Field guide
Choose the ranking mode that answers the question.
A ranking is not only a result. It is a method for making trade-offs. The right method depends on whether you care most about classes, exact order, repeated comparisons, or explicit criteria.
Tier ranking
A tier ranking groups items into meaningful bands such as S, A, B, C, and D. It answers questions like: which books are all-timers, which choices are dependable, and which ones are not for me?
Ordering inside each tier adds useful precision. The first item in A outranks the last item in A, but both still share the same broad judgment. Tiers are especially good when small differences should not pretend to be decisive.
Ordered ranking
An ordered ranking is a strict 1–N list. Every position matters, so ties must be resolved. It is best for draft boards, shortlists, itineraries, queues, and any situation where the next action depends on exact priority.
Ordered rankings are easy to read but demanding to create. A useful check is to compare every item with the one immediately above it. If you cannot explain that boundary, the order may be too precise for the evidence.
Pairwise comparison
Pairwise ranking asks one smaller question at a time: A or B? Repeated comparisons are helpful when a long list feels overwhelming or when several items are closely matched. The resulting wins and losses can suggest an order, but cycles are possible: A can beat B, B can beat C, and C can still beat A.
Use pairwise results as evidence rather than unquestionable truth. Revisit comparisons that conflict with the overall result.
Scored ranking
A scorecard rates every item against the same criteria—for example quality, cost, effort, durability, and delight. Weighted totals are useful when a decision must be auditable, but the weights encode your priorities. Publish those weights whenever the result will affect someone else.
Converting between modes
Tier → ordered
Keep the tiers in sequence, then resolve the order inside each tier. Compare adjacent items and write one sentence explaining each close boundary.
Ordered → tier
Look for natural gaps in quality or confidence. Place a tier break where the reason for choosing the next item changes—not merely after every fixed number of entries.
Comparisons → ordered
Start with win rate, then inspect contradictions and strength of opposition. Re-run the comparisons around disputed positions.
Scores → tiers
Set thresholds before seeing the final totals when possible. If every item lands near a boundary, the criteria need refinement.
A reliable ranking session
- Define the question. “Best” might mean most enjoyable, most useful, best value, or best for a particular person.
- Choose the mode. Use tiers for classes and an ordered list when exact priority has a consequence.
- Set criteria. Name the two or three qualities that matter before moving items.
- Place obvious anchors. Start with a clear high, middle, and low example.
- Resolve boundaries. Spend time where two items could reasonably swap.
- Record uncertainty. A note such as “provisional” is more honest than false precision.
See it applied
Worked examples make the trade-offs visible.
Browse three public example rankings, each with its method and commentary.