About Racecast.io 🦄
The mission of Racecast.io is to provide race forecasts that are a pleasure to use. We use data science to improve our endurance racing models and provide the most accurate representation of how the weather is going to affect your next race.
Racecast.io began in response the Boston Marathon's famous wild weather. We believed that there had to be a better solution than burning through your refresh button hoping that the chance of rain is going to drop by 1%. Because of that we are proud to say that we are “Built in Boston“ - but in reality we are a fully remote team with day jobs. Racecast.io is a passion project.
What is the RunScore?
We've analyzed over 350,000 running activities and solicited feedback on how the weather was for running during that run. Combined with over 350 race results over the past 20 years, we've fed performance data into a machine learning model to predict how running is affected by weather conditions. This model improves every day as more data is processed. That lets us come up with a number to answer the question: “Is it good running weather?“
The forecast we display, the data we feed into our RunScore models, and weather history comes from a variety of different weather sources. Part of this is due to filling in gaps in the weather data. The other part is being aware that different forecasts and weather stations perform better or worse for certain locations. With that in mind, we analyze predictions to score every weather source for each race location we track over time. That let's us use the best source for each individual location.
Our sources, in no particular order:
If you want to help out with Racecast.io, have a problem, don't see your race, or just want to chat, do not hesitate to reach out [email protected].