uHammer — Army Math Workbench

Generalized army math workbench

Editable rosters plus temporary battlefield state.

The public demo starts with editable starter forces. The army editor stores the roster; the matchup panel stores the current battle state, including participating attackers, remaining defenders, cover, range, and temporary conditions.

1 · Edit rosters 2 · Set battlefield state 3 · Read statistics 4 · Plan turns ahead

Armies

Starter armies and roster editor

Use the starter forces as editable examples or select local JSON/CSV army files for import. The file picker is used to load roster data into your browser session; the files are not transmitted to a site backend.

Editing Army A Click an army card below to choose which roster is editable.

Army A

Loaded

—

Import JSON/CSV

Army B

Loaded

—

Import JSON/CSV

EDITING · Army A — —

Selected unit

—

Model groups / one-off models

Use groups for commanders, sergeants, Nobz, special weapons, heavy weapons, or any subset of a squad with different stats or equipment.

Selected model group

Group name Group type Roster count Current count Group tags Group notes

Unit name Role Type Roster models Points Unit tags

BS WS S T W I A LD Armor save Invuln FNP Cover Armor Value Hull Points Vehicle cover

Unit notes

Weapons / attack groups

Selected weapon

Weapon preset Carried by group Preset statusCustom / editable Weapon name Type Carrier count Range Shots / attacks each Strength AP Tags Notes

Validation and assumptions

—

Battle State

Matchup from roster plus current battlefield state

The roster is stable. These fields describe what is happening right now: how many attackers are participating, how many defenders remain, and what temporary conditions apply.

Attacker

Army Unit Weapon / attack group

Current models in unit Participating models Models using weapon Total shots/attacks override

—

Defender

Army Unit

Current models Current wound/HP pool Already marked damage Armor value / facing

—

Changing battle conditions

Range condition Cover override Invuln override FNP override Temporary hit condition Re-roll condition Forecast rounds Attacker output over rounds

No battle-state preset applied yet.

Battle state is temporary. It changes this calculation without overwriting the saved army roster.

Calculation

Current result Press Run Math after changing the configuration.

Current attacker-defender configuration

Blue = attacker · Red = defender · Gold = selected weapon

—

Icon and color key

● attacker/output · ★ leader/commander · ◆ special weapon/group · ▲ heavy weapon · ■ defender/protection · ! warning/custom · ○ inactive/not participating

Army A / attacker · Army B / defender · selected weapon · saves/survival

Results

Exact distribution, cumulative odds, and multi-round projection

Run a matchup to generate statistics.

CURRENT—

Expected hits—

Expected wounds / HP—

Kill chance—

Zero-effect chance—

At least selected N—

Most likely damage—

Std. dev.—

Reliability—

P10 / median / P90—

Overkill chance—

Expected overkill—

Expected rounds to kill—

How to read the charts

Exact damage distribution shows the probability of causing exactly N unsaved wounds or hull points in one attack step. Cumulative odds shows the chance of causing at least N damage. Turn forecast repeats this same attack over several rounds to show survival, kill-by-turn, and expected remaining target strength.

How to read these statistics

Expected value

The long-run average if this same attack were repeated many times. It is useful for comparing options, but it is not a promise that the next roll will be close.

Use it to compare average damage between weapons or targets.

◇

Median and percentiles

The median is the middle outcome. P10 and P90 form a planning band: roughly low-side and high-side outcomes for this attack or forecast.

Use these when you care about realistic ranges, not just averages.

↯

Reliability and spread

Standard deviation describes how spread out the outcomes are. A high-spread attack may have the same average as a reliable one but fail more often.

Use this to identify swingy plans before committing resources.

☠

Kill chance

The probability that the current attack removes the target based on its current wounds, models, or hull points after saves and post-save rolls.

Use this when the target must be gone now.

○

Zero effect

The chance that the attack causes no unsaved damage at all. This is the “nothing happens” risk.

Use this to avoid plans that can completely bounce.

⊕

Overkill

Damage beyond what was needed to remove the target. High overkill can mean the weapon is powerful but inefficient into this target.

Use this to avoid wasting high-value attacks.

Exact distribution

Calculated directly from dice probabilities when the situation is simple enough. This is preferred when supported because it gives the full outcome distribution.

The bar chart shows exact one-round damage outcomes.

⚂

Monte Carlo check

A local random simulation that repeats the same attack many times. It is useful as a sanity check and for future complex mixed-state cases.

Simulation results should be close to exact results when both apply.

Weapon users / participating carriers— Defender current pool— Forecast rounds—

Exact damage distribution

X-axis: exact unsaved wounds or hull points caused. Y-axis: probability of that exact result. Bars at or beyond the target pool count as kill outcomes.

At least N damage—

Cumulative damage odds

This answers “what is the chance I do at least this much damage?” It is usually the clearest chart for kill thresholds and planning.

Math trace, assumptions, and simulation

—

Turn Forecast

Turn-after-turn planning dashboard

This repeats the current attack over the selected number of rounds. It is a planning view, not a full game-state simulator.

Kill-by-turn and survival curve

The rising line is cumulative kill probability. The falling line is the chance the target is still alive after each repeated attack round.

Expected remaining strength with uncertainty band

The center line is expected remaining wounds/models/HP. The band shows P10 to P90 outcomes; a wide band means the plan is swingy.

Round-by-round forecast table

Round-by-round projection from repeating the current attack. Highlight follows the forecast-round slider.
Round	Kill by turn	Still alive	New kill chance	Expected damage	Expected remaining	P10	Median	P90	Threat left
Run a matchup to populate the forecast.

Export

Local downloads only

Exports are generated in the browser. Army imports are local file reads, not server uploads.

About

Static and editable by design

uHammer is a static browser workbench. It does not include accounts, server uploads, a database, official art, official logos, copied datasheets, No official rules text, or a locked official roster database. The starter Space Marines and Orks armies are editable examples for demonstrating realistic army math workflows.

The current statistics engine supports exact binomial-style single-round damage distributions, cumulative probability curves, repeated-round survival projections, and a local Monte Carlo check. Unsupported tags are preserved and shown as assumptions rather than silently applied.