as you know we have been conducting an internal analysis to find the reasons why results have not been close to our usual standard. We have been at this game for a long time but recent results are very disappointing compared to previous years.
As an example, January to December last year produced 144 points profit which included a poor run towards the end of the year. Compare that to this year where we are over 60 points down.
Before we go into what we have found here is a short message from Andrew our Chief Tipster...
"As Cleeve's main analyser I would like to offer every member my personal apologies for this dreadful run of results, I have spent a lot of time reanalysing our selections to try and understand why our form has dipped so much in the past few months.
I came to the conclusion that after a number of years of continuous profit that I had unconciously changed my approach and was becoming distracted by too many data components. I am going back to the proven formula that I used and hope that this will deliver an upturn in results in the coming months, my target is to end the flat season showing a decent profit.
Please note that I back every selection given to Cleeve members with my own money, so I am also enduring this extended losing period with some pain."
We are now going to expand a little on what Andrew means in his statement.
1. Ratings - We have looked at the rating algorithm which hasn’t changed in the period analysed, and they have proven to be as robust in narrowing the field down. We have confirmed that the strike rate for the top 3 rated has changed very little.
2. Over-analysis - with more data tools available, and more and more elements creeping into our analysis there is the obvious danger of over-analysis. Some of these elements are needed like draw bias, pace stats and class moves, but their importance may have been overly weighted.
3. Value – There is a myriad of explanations out there for what value is. We are strong proponents of value and one of our criteria is we don’t back horses under 2/1. What we have found in going back over many race write-ups is that the concept of searching for optimum value has now gone too far.
To give an example. We backed Nahaar each way at Ascot in the Silver Wokingham at 14/1 and he came 3rd. He then came out again recently in a Class 2 Handicap in which he was clear top-rated but we didn’t select him.
Here is the excerpt from the selections post that day...
Nahaarr is clear top-rated and would have been the selection but his price is two points short in my view. He was our selection in the Silver Wokingham where he ran out of puff and finished third. I do think he is the most likely winner, but at 5/2 in a competitive field, he does not offer value.
The fundamental mistake that has been made is that if we think Nahaar is the likely winner but is too short then we should be 'NO BET', not looking for an alternative at a bigger price. In this case, we backed Dazzling Dan who came 5th and we lost two points rather than having no selection in the race.
4. Psychology - as Andrew stated in his message he has been selecting and backing winners consistently for over 30 years. Even for a hardened punter, having a long losing run will skew judgement. This can subconsciously change your approach and lead to looking for bigger priced selections to get back on track... having the added responsibility of keeping members happy can only exacerbate things.
The changes we are making taking into account the points discussed above
1. We will be staying with our original ratings algorithm which we know works.
2. We have pared back the weightings on the new elements that we've introduced.
3. If we feel that the most likely winner is not value, we will have no selection
4. We are going back to how things were done a year ago.
We don't expect results to instantly turn our season around but we do expect to show demonstrable and consistent improvement so we chip away at our loss and return to profit before the season ends.