour plus any natural protection of its own. The defender adds a DRN to it, and a hit wounds only when the damage total is higher, so a tie does nothing. · Full article · back to text
Last updated 2026-09-23 by JakeVerified for 6.37 (current)