Decision tables and DMN: the rules side of a decision engine
The essentials in 30 seconds
- A decision table lists conditions and actions in a grid with 4 quadrants.
- 3 Boolean conditions produce 8 rule combinations.
- DMN is a standard from the Object Management Group with a graphical notation and an expression language.
- Tables apply agreed policy, and the EQOS Asia engine scores 9 dimensions between 0 and 1 (EQOS Asia method page, as of October 2026).
- A table can gate cases before scoring.
A decision table lists conditions and actions in a grid, and DMN is the standard notation that places such tables inside a model of linked decisions. Together they form the rules side of a decision engine: the part that applies agreed policy consistently. This guide explains both, shows a small worked table, and marks where scoring begins. It builds on the comparison of rules engines and decision engines.
On this page
Compare the EQOS Asia offers and request a 30-minute pre-diagnostic.
What is a decision table and how is it laid out?
A decision table is a grid with 4 quadrants. Conditions sit on one side, actions on the other, and each rule is one combination of condition values paired with the actions to perform (Wikipedia, Decision table, consulted October 2026). The format lets policy experts and developers work from the same page. Tables are described as easier to understand and review than code.
A reader scans a rule from top to bottom: if these conditions hold, perform these actions. Every rule sits in plain view. Teams use tables for pricing tiers, approval limits, eligibility and routing, all of which depend on a handful of clear conditions. The grid format makes gaps visible, because an empty cell shows where a rule is missing.
How many rules does a decision table need to cover every case?
A table with Boolean conditions needs 2 to the power of n rules to cover every combination, where n is the number of conditions. Wikipedia’s printer troubleshooter has 3 conditions and therefore 8 combinations (Wikipedia, Decision table, consulted October 2026). Tables that cover every combination are called balanced or complete, so every situation has a prescribed action.
A supplier approval table with 2 conditions needs 4 rules. Is the contract above the approval limit? Is the supplier a local entity? The illustrative figure lists the 4 outcomes. Each added condition doubles the grid, so teams keep conditions few and merge cases where a condition does not matter, marked with a hyphen.
What advantages do decision tables bring to business rules?
Decision tables make policy visible. A compliance lead can read the approval limits without opening code, and a developer can implement the same table. Wikipedia lists shared use by developers and policy experts as an advantage (Wikipedia, Decision table, consulted October 2026). When the policy changes, the team edits one row and records who agreed the change.
Tables also support testing. Each rule is a test case with an expected action, so a team can check a table by running every row. Finding conditions that do not matter for a rule lets tests be skipped and streamlines the table. Reviewers gain a compact artefact to sign off, which suits regulated processes.
What is DMN and who maintains this decision modelling standard?
DMN stands for Decision Model and Notation. The Object Management Group publishes it as a standard approach for describing and modelling repeatable decisions within organisations (Wikipedia, Decision Model and Notation, consulted October 2026). The OMG specification page presents 2 tools: a graphical notation and an expression language (OMG, DMN specification page, consulted October 2026).
The graphical notation shows the dependencies between related decisions, together with the business knowledge and the input data needed to make them. The expression language defines decision logic in enough detail to allow execution, and it includes decision tables among its features. Wikipedia names the language FEEL, A Friendly Enough Expression Language. The standard makes decision models interchangeable between organisations and tools.
How do decision tables fit inside a DMN decision model?
A DMN model arranges decisions in a network. A top decision, such as approve a supplier, depends on sub-decisions such as check eligibility and set the approval limit, and each sub-decision draws on input data. A decision table is one way to define the logic of a node. The diagram shows how the pieces connect, and the table shows how each node decides.
This division of labour helps teams. Business owners read the diagram to see which decisions exist and what feeds them. Specialists maintain the tables behind each node. Because the notation is standardised, a model built in one tool can be reviewed in another, and the vocabulary stays stable when staff or vendors change.
Where does the rules side of a decision engine stop?
Tables answer one question: given these conditions, which action applies? They work best when the policy is agreed and the conditions are crisp. A question about how a regional hub performs on talent, regulation, cost and resilience at once has no crisp threshold, and writing it as a table would require a huge grid or a long list of exceptions.
A scoring layer takes over at that point. The EQOS Asia engine measures a situation on 9 structural dimensions, returns a φ(S) score between 0 and 1, and flags a STOP zone below 0.618 (EQOS Asia method page, as of October 2026). The scoring layer weighs the factors, and the rules layer applies the policies that depend on the resulting zone.
How can a decision table feed a decision engine?
Use the table as a gate before the score. A table can check that a proposal has the documents, the approvals and the eligibility that policy requires, and send incomplete cases back. Only complete cases move to scoring. The engine then spends its effort on the structural question, and the table keeps the process orderly.
After scoring, a second table can map the zone to an action, such as proceed, review or pause. The simulator offers a first score, the documented method explains the 9 dimensions, and a free 30-minute pre-diagnostic by video, scheduled in SGT, helps design the split between rules and scoring (EQOS Asia contact page, as of October 2026). This page is general information, not legal, tax or investment advice.
See also in this guide
Frequently asked questions
What is a decision table?
A decision table is a grid that lists conditions and actions, where each rule is one combination of condition values paired with the actions to perform (Wikipedia, Decision table, consulted October 2026). It is easier to review than code and lets policy experts and developers share one artefact.
What is DMN?
DMN, Decision Model and Notation, is a standard published by the Object Management Group for describing and modelling repeatable decisions. It combines a graphical notation of decision dependencies with an expression language that includes decision tables (Wikipedia, Decision Model and Notation, consulted October 2026).
How many rules does a decision table need?
For Boolean conditions the grid needs 2 to the power of n rules, where n is the number of conditions. A table with 3 conditions has 8 combinations, and each added condition doubles the grid. Teams merge cases where a condition does not matter to keep the table short.
Does the EQOS Asia engine replace decision tables?
The two work at different layers. Tables apply agreed policy to crisp conditions, and the EQOS Asia engine scores a situation on 9 dimensions between 0 and 1 (EQOS Asia method page, as of October 2026). A table can gate cases before scoring and map the resulting zone to an action.
Test your knowledge
1. How many quadrants does the basic decision table layout have?
- 2
- 4
- 8
Show the answer
4.
2. How many combinations do 3 Boolean conditions give?
- 6
- 9
- 8
Show the answer
8.
3. Which organisation publishes DMN?
- Object Management Group
- A national bank
- A software vendor
Show the answer
Object Management Group.
4. What does the DMN graphical notation show?
- Dependencies between related decisions
- Server addresses
- Brand colours
Show the answer
Dependencies between related decisions.
5. Below which φ(S) value does the STOP zone begin?
- 0.900
- 0.618
- 0.300
Show the answer
0.618.
About the author
Stéphane Jambu
President, EQOS Forecast Asia
Stéphane Jambu leads EQOS Forecast Asia and works as a consultant in e-commerce, search visibility, and AI. LinkedIn profile
Sources: EQOS Asia method page (eqos-asia.com/method.html), simulator page (eqos-asia.com/simulator.html) and contact page (eqos-asia.com/contact.html), en.wikipedia.org/wiki/Decision_table, en.wikipedia.org/wiki/Decision_Model_and_Notation, omg.org/spec/DMN, consulted on 22 October 2026. Page updated on 22 October 2026.