Well this is a new one. 🙂
The idea of automated betting software is almost heresy to me. I really enjoy hunting around for bets, and feel it’s both an art and science. And I have to admit, there is nothing quite as satisfying as being proved right when an unlikely punt comes home in front.
So – I suppose all that’s tantamount to a slightly embarrassing admission that the ego does play a part in betting. But anyway, moving swiftly on!….. maybe, jus’ maybe, all this betting vanity is misplaced. After all, computers do everything nowadays…. so maybe they can pick winners too?
All of which explains why I am trialling the Magic Backing Bot. Part of me really wants this to work… I mean, it would save a lot of time for a start! But another part of me hopes, for reasons I don’t wish to analyse too closely, that it falls flat on its face…
Anyway, it’s all installed and good to go. Set-up couldn’t be easier. And the first run will be on Saturday afternoon…..
Time to find out if human tipsters are a thing of the past!
OK, testing starts today. Some further reflections before we begin –
- It’s not really clear to me how the selections themselves are identified by the system: maybe it will all became apparent once we’re up and running?? There is a daily download into a selections table that rather suggests that, behind the scenes, there may after all be a human tipster working his way through today’s copy of the Racing Post with a pencil! Surely not?? 🙂
- Magic Backing Bot provides a testing mode as a sort of electronic paper trade, but I’m not really sure how valid a test it is. In particular, I want to understand at what odds we get matched: and, as far as I can see, the only way to do so is to place real bets. So…. I am breaking my own rules by risking real money on a totally unproven proposition! Honestly, the things I do for you boys! 🙂
- I intend to place small bets @ £2 per point, using the Magic Backing Bot’s own staking system. They do make quite a bit of fuss about this in their manual, though, as far as I can see, it is a pretty straightforward ‘capped ratcheting’ system. Stakes are raised during winning sequences until a maximum level is reached, at which point bets are placed at the same level until we hit a loser. We then drop back to the minimum stake and start again.
- I have to confess to more than a little nervousness at keying my Betfair login details into anything other than Betfair. What’s to stop the system authors using my card details to empty my bank account?! And yet, for some reason I’ve done it, and my bank account is so far untouched…
- There is a further hidden cost of the system in that (fairly obviously I guess) you have to leave your computer switched on all day, using power even when you’re away from home. And of course you have to suspend any ‘go to sleep’ options on your PC. None of which is the end of the world, but still needs to be factored into the overall investment case.
OK, time to switch on…
Everything crossed as ever,
Final Update and Trial Conclusions, 5/12/2011
Although the trial was shorter than I’d ideally have liked, running for a shade over three weeks, I still feel we are able to draw some general conclusions. And overall, I have to say that I have been disappointed by the Magic Backing Bot, and therefore have to consign it to my Failed Systems folder.
We actually managed to place an astonishing 71 bets in this short, intense period, and, according to the Magic Backing Bot support people, should actually have had more – but for the mysterious (and still completely unexplained) missing bets.
Amongst those 71 bets, we had 19 winners, which, at £2 a point, delivered a final system balance of-£25.92, or -12.96 points. The strike rate was around 27%, which is too low for short-priced tips such as the ones provided.
Most of my concerns were initially voiced in my comments to Paul on 22 November (see below), however I would summarize as follows:
- I don’t feel comfortable exposing my personal, financial information, and my Betfair login details, to a third party. If I am to do so, there needs to be an extraordinarily good reason, and sadly the Magic Backing Bot did not provide me with one.
- There were differences between the results I got, and the results reported on the system author’s log. In particular, the Bot inexplicably “missed” bets that should have occurred in the middle of my betting day – something which I found particularly galling when the bets in question had been on horses that came home in front.
- The Bot often secured odds below Betfair SP. As far I can tell, it is not designed in such a way as to take advantage of the long periods of time it has prior to the off to secure attractive odds – surely, a missed opportunity?
- I don’t like having to leave my computer on all day, burning power and money for no benefit. And the fact that BSP could have often delivered better odds than the Bot made me wonder why we bothered having a software robot at all! Why not simply send me a list of selections to place at BSP before switching off and going out in the morning?
- There was a general feeling of instability about the product, with many software downloads being required during the trial. The whole point of a product like this should be its ease of use, but I found myself re-loading the software so often, sometimes on successive days, that it became quite annoying.
In summary though, my biggest objection to the Magic Backing Bot is simply that I don’t understand the
point of it. I had expected to find that either the tipping or bet pricing processes would somehow exploit the power of software – but they didn’t. Instead, the tips appear to have been picked by an unnamed (and presumably human) tipster, who sometimes listed them first thing in the morning at the foot of the previous day’s performance report. And the odds we secured did not, as has been noted, appear to show much sign of ‘software intelligence’.
It has been a frustrating few weeks and I have to say I’m quite relieved it’s over. I genuinely do hope that the people who subscribe to the Bot make a lot of money, but on the evidence of this trial, I see little reason to believe that they will.
Ultimately, short trials (even those that produce 71 selections!) can only offer ‘reasons to believe’ rather than categorical proof that they work. But there is no basis for such faith here, not just because we lost money, but because of the way in which we did so. Reluctantly therefore – as I had really hoped to like the Magic Backing Bot – I have to put it into my ‘Failed Systems’ category.