I'd been meaning to do it for a few months. Finally, prompted by a reader (thanks Martyn C), I asked my database demon, Dave, to crunch the numbers. What numbers?, you may very well be thinking. Allow me to explain...
Geegeez Gold has eight reports in its dropdown list, and the most easy to use - dare I say it, the closest to a pure tipping sheet - is The Shortlist.
Its purpose is simple: highlight those horses whose past form indicates they're best suited to today's race conditions. Here's an example, from Sunday:
The fact that this example contains nine runners, of which six won and two more were second (one of them behind a nominated winner), is coincidental. Because yesterday I got the data back from 'Database Dave'.
The Scoring System
Before I go further, I want to divert your attention to the 'score' column away to the right. That will help to explain how horses qualify to be on The Shortlist.
In order to qualify, a horse must have run at least twice under at least three of the five matching criteria of going, class, course, distance and field size.
Green is for horses with a 33% or greater win record, and scores three points. Thus, a line of five greens (as in Chris Pea Green's and Abi Scarlet's cases above) is worth 15 points.
Amber is for horses with between 15% and 32% win records under the relevant condition, and scores one point.
Grey is where a horse has yet to race under a particular condition, and scores zero. And Red is for a horse that has raced under a given criterion and recorded between 0% and 14% wins. These score -1
The percentages for the place variant of the report are more stringent and are not within the scope of this article.
To appear on The Shortlist report, a horse must score at least nine points; and it is impossible for a horse to have fourteen points. So the eligible scores are 9, 10, 11, 12, 13 and 15.
Headline Figures: Top Rated
So it was that I excitedly - and somewhat apprehensively - opened the spreadsheet containing the historical data, running back to the end of May 2014 (just over seven months) when we introduced the traffic light and scoring system.
I had high hopes that the top rated 15 scores would be profitable, and I wasn't disappointed. In fact, remarkably, they have proved profitable at industry starting price.
I write "remarkably", because many of these types take heavy support during the day as the general public cotton on to their profiles, and the early bird will have done substantially better than SP by availing of Best Odds Guaranteed prices.
Here are the headline figures...
'Line of green' (i.e. max score 15) horses have netted 39 wins from 135 runs (28.9% Strike Rate) for 27.61 points profit at starting price (20.5% ROI). At Betfair SP, those figures rise to 47.63 units profit after commission. That's a 35% ROI.
And, like I say, BOG bettors will have outperformed those profits by perhaps as much as another 10%.
The downside, if there is one, is that 135 top rated picks in seven months is not a lot. Of course, allied to the excellent Stat of the Day and Double Dutch services, that's a handy little portfolio right there, and worth just over 200 points profit in 2014, all told.
Two. Hundred. Points. Profit.
These aren't fancy fictional numbers. They're "pop your bets on in five minutes and get on with it" real numbers, which I - and more importantly loads of you - have been deploying daily for a good while.
But I digress. Back to the sample size, which is small, granted. So what about the overall level of performance?
Headline Figures: Overall Performance of All Qualifiers
I'll be honest here. I expected this to be a big loser overall. After all, we're talking 1617 qualifiers from 30th May to close of play Sunday (when I received the data). And lots of them with Red in their profile, implying a dislike of one or other criteria.
And collectively they did lose. At SP. There was a loss of 76 points on 1617 selections at starting price, which equates to 4.7% of stakes. But, at Betfair SP (less 5% commission of course), the story was different. A good bit different.
That 76 point loss becomes a 104.77 point PROFIT at Betfair SP. With all taxes paid.
I have to be honest, I was pretty bloody delighted with that. I mean, don't get me wrong, I'm obviously a massive advocate of form profiling, and I knew the top rated picks were likely to be a winner. But all of them? That has been really pleasantly surprising.
And here's another thing. The scoring system, whilst not quite arbitrary, was a best guess of what ought to work. And work it surely does.
I reckon, based on a LOT of personal experience, that it is possible to beat SP by around 20% using the best BOG prices in the morning or the night before. (That means getting 6/1 about your average 5/1 winner, so hardly implausible, right?)
But let's just say 15% for the purposes of toning things down a bit. All of a sudden, that 76 point loss becomes a profit of 103 points - amazingly similar to the Betfair SP profit.
Before you think about going away and backing all selections, two things to keep in mind:
1. The profit and loss graph shows it's a bit of a bumpy ride (see below)
2. You'd be looking at an average of 7.35 bets a day
Headline Figures: Second Top Rated
Now this might take a bit of 'getting' so bear with me. Ostensibly, if 15 is the highest available score, and 13 is the next highest available, then 13 should be the second top rated score, right?
Alas, no, it's not as simple as that. You see a 13 will consist of four Greens (4 x 3 = 12) and an Amber (1).
My contention - and not just because the data supports it - is that four Greens (4 x 3 = 12) and a Grey (0) is preferable. But not just any grey. The most likely omission from a horse's score card is that it has never previously raced at today's course.
So, and this has been the way I've looked at The Shortlist since its inception, second top rated picks for me are those that have four greens and a grey for course.
Long and short, these are profitable to follow, and to a similar degree as the 'line of green' types.
'Otherwise all green but no prior course run' types won 30 of 107 (28% Strike Rate) for +32.38 (30% ROI). It's a small sample again, but extremely well aligned to the top rated.
At SP, they recorded a profit of 16.82 points (15.7% ROI). And using the 15% over SP BOG extrapolation, that profit becomes 31 points. Again, spot on in line with the BSP bottom line.
Other Interesting Insights
I've sliced and diced this dataset more than is probably healthy, and as a consequence I've found various things of interest (as well as a heap of irrelevance and happenstance).
By Odds Range
Here's one interesting take. I broke the entire 1617 lines down into (arbitrary) SP odds ranges, as follows:
Greater than 20/1
14/1 - 20/1
8/1 - 12/1
6/1 - 15/2
4/1 - 11/2
Evens - 7/2
And here's how the BSP profit or loss stacked up...
More than 20/1 +88.57
14/1 - 20/1 +13.07
8/1 - 12/1 -9.64
6/1 - 15/2 +0.92
4/1 - 11/2 -45.18
Evens - 7/2 +51.75
Odds On +5.27
Without getting too categorical, those are quite interesting figures, to my eye at least.
We can put them into three bands - 14/1 or bigger, 7/2 or smaller, and 'the inbetweeners'. So doing gives us this:
14/1 and up +101.64
4/1 - 12/1 -53.9
7/2 or less +57.02
I don't really know why this is, but there looks to be plenty to go at for both small staking big pot hunters (14/1+), and those who like consistent winners (7/2 or less).
I looked at all of the scoring categories, and uncovered at least one interesting anomaly. The overall BSP records of each score group was as follows:
I obviously got to thinking about this, and about whether these figures were acts of randomness, or skewed by a single fat priced winner. And, in the case of the 9's, the latter is true.
On the 1st December at Kempton Park, Pretty Bubbles shocked most punters by winning at 50/1. He returned 91.65 at Betfair SP, which rounds down to 87.07 after commission. That accounts for the entire 85 point profit in one hit.
It should be added, in the interests of balance, that the '9 score' group included 39 horses sent off at 25/1 or bigger, so it wasn't an entirely isolated fluke result. [And, happily, I know from correspondence that at least one Gold subscriber was on the Bubbles - well done, Stuart!]
The 15's did predictably well, as we've discussed; and the 11-13 range were unrewarding to blanket support. But the 10's were interesting. Really interesting.
There's only one way to get a 10 score, and it's this: 3 x Green, 1 x Amber and 1 x Grey.
I found this group, and their smoothest of all profit curves - almost linear, fascinating.
Why was this group profitable? And why was it so consistently so?
At around 10% of the total sample, it's not the biggest subset, and conclusions should be treated with caution... but here's where I got to:
My suspicion is that the combination of the known (i.e. proven against three of today's conditions), the unknown (i.e. unproven against one of today's conditions), and the less than perfect (i.e. a middling profile against one of today's conditions) was enough to inflate the average odds on such runners without especially diluting their win chances.
The table below seems to substantiate that.
We can see that in the case of the 9's and 10's, the average win SP was around the 4/1 mark (3.97/1 and 4.16/1). That's higher than the 11's, and significantly higher than the 12's or 13's.
The 15's, generally a better profile proposition, actually have a surprisingly high average win starting price of over 3/1.
Although the data supports my conjecture, it would irresponsible of me to suggest that my suspicions are anything other than conjecture, and should be treated as such.
The best bets remain - and I'm confident of this - the 'line of green' 15 types, and the 12's with no prior course form. Those fellows collectively accounted for 80.01 points profit over the course of just seven months.
That's eight hundred quid profit for £10 stakes on a turnover of just 242 bets. Add it to Stat of the Day and Double Dutch, and you have a very nice little 'low maintenance' portfolio within your Gold subscription, regardless of how deeply you delve into the other form tools, race cards and reports.
p.s. The Shortlist is geegeez.co.uk's free Feature of the Day on Tuesdays, and can be viewed here. [Note, you need to be a free registered member, and logged in, to see it. If you're not registered yet, go here to register]. Or start your Gold trial here.
If you'd like to delve deeper yourself, the link below will enable you to download the dataset (as well as my various tabs of working out). Feel free to play with different scenarios, and do report back if you find anything interesting.
N.B. Microsoft Excel is required to edit this file. It should open, however, in other spreadsheet applications, such as Open Office, which can be downloaded free here.
You can read more about what Geegeez Gold offers from these links: