Last week, British-trained horses received a 'doing' the like of which had never before been witnessed. The Irish minority rode, almost literally, roughshod over the vast numerical superiority of the domestic defences in a manner that suggested this was more than a mere perfect storm.

One leading Irish trainer has mooted that the root cause lies in British racing's infatuation with high value handicaps, but that feels wide of the mark. Others argue that the Irish are better at 'plotting one up': even if that's true, the extent to which they outmanoeuvred both the BHA handicapping team and the British training ranks also feels somewhat of a convenient pigeonhole.

No, as always, the answer is likely to be far more nuanced than 'this' or 'that'; more likely a combination of elements which have been brewing for some time. To understand what went wrong this time, a spot of historical context is required. Let's start with the most basic of barometers, the UK vs Ireland tally for the last five Cheltenham Festivals.

 

Trainer location of winning horses, Cheltenham Festival 2012-17

Year	Races	UK	Ire
2012	27	22	5
2013	27	13	14
2014	27	15	12
2015	27	14	13
2016	28	13	15
2017	28	9	19

 

This chart tells the story rather more succinctly:

 

 

In terms of pure winners, Ireland has been improving its tally significantly since 2013, and actually only enhanced their win score by four this term. That, of course, equates to an eight race swing and the smallest number of prizes for the home team ever.

But win samples are typically small, however, and this one is restricted to just 28 (27 prior to the introduction of the mares' novices' hurdle last year) races. So what of the place data?

 

Trainer location of placed horses, Cheltenham Festival 2012-17

Year	Places	UK	Ire
2012	91	58	33
2013	90	53	35
2014	92	56	36
2015	92	48	44
2016	93	48	45
2017	94	53	41

 

Here's the chart for the place data:

 

 

Notice how there is convergence in the place data but not the overlap of the win graph? This is significant because it suggests that the emerald dominance of 2017, while hardly a surprise, has been magnified somewhat by the microcosm of the winners dataset.

[Incidentally, I prefer places to percentage of runners beaten because, aside from the challenges of quantifying non-completions, many horses are eased off significantly when their chance has gone, thus further muddying what is already at best translucent water]

Before moving on, let us also consider the number of placed horses as a percentage of the number of runners from UK and Ireland. This obviously requires us to know the number of runners from each 'country' taking part, which gets interesting. Check this out:

 

Placed horses as a percentage of runners (right hand columns)

Year	Runners	UK	Ire	Places	UK	Ire	UK%	Ire%
2012	483	356	126	91	58	33	16.29%	26.19%
2013	464	355	106	90	53	35	14.93%	33.02%
2014	487	363	121	92	56	36	15.43%	29.75%
2015	468	321	146	92	48	44	14.95%	30.14%
2016	492	346	143	93	48	45	13.87%	31.47%
2017	488	325	160	94	53	41	16.31%	25.63%

*there have been a few non-UK/Irish runners as well, hence the small disparity between total runners and the UK/Ire aggregate

 

In case you missed it, let me help you out:

1. The home team had a higher percentage of their horses placed last week than in any other Festival in the sample.
2. Ireland registered its lowest percentage of placed horses to runners in the six year sample period last week.

 

Why? Simple. Ireland had their biggest raiding party since 2012 (at least), and Britain had very close to its smallest defensive battalion, 2017's 325 only surpassed by 2015's 321 (spread across one fewer race).

The graph of places as a percentage of runners looks like this:

 

 

In terms of the numerical strength of the Irish team, between 2012 and 2014 their runners amounted to circa 25%, against a British squad of 75%. From 2015 to 2017, that quarter to three-quarters was more like a third to two-thirds. Last week, Irish runners accounted for 32.8% of the entries, their highest figure as a percentage of runners in the sample, and fully ten per cent more in absolute terms than any other year (160 versus their next largest team of 146, in 2015).

So it may actually be the quantity as much as the quality of the Irish runners that is responsible for their huge margin of victory in everyone's favourite pointless contest, the Betbright Cup.

Why?

We now join the ranks of the hand-wringers to ask why the Irish are winning more Cheltenham Festival races. As noted above, the question doesn't relate solely to the most recent renewal, but to each one since 2013. What has changed during that time to bring about such an upturn in Irish fortunes? Let's consider three possible contributory factors:

- Prize money

- Handicap ratings

- Purchase price / source of acquisition

Prize Money

Willie Mullins posited over the weekend that perhaps owners want to have horses trained in Ireland due to the greater prize money, and because of the lesser programme book reliance on higher value handicaps. A quick review of last week's winners lends some credence to Willie's mullings: of the 19 Irish-trained winners, eight of them by my reckoning - Special Tiara, Supasundae, Sizing John, Yorkhill, Nichols Canyon, Let's Dance, Penhill and Rock The World - are owned by 'Brits'.

But with the exceptions of exiled Americans, Susannah Ricci and Mrs Rowley-Williams (now moved back to US), owner of Special Tiara, the others all have horses trained in Britain as well. True, the Wylies seem to be phasing out their Paul Nicholls team, but this looks more in the Gigginstown vein of performance-based decision-making rather than as a result of prize money, though a case can certainly be made for the latter...

The below table shows the five year prize money accrued by four of the top owners to have split their teams across UK and Ireland (figures derived from ownership data at RacingPost.com).

 

Owner	Ire Prize	Ire Runs	Ire £/Run	UK Prize	UK Runs	UK £/Run	Differential
Ricci	£4,262,102	545	£7,820	£2,462,702	116	£21,230	2.71
Potts	£1,644,110	456	£3,606	£769,725	50	£15,395	4.27
Wylie	£1,910,689	174	£10,981	£1,701,885	209	£8,143	0.74
McManus	£8,960,364	4238	£2,114	£7,994,949	2669	£2,995	1.42

 

Although there is unquestionably some 'cause and effect' as a result of these owners having won at Cheltenham, that's precisely why they're included in the table. The 'Differential' column shows that, while the Wylies won only 74% as much from their UK endeavours compared with their Irish portfolios, Teams Ricci and Potts did much, much better with their British teams.

But probably the best barometer of this line of argument is JP McManus. Ol' Green n'Gold supports racing to a huge degree on both sides of the pond, and it can clearly be seen on which side his bread is best buttered. McManus' UK contingent net him 42% more per run than his Irish legion.

The fact is that Willie Mullins has performed incredibly well - peerlessly, in fact - at the Cheltenham Festival for a number of years. That success brings 'overseas investment', regardless of whether there are valuable Graded pots or handicaps in the run of things. Indeed, owners like Ricci are on record as saying that they are not interested in winning outside of Cheltenham in March, a week which is the alpha and omega of their involvement in the ownership game.

So whilst there is some smoke to Mullins' contention, it seems unlikely there is much in the way of fire generating those plumes.

Handicap Ratings

More interesting, perhaps, and going beyond the handicap races, is the allocation of handicap ratings. Much has been made - before, during and since the Festival - of the re-assessment of Irish horses for British races. The consensus beforehand from the Irish camp was that this was unjust. With the raiders claiming seven of the ten handicap prizes, there is less crabbing now than before, but the question remains: why were the Irish horses largely elevated from their domestic perches?

The answer may lie not in the errancy of the Irish handicapper's work, but perhaps in a general overstatement in the British figures. Put another way, it may be that the British horses are rated too highly by the BHA 'cappers rather than the Irish too low by theirs.

To be brutally honest, I struggled to think of an effective (and time-efficient) method to test this hypothesis, and so will leave it as a question that others of appropriate informational means may crunch and confirm/refute the suggestion.

I definitely have a 'feeling' that some horses, especially in the two mile divisions, both hurdle and chase, have been significantly over-rated. Such conjecture should have no place in a pseudo-empirical article, so I'll leave it at that.

UPDATE: I've been made aware of two articles from last year covering the inflation in UK ratings. This one is from Simon Rowlands, and this one from Kevin Blake, are both excellent corroboration of the perception which, it seems, is more than that.

Purchase Price / Source

One thing that fascinates me, as a jealous owner peering through the windows into the Tattersalls Cheltenham sale and the like, is how purchase price and source impact on Festival prospects. As more largely untested stock changes hands for north of £300,000 a head, is there any evidence of a correlation between purchase price and performance in the Cotswolds in March? Or are the winners arriving in the hands of their owners by other means than public auction?

To evaluate this, I looked at the winners of the last six renewals of each of the Champion Hurdle, Champion Chase, Stayers' Hurdle and Gold Cup. That's the same time frame used above for the UK / Ireland comparisons and gives us 24 horses - minus multiple winners - to look at. Remarkably, the only multiple winner in the period was Sprinter Sacre, whose story is an interesting one to which we'll briefly return shortly.

Of the 23 individual winners of the four main Championship races since 2012, 15 were acquired privately. The remaining eight including two home-bred's - Synchronised and Coneygree, both Gold Cup winners - and six purchased for or by their current owners at public auction.

The highest price paid at public auction for a winner of the Champion Hurdle (one), Stayers' Hurdle (two), or Gold Cup (three) was the £75,000 Jim Culloty (on behalf of Dr Ronan Lambe) gave for Lord Windermere.

This year's Gold Cup winner, Sizing John, was bought as a yearling for just €16,000, Thistlecrack cost €43,000, and Bob's Worth (RSA and Gold Cup winner) was a mere £20,000. Using 90p to €1 as a conversion metric, the six Championship winners sold at public auction averaged at £32,717. The median was £24,100.

We also know something of some of those acquired privately. For example, we know that Champion Chaser, Sire De Grugy, was bought for €50,000. And it is reputed that Sprinter Sacre, who won two Champion Chases, was part of a 'job lot' of 22 horses purchased from France for €300,000. While it may be unwise to apportion that price tag equally across the whole draft, we do arrive at a figure of €13,636, or £12,272 using the 90p/€1 conversion principle. For us small-time syndicateers there is something comforting in such mathematical folly.

Perhaps Cole Harden is worth a mention, too. He was led out not sold at £30,000 after winning his debut bumper. Acquired privately soon after, it is highly possible that the purchaser paid in the region of £35,000 given that the auctioneer will usually 'phantom bid' up to just below the reserve price.

It seems that only fools rush in via the sales ring and, although the auction houses probably don't want to admit it, they appear to be doing considerably better than purchasers from these multi-hundred thousand pound/euro deals over jumps: most of the best horses are either bought privately or snapped up for relative pennies.

In Summary...

There are a number of key takeaways from the data posted in this article. Probably the hardest to swallow is that Ireland actually under-performed against their numerical representation this year, in spite of 'winning' 19-9 in terms of race victors.

The natural selectivity of Irish runners - it's a long, expensive journey for a horse with no chance - is also a factor, though this year was one where expense was waived in favour of 'having a runner' more than ever before. This was supported by those higher Irish handicap ratings, meaning more of their horses actually got a run than would have been the case of their domestic pegs.

Tully East (Ire 133, UK 138), winner of the Close Brothers Novices' Handicap Chase, was the most notable beneficiary as his Irish mark was insufficient to make the cut for the race.

There is unlikely to be anything material in the Mullins line about British fascination with a handicap-driven programme, certainly if the major owners are anything to go by. But I'm fascinated by the evidence published by Messrs Rowlands and Blake around potential inflation in UK handicap ratings: it looks like there may well be something in that.

And if you love the idea of owning a Cheltenham Festival champion, it would appear that your best chance is to either a) acquire privately, either from France or from a small stable out of an Irish bumper; or b) buy a relatively cheap ticket at the sales and hope that your luck is in!

So here's to next year, when I expect Ireland to have less winners, perhaps significantly less on the evidence of their overall performance rather than merely the microcosm of the winners' enclosure.

Matt