Individual Performance Factors - Systems specific to one named jockey

Quite often, our query may originate simply when watching a race or hearing a piece of horse racing news or gossip - in any of these cases the best way to begin is to navigate to the jockeys record. Once viewing, you will see any rides today and some recent stats - as we aim to delve deeper press the yellow Profiler button, which takes us to a breakdown of jockey performance in 8 different tests (Track, Going, Distance, Age, Race Quality, Odds Range, Favourites and Trainer).

There are a couple of things to be weary of initially. First ensure a reasonable sample size (strike rate figures are a lot less meaningful when there have only been a handful of similar rides) and second look for consistency (if the profit figure is really high yet the strike rate low, it may well mean a long-shot is distorting the picture). If you are on to something take a note of it for the next step! In the example I am looking at (Greg Fairley) it bounces off the page how profitable he is riding for Mark Johnston.

As I do want to check this out more thoroughly, I press the 'Run Query' button taking me to the System Builder with my jockey details already set. I know I want to add in Mark Johnston and therefore use Quickset Trainers (remember if you know the criteria, Quickset is much much quicker) and add him in.

The massive capabalities of the System Builder provide so many options but with a system of such a simple nature, I normally run two or three spot checks dependent on its type. As this is a trainer/jockey system I'd want to know if the stats are heavily impacted by a few different things (to breakdown use Fullset, eg. to see how the Fairley/Johnston partnership performs by track I'd use Fullset track). I always check Fullset Dates first though as consistency and recent performance are key to my own judgement - ultimately this system criteria has had its day (in my opinion) and any advantage it did have has now been realised and caught up with (7 of the last 9 months have resulted in loss), had you spotted it a few months into their partnership you would have been quids in - but that is life. So for me this one is scored off the list and its time to look elsewhere!

Combined Performance Factors - Systems open to several named jockeys

With combined named jockey systems you have to be careful, often stats can be wildly misleading if you manipulate figures to suit your own wants. For example a common approach for beginners is to go into the System Builder and break something down by Fullset Jockey and then select all the profitable jockeys. This is a mistake in my opinion as its far too random.

Another thing to be weary of is monitoring progress of the system - if there are so many parts (jockeys in this case) then you may not be optimised fully as time moves on and its harder to spot which particular element of the system is failing. One other thing to be very careful of is the criteria you use. Imagine you wanted to follow jockeyA over 5f races and jockeyB over 6f - this requires two separate systems, if you save it in one then your criteria will be jockeyA or jockeyB over 5-6f.

If you can get past the problems though there are good combined jockey systems out there and they are really easy to build. Lets have a look at how to build one effectively. So, off the top of my head - I watched racing at Leopardstown yesterday and it struck me in Ireland the same jockeys always seem to come good but some are really underrated despite their consistency (Fran Berry), particularly when they are on horses who are in a price range up to say 6/1. Therefore my initial criteria is Location - Ireland, Odds - Between Evens and 6/1 (I dont like odds on shot). Then I use Fullset and break down by jockey.

On looking at the stats my decision of the logical system rules are - I will include all jockeys with strike rate above 24% excluding the big jockeys I consider non-profitable due to public perception (this is a common rule for me and includes the likes of Spencer, McCoy & Walsh). Its important to note that a couple of those I am including have overall loss figures. I end up with 9 jockeys. I immediately run the system and the stats are good. The timeline tells me its made a very handsome profit each year except 2003 (which Im happy to accept as many of the jockeys would only have been picking up spare rides back then and would have been inexperienced).

Grouped Categorisation - Systems focusing on unnamed jockeys in preset statistical bracket

When you deal with a specific statistical bracket, it means your system is going to throw up jockeys who are performing to a certain level at any given time, this means on one day you might get the rides of Tony McCoy and on the next you wont. Again its always best to start out with a rough idea and then build as you go. For this test Im wondering to myself if a jockey who has a recent high strike rate is actually more profitable on favourites than one with a low recent strike rate?

To check I need to set my starting criteria, which is in this case simply select favourites. Then I break down by using Fullset. When I do this, I instantly spot a flaw - the stats for 100% jockeys are rubbish and as I am pretty sure they shouldnt be I need to think why. Its fairly obvious it is cause I am including all jockeys and I need to take out the ones who only had 1 or 2 rides during that period.

So now my criteria is Favourites with jockeys who had over 10 rides in the past 14 days and again I am using Fullset to show strike rate of jockeys in the past 14 days. The fifth column is the strike rate of the system and it does appear to raise higher as the strike rate of the jockey grows (first column). My decision is to investigate this a bit further, but Im going to start again and include new criteria about 40% or higher strike rate and remove the favourites definition.

My new criteria is therefore jockeys who have had more than 10 rides in last 14 days and a strike rate of 40% or higher during this period. I break this down by a few different things and have decided its maybe the wrong approach and its likely their price may actually be shortened as it will be common knowledge they are currently in a purple patch. So a new idea has sprung to mind - what about jockeys who have the ability (which can be measured using the 2 years strike rate and runners ranges) but are currently in a slump (poor recent figures)?

So off I go to check this out. The point of all this is its easy to check these things and doesnt take long and as each idea springs into your mind, write it down and in time you will definitely find angles.