Posts

How to Use the Profiler Tab

The second of three articles requested by readers was on the subject of the Profiler tab, which we introduced to Geegeez Gold racecards this time last year. So, in this post, I'll show you how it works and outline some use cases, including how you can get around the inflexibility of 'today's race conditions'.

I've recorded a video for those who prefer to watch/listen, and below that is a written version of the same (or very similar content). These links will take you to any specific section you might be interested in.

 

What is Profiler? When to use Profiler Profiler: Beware! Example Use Cases

 

 

What is Profiler?

Profiler is a racecard tab that breaks a horse's (or jockey's or trainer's or sire's) form down by a range of race condition variables. In the example image below for a horse called Raha, we can see the FILTERS block at the top, common with our Full Form tab, and, beneath that, Raha's Going performance profile. All other variables are closed in the image, to highlight what is available to users.

The cyan coloured row with red edges highlights today's race condition: in this case, the going for today's race is good to soft.

 

WHEN should I use Profiler?

Profiler undoubtedly works best when there is a good amount of evidence in the form book. Its value is in compartmentalising that evidence such that patterns are easily visible. As such, it is typically more useful when researching the profiles of jockeys and, in my opinion, especially trainers and sires.

The example below shows trainer Paul Midgley's profile for going, distance and class.

Ignoring the cyan rows we can immediately see that he is almost a pure trainer of five-furlong sprinters, and his record at that trip is some way above his record over longer. We can also see that his all-weather form is not as good as his turf form; and we can see that his Class 1 and 2 record is inferior to lower class races, in win and place percent terms at least.

This is a five furlong race, so we can use the 'Distance' filter in the top block to narrow our analysis down only to that range. And I've also selected 'Hcap' as that is where the majority of Midgley's horses ply their trade. Let's see how the figures look now, for five furlong handicaps:

His record on good to soft or faster turf and, contrary to earlier perception, in higher class races is excellent. Note that the strike rate in Class 2 is much lower but the ROI is positive, just.

Things to be wary of with Profiler

As with all data sets, including Query Tool results, we need to be sensible and challenge the output rather than blindly accept it because we've been seduced by the black figures in the P/L column. In the example above, it makes sense that a relatively unfashionable trainer like Paul Midgley would be profitable in better class 5f handicaps while only maintaining a moderate strike rate in such company: as bettors, we have to be comfortable with long losing runs if we wish to play 7% hit rate trainers; it's not for the faint hearted, or those whose bankroll may be short stacked!

Outlying percentages or profit figures should always be treated with deep scepticism. Why is this singular condition so profitable? Does it make sense that, for example, Midgley 5f handicappers would show a positive ROI on good to soft ground particularly? Answer: almost certainly not, but his 5f 'cap runners' form on all turf going have a broadly aligned place percentage. That's far more reliable in the round.

Which leads me on to my next point: place percentages are more reliable than win percentages. Percentage of Rivals Beaten (PRB) is more reliable again and we'll look to add that to Profiler at some point before too long.

HOW Should I use Profiler?

So that's what Profiler is, how it works, and a couple of things we need to be careful about. But how should we use it? While you'll quite probably have your own ideas about utility, allow me to suggest a couple. These take the form of either knowing more about today's race, or creating shortlists/QT angles/tracker entries. Let's begin with the latter.

Horse Profiles

For the past two flat seasons here at geegeez.co.uk, we've undertaken a community project to unearth a collection of horse profiles. You can view the full list, along with any qualifiers, here. And if you want a route map for researching horse profiles for yourself, check out this blog post.

These profiles can only be generated when a horse has run plenty, so the list of better class all-age handicaps in the second post above is instructive; but you might also run a query in Query Tool to get a list of Class 5 all weather six-furlong handicaps, for example.

Trainer Profiles

We've already looked at this to some degree with Paul Midgley above. But what if we want to test a specific scenario not linked to today's race conditions? Let's say we wanted to examine the widely-held contention that Venetia Williams-trained runners go well in deep ground over marathon trips; but alas Venetia has no runners today. Then what?

First port of call is the search box on the racecards and, having identified the person (or horse) we wish to investigate, click the arrow by their name to see if there are any upcoming entries:

 

In this case, Venetia has an entry later in the week. So I click on that entry to go to the racecard for it:

 

Hmm, it's a shorter distance good ground novice race. Does that matter? Not necessarily. I go to the Profiler tab, RESET MY FILTERS (important!), and then select the TRAINER button and Venetia's runner:

 

 

In the above, I've also chosen handicaps only and 'All NH'.

Now, looking at win or place strike rates in the going section can be misleading for two reasons. Firstly, races on bottomless ground generally have fewer runners than races on 'terra firmer'. And secondly, as punters, we need to know what the money impact is, so that's the ROI column.

As it happens, in this case, the ROI figures align with the win/place strike rates in that they are more positive. ROI for heavy ground is NOT positive, but it is closer to break even than for the majority of other going categories. The same is true for longer distances in the main, though not extreme distances (beyond 3m4f), although samples are small.

But the fact that this race is at 2m2f and on good ground is a bit of an obstacle to what we want to achieve in looking at deep ground and longer distances. The solution? Go to the Full Form tab for Venetia and find a recent race in which she saddled a runner on heavy ground over a trip. Here's Profiler for Royale Pagaille's astonishing performance in the Peter Marsh in January:

 

Note that I've selected all NH handicaps on heavy ground. We can see that those parameters were unprofitable in the last five years (the 'Heavy' line in the Going column). But looking at longer trips, it is indeed the case that there might be a small bit of juice in Venetia's heavy ground handicappers over trips from two-and-three-quarter miles or so, and up.

Sadly, though not remotely surprisingly, changing the date range to the last two years - something you absolutely should do, compare longer-term with shorter-term time frames - reveals the edge has evaporated:

 

But I don't want to leave you without after all this, so one Miss Williams edge that remains is in small fields:

 

The above is the two-year view of Venetia Williams' heavy ground handicappers (chase and hurdles) in small fields (2-7 runners). In such races she's been consistently profitable over one, two and five years. I was curious as to why, and it looks like most of them go from the front (group 4), or close to the front (3) - thanks QT!

 

[The 'null' run is where the run style could not be deduced from the in running comment]

Sire Profiles

Sires can be profiled in the exact same way as trainers above; and, as with trainers (and jockeys and horses), we can create either tracker or QT Angle entries - with associated notes - to be alerted of our profiles.

 

Unexposed Runners

A second sire use case is when trying to understand more about a horse having its first run, or its first run under different conditions. We may not be able to know how that individual horse will perform but we can get an idea of what might happen based on all runners with at least a 50% common lineage.

Here's Our New Buddy, stepping up two furlongs to a mile and a half tonight on her second handicap start:

 

She's been beaten far enough - 12L, 11L, 9L - in her most recent three runs for many to overlook her chance. But should they be so dismissive? Let's find out:

 

In Profiler, I've selected Our New Buddy and SIRE (New Approach) and Flat/Handicap/Age (3)/Distance (1m4f).

The trip should be a positive and, at this distance, progeny of New Approach have done well in lower level handicaps (see Class 3-6). I've included field size and weight as examples, though I'd not be getting too stoked about them as supporting ballast.

More credible is the bottom row of the Trainer Snippet intel on geegeez, see image below:

 

Our New Buddy may get whacked again this evening, but there are reasons to believe she can show more for the extended range and for the second try in handicap company.

As with all horseracing data - and indeed information in any other sphere - nothing is alpha and omega: we need to get as much awareness as we can in order to make the most informed decisions we can. Profiler is one more tool in our kit bag to that end.

Good luck,

Matt