In the renewed Estimated Transfer Value (ETV), set to release in February 2025, we have given the model a substantial upgrade to align it with recent developments in the transfer market and ensure that our ETV model remains the industry standard for predictive player value modeling.
Estimated Transfer Value (ETV)
The Estimated Transfer Value (ETV) is a Machine Learning (ML) model that learns patterns from thousands of past transfers. By training the ML model on previous transfers, it learns to accurately estimate the ETV of every player in our database based on features like the SciSkill, Potential and contract duration.
Market Developments
In the past decade, transfer fees paid on the market have significantly increased right until the start of the COVID pandemic, after which they temporarily plummeted before rising to record heights in the past two years. The price development can at least partially be attributed to global economic inflation, but also to other factors like increasing investments and commercialization in the sports industry as well as upcoming markets that are willing to pay substantial fees for top players.
User Feedback
Over the past two years, our ETV model has gained popularity in our Recruitment Center application as well as through our partners footballtransfers.com, and has increasingly been adopted by users as key performance metric in their workflow. Thanks to its wide usage, we have received feedback on various cases and applications.
Impact
Due to the market developments and user feedback described above, we decided to retrain the ETV model on the newest data, and upgrade the model with an improved feature set to better align with user feedback. As a result, we managed to reduce the prediction error in terms of predicting actual fees paid on the transfermarkt from ±25% to ±10%, thereby significantly outperforming both the existing model as well as for example the market value as available on transfermarkt.com.
In terms of user functionality, this release will not have any impact at all, as the ETV functionalities across our products will remain unchanged. We do however anticipate a substantial impact on ETV values as can be seen in our products. Effects for individual players will depend on specific conditions, but in general one can expect the following generic patterns to emerge:
Values will increase overall due to market inflation.
Increased ETV for young high potentials: These players are highly valued by the market and as such they will see a steep rise in ETV.
Decreased ETV for older players: Despite their quality, players in their 30 represent little investment potential as they will lose their value over their remaining contract years.
Decreased impact of contract duration: While contract duration is important, its effect has been somewhat reduced thanks to user feedback. As a result, you will see less sudden drops & peaks in ETV value due to contract changes.
Increased responsiveness: The transfermarkt is sensitive to hype and player form, which can be most easily quantified in recent minutes played across competitions as well as SciSkill development. By focussing on player minutes and development in the last 6-12 months, we have made the model more responsive, and as such you will see quicker rising and falling of ETV in case of injuries, form changes etc.
In line with maintaining the same user functionality, you will also retain the ability to filter based on our Estimated Transfer Value (ETV).
Implementation
The renewed ETV model will be released across all our products on February 1st, 2025. To prevent sudden changes in a players ETV graphs, we will retrospectively introduce the new model from 12 months ago, and will gradually change a players value to be represented by the new model over the course of the past year. As such you will also see changes in the historic ETV of players over the last 12 months.
Continuous Improvement
As the transfermarkt is constantly developing, and economic inflation actually causes prices to rise with 0-10% per transfer window, we have to make sure the model remains aligned with the market. Therefore, the model is automatically retrained and updated after every transfer window, to include the last information and account for changes on the market.
To gain a deeper understanding of the various components involved and their impact on the Estimated Transfer Value (ETV), we recommend reading this article.
Explore the various use cases of the renewed Estimated Transfer Value (ETV) by reading this article.