We are back, finally, and part of the plan to kickstart racing in the UK is a disproportionately high volume of early meetings scheduled to take place at Newmarket, headquarters of flat racing in this country, writes Jon Shenton. At the current count, no less than 21 days of racing are planned in Suffolk between now and the end of August; that’s a meeting, on average, every five days or so. It’s high time, then, to give Newmarket some overdue attention in order to try and unearth some of its secrets.

Newmarket Courses

Newmarket has two distinct tracks, the Rowley Mile and the July course. They are to all intents and purposes separate entities, although they both share ground in longer distance races where they join the Cesarewitch course.

The maps below should give a more effective picture of the arrangement.

The Rowley Mile

The home of the 2000 and 1000 Guineas, scheduled for this weekend. It’s a gun-barrel straight track 1 mile 2 furlongs in length, where it joins the shared strip for more extended distance races. Perhaps its key characteristic is “the dip” a furlong from the winning post, where a downhill undulation develops into a stiff uphill finish, placing an emphasis on stamina. The course is exceptionally wide and is generally considered a fair test, in theory at least.

 

The July Course

Host of the July Cup, this course features the 'Bunbury Mile' straight track, again dovetailing with the shared straight adjacent to the A14 for distances greater than a mile. The course map below doesn’t do it justice, but a signature feature of the July track is the stiff uphill finish over the final furlong or so which, similarly to the Rowley Mile, often sorts the wheat from the chaff.

 

Newmarket Racecourse: Trainers to note

The first staging post in these articles is usually analysing trainer performance at the track. It often gives a way into more nuanced findings.

The below table shows the record of all stables which have had at least 100 runners on each of the Rowley and July courses at SP’s of 20/1 or shorter from the 2010 season onwards. The info is divided into a total Newmarket performance section (combined Rowley Mile and July), followed by the individual course records. The table is sorted by the Total Newmarket overall A/E.

I thought it important to illustrate the performance by individual track, essentially treating them separately. That was partly to satisfy my own curiosity, but primarily to establish if there was any discernible variance in a yard’s record between the two that might justify further examination.

Rather underwhelmingly, there isn’t much in terms of the individual course data to get stuck into. Perhaps Mark Johnston's and Richard Fahey’s strong July course records are worth a second glance. Likewise, Roger Charlton’s impressive Rowley Mile data too. Broadly, though, the performance is much of a muchness across the twin tracks.

Mick Channon

Based on the total Newmarket data, the UK No.1 for the second consecutive article is Mick Channon. There may be some related contingencies at play here, however, as in my last edition I commenced with a focus on Newmarket juvenile racing! You can read that here.

Analysing the Channon yard data in more detail, I have some interest in his three-year-old runners, enough to warrant a quick perusal anyway. Performance is a little sketchy over the last three years meaning angle/data confidence is at the lower end of the spectrum. Nonetheless, here is quite simply his record with three-year olds at the track since 2010:

 

And again, equally as simply, if runners are omitted that started at a price greater than 10/1, the following picture emerges:

His 2019 record was 0/5 and there was only a solitary winner in each of 2017 and 2018, so it’s not an approach to follow with your last pound - what is? However, a relatively short-priced Channon entry is probably a good start for a shortlist.

Charlie Appleby

Moving on, as also mentioned in the preceding article the original intention was to analyse Charlie Appleby’s course record with Juveniles. It didn’t quite happen due to some very alluring John Gosden stats! However, this time there is no shiny object to divert focus.

The linchpin of Godolphin trains at Moulton Paddocks, within Newmarket itself, and the overall record at the track of his runners is mightily impressive. Appleby has clocked up 129 victories from 525 runners and is marginally profitable to back at SP, too. Arguably, performance on the Rowley Mile is slighter stronger than the July Course, but not strong enough to solely focus on one over the other.

I'll begin with a consideration of performance by race class, an attribute that regularly provides a pathway to developing stronger angles, and it is no different here:

The numbers clearly show that yields from elite level (Class 1) racing suffer in comparison to the lower-class events in terms of win rates, place rates, P&L and A/E measures. Whilst in angle terms we can exclude Class 1 racing from any systematic approach based on the numbers above, there is a cautionary note which requires observation; namely that the yard is steadily improving at this leading level. The graph below illustrates Appleby’s winning numbers in Class 1’s at Newmarket as well as at every UK course (including Newmarket) from 2013, when he assumed the licence from the disgraced Mahmood Al Zarooni.

2019 notwithstanding there is a clear indication of incremental improvement. With the 2000 Guineas odds-on favourite Pinatubo in the ranks, those Class 1 wins are likely to grow, at least according to the current market. So, whilst excluding Class 1 Appleby runners from a data and angle perspective is a pragmatic move based on historical records, this does not mean that Pinatubo is an unwise wager next weekend. Indeed, it makes no comment either way on the matter!

Evaluating Appleby’s Class 2 or lower runners at HQ by SP brings in further optionality as to how best to utilise the data:

Broadly speaking, all these Appleby prices are potentially worth following from a value perspective. However, I’m inclined to play this one around the 13/2 cut off or lower mark to keep things ticking over, hopefully without too long between drinks.

I’ve checked these runners for further insight and sharpening in terms of race types, ages, days since last run and a myriad of other attributes but in truth the yard delivers consistently across all variables at HQ.

It’s not rocket science, as they say, but by backing the charges of Charlie Appleby at HQ in Class 2 or lower racing, reliable returns have been garnered. Here is the record of all SP 13/2 or shorter runners from the yard. It is profitable in every year aside from a small loss in 2014.

This year may be different given its unique Covid-infused nature, but it’s a very solid angle which should continue giving up some value all things being equal.

Aidan O’Brien

I fully expected the doyen of Ballydoyle to feature heavily on the trainer list. However, surprisingly (to me anyway), he hasn’t had the century of runners on the July Course to qualify for inclusion on the overall trainer data table. I’m hopeful he’ll get over it.

However, APOB still merits microscopic focus at Newmarket, where his record is exemplary; and I think I’ve found an aspect of it which demands closer scrutiny.

The table above represents the yard's 'all in' Newmarket record from the 2010 season to present. Backing every runner from Ballydoyle at SP would have resulted in a 10% return at SP and a whopping 37% at BSP. Go figure.

As may be expected from such a powerhouse stable, there is a large focus and concentration of runners, wins and overall stellar performance in Class 1 events:

And by drilling down further into those Class 1 races into their individual Pattern status there is even more to ponder upon:

Based upon these numbers, it would strongly appear that the greater the competition, the sharper the performance of O’Brien horses. Perhaps at this elite level the increased quality of rivals ensures that some value can be attained by backing the O’Brien contingent; maybe it relates to second-, third- and even fourth-string entries possessing the requisite ability to 'pull rank' on better-fancied stablemates.

It’s not one for the wise guys but, nevertheless, blindly backing all Group 1 and 2 entrants trained by O’Brien at Newmarket since 2010 produced 41 wins from 191 runs and would have netted a roughly £85 profit to a level stake at starting price, or a 45% return on investment if you prefer. The exchange SP ROI is closer to 85%.

Duty leads me to point out that there are three winners in the sample at 25-1, which certainly puts some fizz into the numbers. All the same, to be in the black by backing arguably the premier trainer in Europe is not to be sniffed at. This probably works as an angle in its own right.

However, there is a possible downside to this approach: as mentioned, often the yard has numerous entrants in these top-class events, which would result in several wagers in the same race. Mulling these multi-runner jamborees opened a further potential way to play, at least theoretically.

How many times does a seemingly second, or third, fourth or fifth string horse from the yard deliver the goods on one of these big days? Always seemingly obvious after the event, too.

Testing a hypothesis that bounties may just be greater where the O’Brien money is split rather than focused on a solitary runner appeared to be a worthwhile exercise. Here are the numbers for O’Brien’s Group 1 and 2 entries at Newmarket by the number of Ballydoyle runners in each race.

I'm pretty interested in this. Solo representatives from the yard undoubtedly perform well, with 24% winning and 40% placing and there is limited damage on the profit and loss front. However, where there is a multi-O’Brien entry in a Group 1 or 2 race at Newmarket, there appears to be value in backing all of them. The annualised split of these mass runner Newmarket accomplishments is detailed below:

Notably, the volume of races with multiple runners is increasing in recent times. The column “Races” represents how many individual events have had more than one runner from the yard. So, for example in 2019, O’Brien had more than one horse in nine G1 or G2 races at Newmarket, with a total of 25 horses running in those nine races.

I checked the results individually of all 2018 and 2019 races from the table above and pulled together the view below:

The column key is:

“AOB run” - number of runners in a race from the O’Brien stable alone
“Tot Run” - the total number of runners in the race
“AOB best” - the best finishing position of a Ballydoyle runner
“AOB2” - the second-best placed AOB runner and so on (AOB3, 4, 5)
“BP AOB” - the SP of the best-placed AOB horse, with winning SP price in bold.

The profit and loss data is compelling with a super roll call of horses in the mix to boot. In 2019 Magna Grecia, Hermosa and Ten Sovereigns cleaned up in the G1’s for example. I’m intrigued by this. Often, I’ve thought that following O’Brien second string (or other strings) entries may be an enjoyable pursuit and, in this case, it seems to have foundation.

The Achilles heel of this premise is that it does not translate to any other track in the UK. Multiple Ballydoyle runners don’t add up at Ascot, Doncaster, Epsom, York or anywhere else for that matter. That does beg the question why it occurs at HQ alone. It could be course configuration, Newmarket being a focal point for the yard, a quirk of the numbers, or a multitude of other potential reasons.

It’s not the most secure approach to wagering, but personally I will be tracking and dipping into this angle in a small way over the coming months. At the very least it’s quite interesting and it might be fun. Of course, the 2000 Guineas will be a test, a multiple O’Brien entry competing against an odds-on Appleby hot pot in Pinatubo. Hmm.

Pace and Draw on the Rowley Mile

Sadly, there are no quick wins (in my view) when it comes to analysing the draw at Newmarket, when focusing on the Rowley Mile at any rate.

However, before addressing that, under normal circumstances utilising the Pace Analyser and Draw Analyser tools is an ideal platform to perform detailed analysis on any course. Sadly, there is more of a challenge where Newmarket is concerned given that the pace and draw data is combined for both the Rowley and July courses.

Where there is a will there is a way, however, and in this instance the Query Tool can be used to gain all the data and intel required, with a small caveat.

With only one or two exceptions, action on the Rowley Mile occurs in April, May, September, October and November. The July course is deployed in June, July (duh!) and August almost exclusively. Thus, by using the month filter in QT, draw and pace indications can be obtained for each individual course. The only cross-pollination I can see since 2011 relates to May 2014, when there were fifteen races on the July course at the end of the month. The net result is that Rowley pace and draw data is slightly 'contaminated' by July course info. It’s relatively trivial in nature, but need highlighting.

Draw and Pace over 7-furlongs on the Rowley Mile

I have evaluated most race distances on the Rowley Mile course and it’s fair to say that any leverage from stall position is hard to find. As the course is wide, and the rail and stall positions change with regularity it is difficult to build up a consistent picture. I spent some time trying to split the draw outcome based on starting stall situation on the track (far side, middle, near side), eventually concluding that I was guilty of looking for something that was not there.

Pace is a slightly different story though as we’ll see.

The below table shows the IV3 ratings, manually derived and calculated from QT data for performance by stall position and field size. I won’t explain what IV and IV3 are in this article as Matt has done a much more comprehensive and clearer version in this blog post recently.

 

The draw data in the left section, using actual stall position (accounting for non-runners), perhaps illustrate a marginal benefit to a stall berthing on the wings, certainly in larger fields. The advantage of being located on the outer could vary by stall position on the course. For example, if the stalls were situated on the far rail, then it is not inconceivable that a low draw* may be slightly favourable and vice-versa. There is nothing solid to recommend here though.

*A low draw position is on the far side of the track based on the standard TV camera angle, obviously a high draw relates to the near side (demonstrated to full effect in last years 2000 Guineas by Magna Grecia)

The section to the right of the table contains IV ratings for Pace. That section clearly illustrates that early speed is an advantage over this distance. Given the vast expanse of the course, it may have been expected that hold up runners would have performed better, given that bad luck in running should be less of an issue. However, the early pace in general holds up. It’s seemingly an uphill battle to make up ground over that stiff final furlong or so.

 

Pace over 6-furlongs and a Mile on the Rowley Course

To check that applies over other distances the same process has been applied, for pace only, to two of the other trips raced along the Rowley Mile.

 

The pattern is the same, early speed plays an important role across all field sizes and race distances form six furlongs to a mile. Identification of the pace profile of the race would appear to be of primary importance, certainly when compared to stall position.

 

July Course Pace Data by Distance

The July course pace composition is of a similar nature to that of its Rowley Mile counterpart. Being in the front rank early is pretty much always advantageous. The stiff finish again makes it tough to make up significant amounts of ground in the latter stages, hypothetically anyway. The tables are unambiguous, not just in terms of early leaders but also in the herculean task most midfield and hold up (depending on race distance) types must overcome.

Newmarket Punting Pointers Summary

There's no doubt that both the Rowley Mile and July courses are difficult to grasp in terms of the use of data to construct wagering strategies. My own main takeaways are that for once I’m not going to stress too much on the draw aspect of races. Building a picture of pace will form the starting point of race analysis on either track for me; unless, of course, there is an Appleby runner in a non-Pattern race, or a multiple AOB entry in a Group 1 or 2, in which case I’ll just back those and move on!

Enjoy the racing, it’s good to be back.

- JS