What people building on the API say

Going in, my main doubt was whether a relatively new API would have enough coverage and consistency for what I was building, particularly across full Australian race cards on busy Saturdays, and how quickly genuine gaps would be dealt with once they showed up in real use. The breadth was better than I expected. Once I was using the endpoints together there was enough to build around rather than just use it as an odds feed, and while there have been gaps and edge cases, they have generally been visible rather than hidden. The most useful part has been having acceptances, results, horse data and market data through the one API; tying those together has been worth more to me than any single endpoint. The biggest benefit beyond that has been the turnaround. The acceptance-card work, a scratching issue and a few identity problems we found were investigated and addressed rather than brushed off, and when something upstream isn't working properly we are told, rather than handed data that looks complete when it isn't. I have probably pushed it harder than the average user, so I have seen both the good parts and the things that still need work. For someone building their own racing analysis or automation rather than looking for a finished tipping product, I would recommend it: it gives you a broad set of building blocks and the development has been very responsive to real-world use.

I’m testing the API alongside our existing source rather than replacing anything. My main concern was whether the API data would reconcile reliably enough with the other racing data First4 Pro uses — runner identity and matching accuracy particularly, because a silent mismatch could contaminate the analysis. The integration has been straightforward and the matching has been very good: in one validation run First4 Pro matched all 41 active runners across five races with no ambiguous matches, and when we did find an issue with missing runner references, the diagnostics gave us enough information to identify it precisely. Stable race and runner identifiers, together with market data, give First4 Pro a much stronger provenance trail and let me verify exactly which provider record was used in the analysis. What I value most is that I’m able to treat PuntersEdge as a factual data source rather than a tipster: First4 Pro does its own evidence assessment and probability work, so reliable underlying data and traceable identifiers are more important to me than somebody else’s selections. First4 Pro is still in prospective validation.

My main question was whether the data would have enough depth, granularity and stable identifiers to build a proper quantitative racing and backtesting platform around it, rather than simply being another odds and results feed. It's been much more comprehensive than I initially expected. The timestamped bookmaker price movements, stable race IDs, results, closing lines and now runner_ref give me a very clean structure for connecting market behaviour to individual horses and races. The permanent price-path history is probably the standout: being able to follow price changes by bookmaker and then link them through runner_ref to results, closing lines and horse form is particularly useful for the type of quantitative research I'm building. The responsiveness has been exceptional. I've raised questions and requirements around stable runner identity, bulk exports and Parquet, and Hamish has either explained the existing capability in detail or implemented improvements very quickly. The transparency around nulls and edge cases has also made integration much easier. The biggest positive for me has been the willingness to explain what the data can and can't do. Things like historical coverage, point-in-time limitations and the Betfair licensing restriction have been explained clearly rather than glossed over, and that matters when you're trying to build and validate models properly. For someone building their own racing analytics, modelling or backtesting platform, I think it provides a very strong data foundation without having to piece together market history, results and horse identity from a number of unrelated sources.

Before I signed up I was skeptical the odds would be refreshed often enough to be useful. I discovered they were updated well within the timeframes I needed, and I'm refreshing the markets multiple times per minute! It was so quick and easy to get started and I had immediate visibility across all the bookmakers I cared about. What I've appreciated is how simple it's been to upgrade as my usage has grown. The system has also been very reliable, and they've kept adding features I can use. Trying to build and run something like this myself would have taken months of effort and been a distraction from the actual systems I want to build. The few times I have had a question, customer support has sorted me out quickly.

RaceVision uses PuntersEdge data across its Australian racing platform, including automated race results and historical market data as we develop our closing-line analysis and model evaluation capabilities.