Racing Alpha v1.0 Review: Tough Day but George Wickham Delivers

A Testing Day at the Office

Yesterday was a reminder that even the best models can have quiet days. Racing Alpha's v1.0 card covered 40 races across six courses — Wolverhampton (AW), Beverley, Ffos Las, Leicester, Downpatrick, and Killarney — but only one of the six settled selections found the winner's enclosure. The overall P&L for the day was -2.75pts, a result we take on the chin.

The Headline: George Wickham Strikes

The standout performer was George Wickham, who obliged at the advised 3.25, returning a profit of +2.25pts. That win prevented the day from being a complete washout, and it's a reminder that when the model does find one, it can be at a workable price. The full detail of every selection is logged on our public performance ledger, where you can see the raw numbers — no spin, no filtering.

The Losers: Honest Assessment

Of the five that didn't fire, none were close enough to point to bad luck. Sanditon (13.50), A Daughters Love (8.60), Ghost Story (10.50), Ritz Plan (5.40), and Morell River (6.00) all failed to justify their ratings. Each was a legitimate selection based on the model's analysis, but on the day they simply didn't run to expectations. That's racing. We publish every race the model rates, not just the fancied ones, which means days like this are part of the deal.

Why Full Card Matters

Racing Alpha's philosophy is to lay out the full canvas — every race the AI system assesses, not a curated shortlist. That means you see the misses alongside the hits, and you can judge the model's performance over the long haul rather than cherry-picked results. A single day's -2.75pts is a blip, not a trend, and we'll be back to flat the model against fresh cards today.

Forward Look: Today's Racing

The model is already crunching the data for today's fixtures. We'll have a full v1.0 card covering all UK and Irish meetings, with the same transparent approach. Head to today's racing page to see the selections, prices, and the AI's confidence ratings. No guarantees, just analysis — and we'll be honest about the outcome, whatever it brings.

As always, the aim is to provide a consistent, data-driven edge. Yesterday didn't deliver much profit, but the process remains sound. One win from six is part of the variance; we'll keep grinding.