Arc de Triomphe: Is the going typically "soft" at Longchamp on the first Sunday of October?

The last week of September is upon us, and trainers in France (including those undertaking a prolonged visit from Japan), the UK and Ireland are finalizing preparations for the most famous race in Europe’s autumn calendar. Many are also looking at weather forecasts nervously, mindful of their horse’s preference for softer or firmer terrain on the big day.

One of those twitching on the weather is Marco Botti, whose 6-year-old stable stalwart Giavellotto defeated longstanding Arc favourite Kalpana at Kempton a few weeks ago. That performance prompted connections to aim their horse at the big one in Paris next week…with the proviso that the ground is on the fast side.

The AntePoster was particularly struck by Botti’s comments on this subject as reported on At The Races a few days ago:

“I’ve always said that if the ground is on the soft side I definitely won’t run [Giavellotto] because we know he’s not going to show his best and it’s probably pointless to waste a run… having said that, you never know as once every 20 years you can get good ground… we’ll just watch the forecast and see what happens.”

Was Botti right in his resigned acceptance of the overwhelming likelihood of soft terrain on the first Sunday of October in the Bois de Boulogne?

The AntePoster, having attended most of the Arcs run over the most recent timeframe referred to by Botti (20 years), was instinctively sceptical, but without any hard data to back up that reaction.

This prompted us to investigate the state of the ground for the last 20 runnings of the Arc de Triomphe, taking the official France Galop description and comparing this with the ground as retrospectively inferred by Timeform based on analyses of race times over the whole card. In our view, the latter provides the only reliable description that can be equated with conventional British descriptions of going types.

The resulting data is set out in the table below, together with the name of the winning horse and final winning time for the year in question. For clarity, we have provided a shading system for each going description whereby the lighter the shade of grey, the better the going.

* In 2016 and 2017 Longchamp racecourse underwent a rebuild and the Arc de Triomphe meeting was held at Chantilly racecourse, situated some 30 miles north of Longchamp.

Findings

Analysis of this table has allowed us to come up with the following key findings, with which the reader may choose to agree or not (comments welcome).

We have included the runnings at Chantilly in our first two findings despite the distance between the racecourses and their slightly different soil types, not least because this data does not materially change the overall picture whether these editions of the race are left in or stripped out.

1. The word “soft” doesn’t appear in the majority of Arc de Triomphe ground descriptions (in UK parlance) over the last 20 years

According to Timeform, no less than 11 of the last 20 runnings have been run on ground that should not have had the word “soft” anywhere in the description.

• Six of the last 20 editions of the race have taken place on “good to firm”, more than on any other going type

• Plain “good”  or is the next most common type of going, together with “soft” (five runnings each)

• “Good to soft” and “heavy” have been the rarest going types, with two classifications apiece
  

But even if the official ground descriptions are used, there have been seven good ground Arcs since 2005 – slightly more than one in three.

2. The “real” going on Arc day over the last two decades has predominantly been at least one category faster than that described by France Galop

In keeping with what is generally agreed about French going descriptions (which are adopted verbatim by the Racing Post and preserved in their database, nota bene), the great majority of Arcs over the last two decades are described as softer in official parlance than is actually the case for the British understanding of going conditions on the day (per Timeform). The typical rule of thumb is a deviation by one “category”, i.e. “good to soft” usually means something in the region of “good” ground, while “good” typically translates as “good to firm” at the very least. In the table above, this holds true in the great majority of cases.

On four occasions (20% of the data set) the ground has been described identically by France Galop and Timeform. There is unanimity that the ground was essentially a bog (“heavy”) for the victories of Solemia – sensationally from Orfevre, breaking thousands of Japanese hearts – in 2012 and Sottsass in 2020. Similarly, both parties agree that the going was plain “good” for the victories of Sea The Stars (2009) and Enable (2018).

By contrast, there is a significant discrepancy of two categories in the going description for the triumph of Ace Impact in 2023 – “good to soft” according to the official French description but “good to firm” according to Timeform. For those who might doubt the latter’s interpretation, note that Ace Impact’s winning time of 2:25.50 was the second-fastest Arc winning time over the last two decades (excluding Found’s victory in 2016, which can be ignored in this context as it was achieved at Chantilly racecourse).

3. Testing conditions have clearly dominated for the last six years

The longer trend notwithstanding, it is notable that five of the last six Arcs have been run on testing going of one kind or another. Even after the typical Timeform “downgrade”, the going has been classified as “heavy” on one occasion, “soft” on three, and “good to soft” on another. Only Ace Impact’s victory in 2023 bucked this trend.

Quite why softer terrain has taken over the trend in recent years is another matter. It may simply come down to random meteorological variance, but some may ponder the increasing impact of climate change.

But certainly if Marco Botti had talked of a “1 in 6” chance of a good ground Arc, he would have been correct if only a very limited dataset of the last six years were the sole timeframe considered.

Going musings for the 2025 Arc de Triomphe

At the time of writing, the Longchamp ground per France Galop was “très souple”, with a French going stick reading of 4.0. This is one notch softer than the going stick measurement recorded on the day of the Prix du Prince d’Orange on 14 September, which featured the clash between Croix du Nord and Daryz. Timeform has classified that ground as “good to soft”.  So the ground right now may well be “British” soft.

The million-euro question, of course, is how much it will dry. The majority of forecasts at the time of writing point to a predominantly dry run-up to the Arc, with negligible drizzle and then sunny weather in the five days leading up to the big day. However, ten days can be an eternity in the world of meteorological evolution, and weather modelling is simply not able to capture the complexity of potential development this far out. Something may well change to bring more rain next week and halt (or even reverse) the transevaporation process, so this jury is well and truly out as to what ground Longchamp will serve up for the 2025 Arc.

That said, it’s worth pointing out that the Longchamp ground has been drier than usual over the course of 2025 generally. Given the remarkable spring and summer of 2025 in northwestern Europe, dry conditions have seen records broken and Longchamp’s marquee meetings consistently run on fast going (descriptions per Timeform):

- Prix Ganay (27 April):                                      Good

- Poulains/Pouliches (Guineas, 11 May):   Good to firm

- Prix Ispahan (25 May):                              Good to soft

- Grand Pris de Paris (13 July):                         Good to firm

- Arc trials (7 Sept.):                                          Good to firm

What can be reasonably inferred from this is that Paris has had a dry spring and summer too, and that the watering regime implemented by the Longchamp ground team has in no way been over-zealous. This is something to bear in mind as Arc day approaches, as any drying could potentially happen faster than in years where the ground has been soaked continuously by the elements.

In summary, like everyone else on this planet The AntePoster has no idea of what weather surprises may lie in store for Paris over the next ten days. It is perfectly possible that a form of “soft” ground will prevail in the Bois de Boulogne on 5 October if the weather picture changes.

As things stand, Longchamp racecourse is currently predicting "souple" for Arc day, or "good to soft" in UK speak. But based on the long-term trend of Arc day ground over the last two decades, the very dry spring and summer of 2005, and the current picture of drying terrain next week, Marco Botti might want to at least start thinking of his transport arrangements to Paris for the admirable Giavellotto. This “1 in 20” shot really could come in.