Last updated 16 October 2026
SnusBattle does not make objective quality assessments of products. The ranking shows what users have chosen when two products were placed side by side. It is not a health recommendation and says nothing about how products compare in terms of risk.
You see two products and pick one. The winner stays and faces another product of the same product type. The challenger is selected at random while recently shown products are avoided and comparisons across different brands are favoured. Each comparison is called a battle. After five consecutive wins, the winner retires from that run and a new battle begins.
The ranking uses Elo, the same system used to rank chess players. Every product starts at 1500.
Before each battle an expected result is calculated from the difference between the two ratings:
expected = 1 / (1 + 10^((opponent rating − own rating) / 400))
After the battle both ratings are adjusted:
new rating = old rating ± 32 × (1 − expected)
The winner gains, the loser loses the same amount. The factor 32 determines how fast ratings move.
| Situation | Effect |
|---|---|
| A favourite beats a low-rated product | Small change — the result was expected |
| A low-rated product beats the favourite | Large change — the result was unexpected |
| Two evenly matched products meet | Around 16 points change hands |
A product with few battles has an uncertain result. Public product lists require votes from at least three different visitors. The 30-day ranking also requires at least five battles during the period. Battle and voter counts are shown so you can judge the basis of a position.
| List | Based on |
|---|---|
| All time | The Elo rating, all history since launch |
| Last 30 days | Win rate during the period, minimum five battles |
| Most contested | Products whose all-time win rate sits closest to 50/50, minimum ten battles |
| My list | Only your own votes, stored in your browser |
| Champion of the month | Win rate calculated separately for each calendar month; previous months remain available |
| Brand battle | Share of battles won against other brands; battles within the same brand do not count |
| Countries | Win rate among counted votes assigned to each country, with the same minimum requirements as the public product lists |
The more people vote, the more the lists move. A position is a snapshot, not a final result.
One visitor voting many times should not be able to steer the list alone. Each visitor can give the same product at most five counted wins. Further votes are saved but do not affect the public ranking. A product only appears in public product lists after at least three different visitors have voted on it. The number after 👤 shows how many visitors have contributed counted votes.
A pseudonymous browser identifier, rate limits and pseudonymised technical signals are used to reduce repeated or automated voting. These protections cannot guarantee that every vote represents a different person. Battle and voter counts are therefore shown openly.
Nicotine content, flavour and format are product data and do not affect the ranking. The ranking is based solely on how users voted.
Votes cast through a shared benchmark link are kept separate and never affect the public ranking.