Welcome to Winsmore — a project that began when I was still at school, rating horses against each other using criteria such as going, trainer, course form, recent form, speed, and more.

The original idea was simple: compare every horse in a race across multiple categories, identify the strongest in each area, and combine those rankings into an overall score.

Over time, the system evolved into a weighted model. Each winning horse had its criteria scores added to a database, allowing the system to identify which factors were historically most important in finding winners.

For example, if races were being rated on only three criteria — going, distance, and speed — and winning horses consistently ranked highly for speed, then the speed category would receive a stronger weighting in future calculations. In other words, criteria proven to be more predictive over time were given greater importance.

In the early days this process worked well, but analysing a single race manually with the Racing Post and Sporting Life could take over an hour. Today, modern technology allows Winsmore to process far richer data, in far greater detail, and at a speed that simply was not possible before.

Winsmore still follows the same core philosophy today.

Each race is analysed using a wide range of data points, with predictions generated through the scoring and weighting of factors such as going, trainer, jockey, speed figures, Racing Post Ratings (RPR), course form, recent form, and more. Every horse is ranked against the others within each category, and those rankings are multiplied by dynamic weightings generated from historical winner data. The final combined score produces the system’s prediction, shown as the BACK bet in the “Today’s Racing” section.

Unlike many tipping services, Winsmore does not selectively choose races. Maidens, novices, hurdles, changing ground conditions, first-time blinkers and other difficult race types are all analysed equally by the system.

Despite this broad approach, the results have been consistently profitable. Since the site went live, backing every BACK selection has produced a profit of +65.7 pts in May.

That said, blindly backing every selection may not be realistic for the average punter. Winsmore is designed not only as a prediction system, but also as a detailed race analysis tool.

Using the BACK bets as a guide — alongside the detailed race cards, comparison tables, jockey and trainer statistics (overall, recent, course and horse-specific), horse profiles, and other data available on the site — users can apply their own judgement to narrow down the strongest opportunities and potentially achieve even better results.

No system can predict racing perfectly, but Winsmore aims to provide a structured, data-led approach that gives users a clearer and more informed view of every race.

Every runner row is packed with information — ratings, stats, form figures and tags. This guide explains what each piece means, section by section.

These small +/- pills (e.g. SPD -1.49, DRAW +0.64, GOING -0.59, DIST -0.58) show how this horse compares to the race average on each individual factor the model scores. Green = above the field average on that factor (helps the horse), red = below average (hurts it). The pills shown are whichever factors most influenced the horse's overall score — common ones include: