Understanding the ELO score
The ELO score lets you compare riders' levels over time, even those who have never raced against each other, by adjusting after every race.
In summary
Every race becomes a chance to track your progress. The more encounters you accumulate, the more faithfully your score reflects your level relative to all the other riders in the community.
ELO, a score on a scale
In practice, ELO is a simple number that places each rider on a common scale: the higher it is, the stronger the level. On Floatelo, the scores of the ~310 riders on record range from about 500 to 1350, around a median of ~980. A few real benchmarks:
Benchmarks computed from each rider's latest ELO score (~310 riders). Median: half the riders are above, the other half below (≈983). Top 10%: score exceeded by only 10% of riders (≈1120). Best: highest current score (≈1348).
~700Bottom of the pack
~980Median
1000Start
~1120Top 10%
~1350Best
Benchmarks based on each rider's latest score (May 2026). What matters is the relative position of riders compared to one another.
The general principle of ELO
The ELO system is named after Arpad Elo, the physicist and chess player who invented it. It is widely used for chess, tennis, video games… and now for onewheel.
It is not a fixed ranking but an evolving rating method: points are added to or removed from riders based on the results of each race.
The basic rule
You beat a rider stronger than you → you gain a lot of points (and they lose some).
You beat a rider at your level → you gain few points (and they lose few).
You beat someone weaker than you → you gain almost nothing (and they lose almost nothing).
Illustrative figures. The number of points exchanged follows the ELO formula: it depends on the level gap. Beating a stronger rider (unlikely win) earns a lot; beating a weaker one (expected win) earns little. The loser loses as much as the winner gains.
You1000
beat
Stronger1200
+24 pts
Beating someone stronger earns a lot.
You1000
beat
Weaker820
+6 pts
Beating someone weaker earns almost nothing.
As you race more and more, your score becomes a faithful reflection of your real level.
Practical consequence: taking part in a low-level race won't artificially boost your ELO. Taking part in a very high-level competition won't unfairly lower your ranking: you are simply compared to riders stronger than you.
The essentials
What is it for?
Comparing riders with one another over the long term, regardless of which races they enter. It is a representation of riders' relative level.
How does the score change?
- A beginner rider's score moves fast: the system is still looking for their true place.
- An experienced rider's score is stable: it accurately reflects their real level.
Illustrative curves (fictional riders), plotted within the real range. The starting score is 1000; the green line is the riders' real average (~970). A beginner varies a lot at first (low reliability), then stabilises once their score reflects their true level.
Example: a beginner's score (dashed blue) swings then stabilises; an experienced rider's (solid blue) barely moves.
What is taken into account?
- Qualifying times — mainly to give newcomers a first score (low reliability).
- Head-to-head races (pools, semi-finals, finals) — where riders truly compare themselves.
The onewheel adaptation
On Floatelo, we distinguish two types of performance: qualifying times and races between riders (2 to 8 riders per pool).
1. Qualifying times — useful but not very reliable
A qualifier places a rider based on a time set alone. It's a good starting indication of their speed, but it doesn't tell the whole story about head-to-head level, since the stakes are lower. They mainly serve to give a base score to riders not yet ranked.
→ These results receive a low reliability.
2. Head-to-head racing — the true measure
Races pit riders against each other in comparable conditions with high stakes. It is the most solid measure of real performance.
→ These results carry a higher reliability than qualifiers.
3. Reliability grows as the competition advances
The further you go, the more decisive the encounters and the more representative the result is of the level.
Reliability measures how representative a result is of the level. A timed qualifier (no duel) counts for little; the weight increases each round (round of 16, quarter, semi, final), because the opposition is tougher and the stakes higher. The higher the reliability, the more the result counts in the ELO calculation.
| Stage |
Reliability |
| Timed qualifiers | Low (no head-to-head) |
| Head-to-head qualifier | Fair (multi-rider head-to-heads) |
| Round of 32 / 16 | Fair (structured head-to-heads) |
| Quarter-finals | Strengthened (more even level, higher stakes) |
| Semi-finals | High (top level, maximum pressure) |
| Final | Maximum (most representative result) |
4. The result: a fair and representative score
By combining qualifiers and races, Floatelo produces a ranking that favours the most comparable and most meaningful performances.
The advantages
Universal comparability
Comparing riders who have never raced each other becomes possible.
Evolving stability
Volatile for beginners, stable for experienced riders.
Independence
A score that takes all races into account, independent of the rules specific to each race or league.
Accuracy
A reliable ranking, based on a recognised mathematical methodology adapted to the specifics of onewheel.