Who can use it: Proform Premium members.
Data Export is the desktop Exporter without the desktop: runner-level CSV files of UK and Irish racing from 1 January 2018, with the columns you choose, built for you in the background and delivered as a download link. It is part of Proform Premium. RaceMetrics Pro includes the on-screen tools but not Data Export.
What a file contains
One row per runner, and nothing is ticked when you arrive — the file carries exactly what you choose. The first group, Standard columns, holds the ones most people want: date, time, course, horse, card number and race type, and on a results export the finishing position, whether it won or placed, the speed figure, the starting price, the Betfair SP and the Betfair place price, along with the race title, the number of runners and the lengths beaten. They do not count towards your 150. After them come the columns you picked, grouped the same way as the System Builder's filters: race conditions, course and going, the horse (its form string among them: the last ten finishing positions going into the race, as on the racecards), market and prices (the Power Odds among them: the price the Power Rating makes each horse), Proform ratings (the Main and Main+Stats ratings among them), RaceMetrics ratings, RMP model prices, connections, records at the course, distance, going and class, breeding, and up to four runs back (last time out's official rating and Power Rating among them). Race conditions include the speed figure class par, the par the Proform software sets for each race. The System Builder's Yes/No filters export as Yes/No columns (Handicap, Maiden/Novice, Pattern/Listed, first-time headgear of each kind, first run for a new trainer and so on), a race's classifications export as one column listing every classification it carries, and the live ranks (evening, breakfast and morning price rank, weight carried rank) are ranked within the race over the runners in the file. A rating rank (official rating, PFR, Main and the rest) is the Proform software's own: every runner in the race is ranked, one without the rating at the bottom, and joint ratings share a rank. A declarations file covers declared runners with pre-race fields only: no results, no starting prices, and prices as they stand when the file is built. The ranks come with it: every rating rank, the weight rank and the evening, breakfast and morning price ranks, ranked across the declared field.
Take only what you read. A column you leave unticked is not written and not even read from the database, which is worth more than it sounds: the race title alone is over a third of the size of a file with a couple of other columns, because it averages 55 characters on every row. Everything you tick, standard or otherwise, is listed in the box above the groups in the order it will appear in the file, and each can be removed from there. Start again clears the lot. An export with nothing ticked is refused — a file of join keys is not an export.
Every file ends with three columns: RaceKey, RunnerKey and ExportRef. The two keys are stable, opaque identifiers: the same race or runner has the same key in every RaceMetrics export, so a daily file joins to a history file perfectly. They are not Proform ids. ExportRef is the reference of the export the row came from; every file is traceable to the account that requested it. Those three are the one part you cannot switch off, because they are what lets the files join to each other. A README inside the zip lists the columns you chose and the date each one is available from, so blank cells in 2019 for a rating that started in 2023 never come as a surprise.
Building an export
Open Data Export from the System Builder menu. Give it a name, choose results or declarations, set the dates (or use Yesterday, Last 30 days or All history), narrow by country, race type, handicap or class if you want, and tick columns by group; up to 150 in one file. Estimate rows tells you how big the file will be and whether it fits your allowance before you queue it. A small export comes back within seconds, ready to download on the page. A bigger one runs in its own queue in the background, reading two weeks at a time so it never gets in the way of anyone else and standing down for the morning results processing — a set window of 05:30 to 08:30 GMT, which is 06:30 to 09:30 while British Summer Time is in force, so a big export queued inside it waits and starts as the window ends. Exports of a day or two, and every daily file, are not held up by it. How long it takes depends far more on how many columns you pick than on how many rows you asked for: a few hundred thousand rows with a handful of columns comes back in under ten minutes, while the whole history since 2018 with eighty columns takes a little over two hours. You get an email when it is ready.
Yesterday's results and today's card in one file
A results export stops at yesterday and a declarations export starts today, so on their own neither spans the two. Tick Include declared cards on a results export and you get one file with both — your date range, plus every card already declared up to two days ahead. It is the same tick the Proform Windows Exporter has, and it works the same way: one date range, one file.
A declared runner has not run, so on those rows every column that describes a result is blank: finishing position, runners, won, placed, lengths beaten, the speed figure, SP, BSP and the place price, along with any other post-race column you picked. Everything known before the off is filled in as normal, and the rank columns rank each race by its own rule — a declared race on the weight a horse will actually carry, a resulted one on the weight it did. Sort on the date to separate the two halves.
Save that export as a daily and you get one file each morning covering yesterday's results and the cards through to two days out, rather than a results file and a separate declarations file.
Entries: the races before their declared cards
Choose Entries instead of Results or Declarations and the file holds the entries for every British and Irish race from today to six days ahead that has no declared card yet, one row per horse. Tick the stages you want (5-day, 4-day, 2-day, overnight), narrow by country, race type, handicap or class if you like, then pick your columns: the race (stage, class, distance, going as published, prize, eligibility, maximum runners, whether the race was divided), the runner (status, draw, age, sex, weight with the penalty shown on its own, official rating, headgear, trainer, jockey and owner, with the trainer and jockey also as the entries feed writes them), the RaceMetrics Ratings as they stand today for the horse and its six connections plus the Combined Score, breeding, and the last run. A horse taken out stays in the file, marked Taken out in the Status column. A race whose declared card is already out is in the declarations, never in both.
Tick Changes file to add a second csv listing every change since the entries were published, in plain words. Preview five rows to check your choice, then queue it: an entries file is small and comes back in seconds. Saved with Save + run daily, a fresh entries file arrives at 4pm every afternoon, today to six days ahead. An entries file carries no sectionals or tote.
Daily files
Save an export with Save + run daily and two kinds of file follow on their own. Each day's results, with your columns, once the morning results processing has finished. And each declared card, with the pre-race columns, as soon as it is published, which is about two days ahead of the race, so Thursday's card arrives on Tuesday afternoon. Prices and RMP do not exist at declaration time; they come with the results file, which carries every pre-race price snapshot alongside the result. Every file arrives by email and sits under My exports. Your first files are built within the quarter-hour of saving. You can run up to two daily exports at a time and switch them off from the Saved exports list.
One saved export gives you both — yesterday's results and today's card — so you do not need a second one for the declarations. And each daily file covers that one day, whatever date range was set when you saved it: an export built on All history still produces a single day each morning, not the archive again. So take the history once as an ordinary export, then save the same thing as a daily and it keeps itself up to date. Daily files do not count against your five exports a day.
Before you queue it
Estimate rows tells you how big the export will be; Preview 5 rows shows you what it will actually look like — the first five rows of the most recent day in your range, with the columns you have picked, so you can see the figures before committing to a pull. The preview is a whole day's racing, so ranks read against complete fields; it costs nothing against your allowances.
Your chosen columns are listed above the groups in the order they appear in the file, and each can be removed from there without hunting for the tick that set it.
Small exports come back while you wait
Small exports run in their own queue, so a day or two of racing is never stuck behind somebody's full history. Queue one and it appears in My exports within seconds, ready to download — no email, because there is nothing to wait for. Bigger pulls queue as before and email you when they are done.
Picking up where you left off
Every export in My exports has a Load into builder button that puts its columns, dates and filters back into the form, so you can adjust one and run it again without starting from scratch. Saved exports have the same button.
Downloading
The link in the email opens the Data Export page, which needs you to be signed in on the same account; a forwarded link does not work. Each link lasts seven days and allows ten downloads, after which the file is deleted. Run the export again for a fresh file.
Allowances
An export is capped at 1,200,000 rows, which comfortably covers every run since 2018, and you can pull up to 3,000,000 rows in any 30 days, five exports a day and three full-history exports a week. Sectional exports have their own allowances on top: 600,000 rows in one export and 1,000,000 in any 30 days, counted on the runners that actually carry a sectional record rather than on every row of the export. The page shows where you stand. Data Export is for your personal use under the terms of use: files must not be redistributed, resold or supplied onward, and there is no trial.
Sectionals
Sectional columns (the summary metrics the System Builder uses: run-out speed, upgrade, finishing speed against par, early and late position, hard-luck flags and the like) are included with the TPD Sectionals (ST) add-on on Proform Premium. RaceMetrics Sectionals (RMS) gives on-screen sectionals only and does not include export, CSV or download. No sectional data leaves the site in any file unless you are a Proform Premium member with the (ST) add-on.
Members with that add-on can also tick Add the full section-by-section sectional record. That puts a second file in the same zip holding the complete sectional record — every section of the race with its time, speed, finishing speed percentage, the adjusted par time and par finishing speed, position and lengths behind the leader, stride length, stride frequency, number of strides and the upgrade, most with a rank within the race. It is 849 columns, one row per runner that has a record, and it joins onto the main file on RunnerKey. The columns carry the same names as the Proform desktop Exporter, so the sectionals data key describes them one for one, and they do not count against the 150 columns you may choose for the main file. Results only: a declared card has no sectional record yet.
Sectional coverage is British racing only — there is no sectional data for Ireland in any year — and it begins when each course joined: 8 of 60 British courses in 2018, rising to every British course from 2024. A blank sectional cell means the race was not tracked, not that the figure was zero. Sectionals also arrive two to three days after the race, so a daily results file built the morning after carries empty sectional columns. You do not have to do anything about that: if your daily export includes sectional data, a second file for the same day is built automatically once the sectionals have landed, named “— sectionals” and the date. It is the same export over the same day, so use it in place of the first, or join the two on RunnerKey.
Tote returns and dividends
Under Market & prices there are three tote columns: Tote Enhanced Win Price, Tote Enhanced Place Price and Tote Dividend Returned. The first two are what a £1 stake would have returned from the win and place pools, stake included, and they are carried for every runner, not only the ones that were paid — so you can compare the tote with the starting price or the Betfair SP for any horse in the race. The third says Yes when that runner actually returned a dividend.
The pools themselves come as a file of their own. Tick Add the tote dividends and the zip carries a second csv, one row per dividend: Exacta, Trifecta, Swinger, Superfecta, Quinella, Double and Treble race by race, and the Placepot, Quadpot, Jackpot, Irish Jackpot, Placepot7 and Scoop6 pools meeting by meeting. Each row gives the date, course, race, bet type, the winning selections as saddlecloth numbers and the dividend to a £1 stake. A race dividend joins the main file on RaceKey; a meeting pool has no race, so its date and course place it. There is a row per dividend rather than a column per bet because a race pays several — the Swinger alone pays on three pairings, and some meetings run two Placepots. This file counts against neither your 150 columns nor your row allowance.
Tote data runs from November 2023, so earlier dates hold none of it. It arrives with the results, in the same overnight file, so a daily results export carries the previous day's tote in its first cut — unlike sectionals, there is nothing to wait for.
Other CSV buttons on the site
Data Export is the way to take runner-level data out of RaceMetrics in bulk. The CSV buttons elsewhere (System Builder runs and qualifiers, Ratings Analysis, Ask RaceMetrics, form lines, My Card) give you what is on your screen: up to 2,000 rows a pull for Proform Premium members and 100 rows for everyone else, and a clipped file always says so. My Ratings files, which are your own ratings, are not capped.