The BikeRide Score
The BikeRide Score ranks each bike from 0-100 to indicate how much it is liked by users and experts. The score is calculated based on reviews from all online sources. Instead of spending hours browsing through reviews, you can use one simple score to understand how good a bike is.
How the BikeRide Score is Calculated
To calculate the BikeRide Score, we gather all expert bike reviews and user ratings that can be found online. We calculate the average of all reviews and ratings from 0-100. We then adjust the score based on the credibility and experience of each expert and the number of user and expert ratings. Low quality or biased reviews and spam ratings are filtered out. See how this works below:
Anyone can add a user rating for a bike that he or she has tried. We also gather ratings from other reliable online sources.
A bike with fewer user ratings offers a less reliable score, as it is the opinion of only one person or a few people. To prevent bikes with fewer user ratings from reaching the top of our ranking lists, we make the following adjustments:
- 1-10 user ratings: subtract 10%
- 11-30 user ratings: subtract 5%
- 31-99 user ratings: no adjustment
- >100 user ratings: add 5%
We include expert reviews that are either published on BikeRide or on other websites.
Expert reviews are converted into the scale of 0-100. For example, an expert rating of 6/10 becomes 60/100 and an expert rating of 4/5 becomes 80/100.
Experts are given a level of 1-5 based on how many reviews they have written:
- 1-2 reviews: level 1
- 3-5 reviews: level 2
- 6-10 reviews: level 3
- 11-25 reviews: level 4
- 26+ reviews: level 5
The higher the level, the more the reviews are weighted in the score calculation:
- Level 1: 100%
- Level 2: 200%
- Level 3: 300%
- Level 4: 400%
- Level 5: 500%
This adjustment accounts for the experience the expert has in writing reviews. It also serves to avoid excessively weighting a new expert, before we have had the chance to fully judge the quality of his or her reviews.
Unrated Expert Reviews
For expert reviews that do not have a rating we use the following framework alongside the experience of our writers to build a rating, so that these reviews can be included within our algorithm. This bike model is made up for demonstration purposes:
Quasi Faux-Faux 2.0 (2022):
- Build quality: 90
- Performance / Innovation: 80
- Design: 80
- Ride quality: 90
- Price / Value: 78
Average rating: 84/100. This is used as the rating of the expert review.
Quality, Unbiased and Spam-Filtered Ratings
Some bike brands try to manipulate the BikeRide Scores by adding positive user ratings to their own bikes. In order for us to avoid this score manipulation, we have built several filters to detect and filter out spam. This includes IP user restrictions and other filters that check for certain patterns. We also manually monitor all user ratings.
We go through each expert review to make sure that it is unbiased and of high quality. This is done manually. We also disregard reviews that we judge as biased. This includes situations where the focus is on promoting a bike in return for a free sample or commission, rather than on impartiality. Elsewhere, some ‘reviewers’ may write low-quality ‘reviews’ of bikes without having tested them. We have high standards. If we are not certain that an expert review is unbiased and of high quality, then we do not include it.
The Final Score Calculation
The final BikeRide Score calculation is made by weighting the expert reviews and user ratings:
- User ratings: 50%
- Expert reviews: 50%
When there are less than 30 user ratings, we weight the user ratings 25% and the expert reviews 75%. This adjusts for the less reliable score given by fewer user ratings.
We also make the following adjustments when making the final score calculations:
- If a bike is missing expert reviews: subtract 5%
- If a bike has 3-4 expert reviews, or 1-4 expert reviews and 1+ user ratings: add 5%
- If a bike has 5 or more expert reviews: add 10%
These adjustments are made because scores based on many expert reviews and user ratings are more reliable.
The color of the BikeRide Score depends on how high the score is:
- 80-100: Green
- 60-79: Yellow
- 00-59: Red
Potential Flaws in the BikeRide Score
The BikeRide Score is built to be as useful and reliable as possible for readers. There are, however, some potential flaws to discuss:
- The BikeRide score is calculated based on reviews and ratings from many online sources. When experts and users choose their ratings they consider how good a bike is, but they also consider the value you get in comparison to the price of the bike. This affects the BikeRide Score, meaning that a high score does not always mean that a bike is better than a bike with a lower score. It can just mean that it is a great bike, based on its price.
- We weight expert reviews depending on how many reviews an expert has written, as explained under “Expert Reviews”. However, the fact that an expert has written several reviews does not necessarily make that expert’s reviews better.
- Brands have a special interest in seeing high scores on their own bikes and low scores on all other brands. They may try to manipulate the score by adding extra user ratings. We have built filters to detect and delete this type of spam, as discussed under “Quality, Unbiased and Spam Filtered Ratings”.
- Experts may write biased reviews because they receive free samples, or write low quality reviews without testing the bikes first. We have high standards. If we are not certain that an expert review is unbiased and high quality, then we do not include it. This means that the BikeRide Score does not include all reviews and ratings but it does include those that we deem to be fully trustworthy.