Tire evaluation metrics that we use on our website

Our website has been designed to give its users an opportunity to find tires with the best price/quality ratio. In order to achieve that, we use such key metrics as CoreScore and Popularity.

CoreScore

The CoreScore is a 0-100 score that indicates how liked a tire is based on reviews from users and results of professional tire tests from numerous experts.

Parameters, on the basis of which the CoreScore is calculated:

  • Consumer reviews on the top tire selling websites.
    We gather such data as the numbers of reviews and the ultimate score of the tire model on numerous tire-related websites such as amazon, tirerack, oponeo, simpletire, market.yandex, tyrereviews, and others.
    We discriminate between the websites that have the so-called “verified” reviews (amazon.com / amazon.de /amazon.co.uk, and tirerack.com), and all others.
    The coefficient or the “weight” of the reviews belonging to the first category is much greater.

    We have no right to display outside reviews on our website but what we can do is aggregate the data on the tire evaluation from all the relevant websites and use them for calculating the CoreScore.

  • Participation of the tire models in professional tire tests and their test scores.

  • Reviews written by the visitors to our website tiresvote.com.

CoreScore Distribution

Popularity of the model

Unfortunately, for new models there are usually too few reviews to make an objective conclusion about the quality of the tire. We solved this problem in our own way by adding yet another metric: Popularity.

Popularity is a score that indicates how widely the tire is known and how much demand there is for it. Unlike the CoreScore, this score does not have an exact maximum value, its average value for most of the tires being equal to 30. The most popular models can score 60 or even slightly higher.

Parameters, on the basis of which the Popularity is calculated:

  • Year of manufacture
    We proceed from the premise that usually a new model is more popular than an old one.
    Of course, there are exceptions to this rule but usually when a manufacturer launches a new tire, it comes as a replacement of another tire of the same line.

  • The premium brand factor. (premium / mid-range / economy)

  • The number of video reviews of the tire on youtube.com, uploaded both by consumers and the tire manufacturer.

  • The number of news items about the model.
    We constantly monitor the top websites of the tire making industry, and keep our eyes open for the news items about new sizes of tires being introduced, launching this or that tire model on new markets, the use of this or that tire as original equipment (OE) on some cars, and so on.

  • The number of times the tire was involved in professional tire tests.
    In this case, we take into account not so much the place in the tournament bracket that was scored by a tire model as the very fact of the tire taking part in such tests and rankings.

  • The fact of the tire being on the bestseller lists on such reputed websites as amazon, tirerack, oponeo, simpletire, and others.

  • The number of tire types and sizes.
    Here we proceed from the assumption that the more types and sizes a tire has, the more popular it is. Of course, this cannot always be the case, but statistically this assumption is more than justified.

Popularity Distribution

How the CoreScore is calculated

Possible value:
“0 – 100” or “Not available” if the conditions for calculating the model’s ranking are not satisfied.

CoreScore: N/A, not N/A Distribution

Calculation pattern

  1. For each of the parameters that determine the model’s ranking, the value of the parameter, calculated by some formula (which differs from parameter to parameter), ranges from 0 to 100.

  2. For each of the parameters, the number of reviews / tests / top charts is transformed into an equivalent of a measurement unit.
    Here we make sure to take into account the weights and the coefficients of every data source that we use.
    Example 1: The review and the ranking of a tire model from amazon.com will always have more “weight” than the review posted on the tire maker’s official website. This has to do with the fact that on amazon.com a review can be written only by a real person who bought this tire, while on other websites such review can be written by anybody.
    Example 2: a professional tire test carried out by the top tire industry publications as Consumer Reports or Auto Bild will always have more weight than a tire test conducted by a less reputed organization.

  3. Then we calculate the weighted average value on the basis of the values from 0 to 100 and the aggregate equivalent of the measurement unit.

  4. Then we smooth out the ranking for the models that did not participate in any tire tests and are not on the top charts.

How the Popularity is calculated

Possible value

  • 0 - 10 * (the sum of all weights)

  • in reality, somewhere from 0 to 70

Calculation pattern

  1. For each of the parameters that determine the model’s popularity, the value of the parameter, calculated by some formula (which differs from parameter to parameter), ranges from 0 to 10.

  2. The value that we get by using such formula (different for each of the parameters) is then limited by the numbers 0 and 10 if it is not within the segment.

  3. Each of the parameters has its own weight (approximately from 0.5 to 2)

  4. Then we add it to the weights.

As for the formulas for the parameters, we do not provide them here for a number of reasons.

Potential drawbacks of our metrics

CoreScore is not the ultimate truth about how good a tire is. Here are a few potential drawbacks of such method:

The tire manufacturers show particular interest for positive reviews about their product and negative reviews about other brands. They may try to manipulate the public opinion by adding a multitude of fake consumer ratings on various websites that we use in our calculations. We know about this, and this is actually a problem. This is specifically why it is very important for us to use data from the websites that feature the so-called verified purchases. In addition, we have a serious respect for the results of professional tire tests. 

As for the user reviews on our website, which we also use for calculating the CoreScore, each user can write only 1 review per tire model. In addition, we have a spam detector that tracks some certain review writing templates, which we are not revealing here for obvious reasons.

Unfortunately, our metrics depend on the data that we get from other websites. Our resources are limited, and we do not always have the time to update the information on each specific tire model (the numbers of types and sizes, news, using the model as OE, involvement in this or that specific ranking, evaluations and the number of reviews on other websites). All of this naturally affects the final scores and metrics. 

We are encouraging you to send us any useful information by using the links on the product’s page. These links are marked with a pencil icon: