End of season report

 Introduction

 
A number of changes were made to the TVB service for the 2020-21 season.
 
Perhaps the most significant, was the formal splitting of the service into ‘official’ bets and ‘unofficial’ bets:
 
The official bets were all sent out by email, after 9:00 on race day - and it should have been possible for all subscribers to back them at close to advised prices (they were all in big races and issued only when the markets had matured).
They were then followed by a second email, stating the logic behind the bets.
 
The unofficial bets were all posted in the forum, at various times and in different forms.
It wouldn’t have been possible for all subscribers to follow all of the bets - but most subscribers will have been able to follow some of them (if so inclined).
 
In terms of results, then it’s only possible to assess the performance of the official bets (which were all proofed by the SBC).
 
The official bets were offered in 2 forms: Best bets & Matrix bets.
 
 
Best bets
 
The tips from previous seasons, became Best bets this season…
 
Over the course of the season (Oct 31st – Apr 10th) a total of 111 Best bets were issued across 103 races, with 61pts staked.

If you had managed to secure the advised price on all of the Best bets, you would have achieved a profit of 22.08pts (36% ROI).

If you had backed the Best bets at BSP, you would have secured a profit of *13.82pts (22% ROI).

Whilst it was a little disappointing that the profits at BSP, were less than those at advised prices, I think the reason is quite simple:
TVB has become a bit too popular ! - particularly with people who can only bet on the exchanges...
 
This became apparent very quickly, with the exchanges prices of the Best bets issued over the first 2 weekend of the season, quickly crashing - and never fully recovering.
My estimate is that over £10K gets placed on each Best bet I issue - and that is likely to skew all but the biggest horse race markets.
To compound things further, far too much money was requesting BSP. As a result, the BSP of the Best bets, was generally 10% lower than the last traded price on Betfair (so the price available in the live market, literally 1 second earlier !).
 
In fact, it’s very interesting to note that if you had backed all of the Best bets in the live Betfair market, 2 mins before the off (prices taken from Betmover), you would have secured a profit of *18.55pts (30% ROI).
 
That’s almost an additional 5pts of profit across the season - just for placing your bets 2 mins before the off.
That number is more in line with my expectation, as I do believe that most of the advised prices, are ‘fair prices’ (as they are relatively stable and in mature markets).
 
To try and counter the issue with over-betting on the exchanges, I refunded 20 people after the first fortnight - and also asked everyone who remained on the service to rein back their stakes.
I can only imagine what the BSP figures would have looked like, if I’d not taken these actions !
 
Suffice to say, the service is clearly at capacity, in terms of the amount of money that is being placed on the Best bets (any increase is likely to eradicate the service edge at BSP).
 
*BSP figures are before commission (which can be set to 2%)
 
 
Matrix bets
 
Following the success of the Matrix bets in big handicaps at the end of last season, I was keen to incorporate them into the official offering this season - the problem was how best to do it…
 
I eventually decided on a solution which saw the Matrix bets covering multiple selections in a race - but also horses which last season, would have been strong mentions or savers.
Whilst I don’t think I managed to crack the Matrix bets completely, I do think they showed significant promise - and I will be looking to evolve them further, next season.
 
In terms of results: then over the course of the season, a total of 312 Matrix bets were suggested, with 1037 units staked.
 
They weren’t really designed to be followed at advised prices (more on that later !) - but if you had taken the advised price on each of them, you would have achieved a profit of 161 units (15% ROI).
 
As 10 Matrix units equals 1 Best bet point, it can be seen that the Matrix bets didn't perform quite as well as the Best bets (16 v 22).
 
However, the 2 sets of bets were running at almost exactly the same level of profit, entering the final month of the season.
A poor April just had a bigger negative impact on the Matrix bets…
 
Somewhat ironically, my belief at the start of the season, was that following the Matrix would reduce volatility - however, the reverse turned out to be true !
I think a lot of that was simply down to luck - as the Matrix suffered some shockingly bad luck during the season (3 odds on IR losers at Aintree, was just the tip of a very large iceburg !).
 

Analysis of the Best bets
 
I didn’t record sufficient details to fully analyse the Matrix bets (there were just too many of them !) - however I did record extensive details for the Best bets and that, amongst other things, enables me to to establish whether they were particularly ‘lucky’ or ‘unlucky’…

During the season, the Best bets achieved the following placings:

1st – 16
2nd – 14
3rd – 14
4th – 14

A very even spread of finishes suggests no obvious luck bias.
 
The better measurement of ‘luck’, is how many of the tips that traded at 2.1 or less, subsequently went on to win…

This season, there were 30 horses that traded at 2.1 or less.
16 of these won – which suggests that there was ‘average’ luck, through the course of the season.
Instinctively, I feel that probably was the case - though as the in-running markets get weaker (which they are doing), that metric becomes a less accurate predictor (if a horse hits 2.1 in running, it is now more likely that it will win).

Again, as with last year, there was a fascinating spread of low trades, across the season:
Broadly speaking: the first month (or more accurately, 3 weeks) was unlucky; the next 3 months were lucky; and the final month was unlucky.
Provided luck continues to even itself out over the full season, I’ll have no complaints - I just hope that 5 months remains long enough for this to be the case (because it wasn’t for the Matrix bets !).
 
The number of Best bets issue this season, was similar to the number of tips issued last season.
I always feel there should be more, but I think I have to accept that the UK weather is invariably going to make things difficult for chunks of the season: whilst market strength/defensive pricing by bookmakers, continues to get worse.
 
In terms of the races in which I issued Best bets: then 19 of the Best bets were in races run in Ireland - they yielded at small profit (3pts).
However, that was down solely to Heaven Help Us - which is a slight concern…

Of the remaining 92 Best bets: 48 were in class 1 races (over half) – with 28 in class 2; and 16 in class 3. There were none in class 4 races or below.
The Best bets in class 1 races, broke even: whilst in class 2 races, they yielded a profit of 15pts; and in class 3 races a profit of 4pts.
I don’t think much can be read into any of that…
 
As has become the norm since I switched to ‘big races’, very few of the Best bets, crashed in price.
Slightly more had a better advised price than BSP (64 v 44).
However, I suspect a comparison of advised price with the ‘2 min’ price, would have seen a near 50:50 split.
Across the entire season, only 15 tips shortened in price by more than 40% - and, as was the case last season, none of those won (maybe there’s a system in there !)


Unofficial bets
 
The TVB forum now provides a single point for all unofficial bets/advice.
The use of sub-forums (for structure) and of email notifications - help greatly with the forum usability.
 
The switch from ‘write-ups’ to ‘previews’, was one of the big changes made this season.
 
In previous seasons, the write-ups were issued after the tips had been sent - however, timings became very tight, and that caused problems.
As a result, I switched to previews - which I produced the day before the racing and posted in the forum.
To support this, the emphasis had to change from being a justification for a selection, to being an overview of all the key runners in a race.
I’m hopeful that the previews were a useful resource: providing an insight on how I dissect races and also helping people make their own judgement on races/bets.
That said, I’ll be using the Satisfaction Survey to help me more fully gauge their value…
 
In addition to the previews, then the forum was also used to to cover the pre-season; mid-week bets, ante-post bets - and Live threads.
 
The main purpose of the pre-season, is to get me back into the groove - and also test out any changes to the way the service will work.
It was therefore of particular value this season - though the bets suggested during the pre-season didn’t cover themselves in glory.
That’s always going to be a danger, because to an extent, I have to force things in order to get the benefits (test out any new methods/procedures).
Whilst I always make a disclaimer at the time of issue, anyone following the bets in the pre-season should do so knowing that they may not make a profit.
 
I was hopeful that the mid-week bets would make a profit - and they did - probably !
To an extent, it will have depended on the prices you took (if you followed them).
At advised prices, they finished about level - but almost every advised price was beatable on the exchanges.
I would intend to continue to handle the mid- week bets in the the same way next season - accepting that the quality of the mid-week racing (I insist on class 3 or better) means that there aren’t many days on which I can suggest bets (there were just 13, this season).
 
I posted ante-post bets in 2 different varieties: Weekend and Long range.
Neither set was profitable, which clearly isn’t ideal - though that wasn’t my main issue with them…
There is always a danger of long losing runs, when you are suggesting bets at big prices (which all of the ante-post suggestions were).
A single season is never going to be long enough to fully assess whether you have an edge…
As a consequence, I’m not too concerned about the results (accepting I would have liked them to be better !).
My bigger issue was that once I’d studied an ante-post race and committed to a selection, I found it very hard to alter my view.
Part of that was because I didn’t want to change my mind on the race day (even though it may have been perfectly reasonable to do so); and part was because I didn’t want to tempt Sods Law !
The bottom line was that I personally, wasn’t convinced over the value of the ante-post bets - but again, I’ll be interested to hear peoples views, via the Satisfaction survey.
 
The final significant unofficial offering this season, was the Live thread.
This was something that I introduced at the Cheltenham festival, last season.
It was my plan to only run 2 or 3 Live threads during this season (at the major festivals) - however, due to lockdown, I ended up running them almost every weekend of the season.
In truth, it is hard for me to assess their value to people.
I know that some wanted it to be live tipping - but it wasn’t that.
I viewed it more as an extension of the TVB community: though it did also mean that I could update my views on some of the races - if certain things had changed from the morning (due to non runners, ground conditions etc.).
I'll need to mull over, whether the Live threads have a future - and if so, how that future will look.
Suffice to say, it’s another area on which I’ll be seeking feedback...
 
 
Plans for next season
 
Unlike last year, I don’t plan to make any major changes for next season.
 
I feel that overall, the changes I implemented this season, improved the service.
I didn’t implement them all flawlessly - but they were generally the right things to do, so my focus going forward will be on tuning them.
 
Splitting official and unofficial bets, worked well - with email used for all official bets and the forum used for anything unofficial.
That broad structure will remain.
 
I adopted some flexibility with the time of issue for the official bets (9:00 or 10:00 - depending on the market strength) and that also worked well.
 
What didn’t work as well as I would have liked, was the Matrix bets.
The main issue was with the big field handicaps, where the enhanced place terms offered by the bookmakers, meant that the odds on offer were significantly lower (25-30%) than they would have been in a win only book.
Next season, I would intend to quote Current Exchange Price (CP), when suggesting Matrix bets.
Those prices won’t be achievable by everyone - so I will use BSP to assess the profitability of the bets.
My hope is that in these kind of races, the drifters and steamers will offset each other, so people will meed to use their judgement on the best time to place bets.
 
For the unofficial bets, the major question mark in my mind concerns ante-post - and whether I should continue to suggest such bets.
My inclination is to drop them from the service - but I’ll wait and see if this backed up by the survey.
 
The only other significant issue in my mind, concerns how I should deal with the really busy periods.
I would define those as the times in the season, when I have to study the racing on 3 or more consecutive days (specifically, Christmas, Cheltenham and Aintree).
Simply, I struggle to make things fit during these periods, as the amount of time to study, write, tip & watch tends not to quite fit into the 14 hour daily window that I have available !
My tendency is to hold back on the tipping - but even doing this hasn’t delivered the results I would have hoped for.
I suspect there isn’t actually an answer to this particular issue - but I wanted to flag it regardless…
 

I think that covers just about everything - apart from the customary words of thanks…

Reassuringly, the guys who have supported the service for many years, continue to do so.
Key amongst them, remains Chris.
 
During the first lockdown, he created ‘Profile tables’ (featuring CSR !) - and they have proved to be a real success. They were followed, at the beginning of this year, by ‘The win Vector’, which peaked when suggesting the winner of the Irish national – at odds of over 300/1 !
Chris continues to diligently post both the Profile tables and The win Vector, on a daily basis - and also helped me with numerous random queries, throughout the season.
He already knows how much I appreciate his efforts - but just to confirm !

Also worthy of special mention, are Hayden and Craig.
They both stepped up to handle the admin for competitions during the season: Hayden, took care of the December Naps comp; whilst Craig defined and ran a brand new Cheltenham tips comp.
Both were a big success - and handled with real aplomb.
Many thanks to both….
 
So that’s it for another season.
I’ll be issuing the satisfaction survey in the next few days, and I’d be grateful if you could all complete it (it really helps me - both in assessing satisfaction levels and with the charting future direction)
 
All being well, I’ll then be back in touch in the autumn with details of what will be on offer for the 2021-22 season.
That will be my tenth season - it barely seems possible !
 
Wishing you all a really good summer :)

TVB.