I’ve documented my attempts to back every tipsters tips from our tipping competition before and the results weren’t very impressive. Contrary to the advertised returns, this was of course because I chose to back the tips automatically using betfair and the conclusion seemed to be that the profits that we see listed on our bulletin board are only achievable with the bookies. I noted then that only placing bets at odds of 10/1 or more provided much better returns and for a while I chose to do only this but a few periods of favourites doing well caused me to rethink this.
For a large part of February I decided to change something again and the results have been a LOT better. I initially continued submitting bets for all of the tips at exactly the same price as the tipster listed the tip at (instead of taking the best price available at the time) or it either didn’t get placed or I made a judgement call as to whether to accept lower odds later. I documented towards the end of the last update that this seemed to see a marked improvement but I still felt I was losing out especially on the selections under 10/1.
Of course getting exactly the same price with betfair is still not going to live up to the bulletin board profits because of commission. A week or two into February, I therefore decided to submit slightly higher odds than the tipster listed to factor in this commission and sometimes a bit more and was quite surprised just how many of my bets were matched. The markets on betfair are sparse the day before and overnight, and the illiquid markets are your friend when it comes to trying to get matched at generous odds but only when the stake is small.
In this case my bets are still £2 and all this reminded me of a post I saw years ago on the betfair forum which detailed the power of the £2 bet very astutely.
The initial post by Betfair forumite Aye Robot was this…
“The “interface cluedo” (not a term I agree with but never mind) cuts both ways too as one of the quirks of the BF interface is that it hands a massive inherent advantage to small bets (especially in running), which makes things much, much easier for small or new players trying to make a few quid to get going with, the £2 advantage is a far stronger force than the market maker advantage if you’re happy with relatively small returns and you don’t even have to know that it exists to benefit from it.”
and when asked to explain exactly what he meant by this he added…
You’re betting IR on a horse race, some other customer puts in a back order on horse X for £10 at 2.
You reckon the correct price is 2.5, so you’re firing in your lay for £10 at 2.4, now- assuming that he bet a split second before you you’ll get his price, your lay is matched at 2.
Problem is if you bet £20 then you’ll only get £10 at 2, and the rest will be left for the next customer at 2.4 (even if he only asked for 2) – so on average you get a worse price. Add to this the fact that your bet is left sitting, unless you are actively managing the bet (hovering over the cancel button) this makes it a fundamentally bad value bet because (like a keep bet) you don’t know what you’re betting on.
The implication of this is that the smaller your bet the more likely it is that you’ll get a better price than you asked for, and in gambling price is everything, but everything.
I know this doesn’t sound like much, but trust me, I’m currently matching about 2-3000 in running bets every day and I can guarantee you that it’s an extremely powerful effect.
Another example to make the point is the archetypal 1.01 override backer, he simply sets his price to 1.01 and then allows best execution to get him the best price available- well the more he asks for the worse the price he gets (obvious- right?). The least he can ask for is £2, so that will always get him the best price- at the other extreme if he asks for £1,000,000 he’ll get most of it a 1.01, in-between there is a continuum and the trick is finding the right point on that continuum where you’re getting the best price and still matching decent amounts- you’ll understand that I’m not going to help anyone with that.
This effect is the reason many players find that they do well with small takes but fail as soon as they scale up. The same effect also applies in pre-off markets although it’s much less powerful.
If you don’t believe me (and you’ve got a few quid to spend) then make up a simple Gruss/excel bot that lays the shortest price runner in running once every second, and run two versions of it (ideally on separate computers)- one using £2 and the other using £20 stakes. Obviously you might win or lose over the course of a day, but what you will definitely find whatever happens is that your £2 does far, far better than your £20 bot. In this scenario it wouldn’t be at all surprising to see a decent win from your £2 and a nasty loss from your £20 with very similar bets on exactly the same races.
Aye Robot’s posts were always very informative and I believe he went on to suggest that his optimal stake per bet was less than £10, certainly a lot less than you would expect for someone turning over hundreds of thousands of pounds and finishing with many thousands of profit most months. Being a “2 pound punter” was and probably still is thrown around as a derogatory label to other users on the betfair forum but staking big amounts like Johnny Big Bollocks isn’t always necessary to make good returns.
Obviously Aye Robot’s activity was in running, he was submitting many transactions per second and what I am doing is pre-event but I’m still seemingly seeing a similar type of benefit. The tipster results for February were however, the 2nd best on record, so how much of these improved results are down to those above average returns and how much are down to the tweak in my methods?
Well, the £397 profit shown in the profit/loss screenshot doesn’t include £33 of profit with bet365 (still using £2 stakes) where I saw significantly bigger prices on a few selections and I had time to compare. I also frittered away £20-30 on a horse racing trading strategy that I’m trialing. That would give me roughly £450 profit overall and times this by 5 (to mirror the bulletin board’s £10 stakes) then I would have seen a profit of £2,250. This is still not quite the £2800+ seen on the board but it’s significantly closer than what I was seeing before.
When you also factor in the fact that I didn’t start the new approach until a week or possibly two into February and that there are tipsters I can’t yet automate for and didn’t place their bets, then I think we are getting very close to the profits shown on the board. Alcuni, Coldgold and AntaeusTen all made very good profits but I didn’t bet on their tips. This is of course comparable at least in terms of ROI, because as described by Aye Robot, increasing the stakes may present new problems in that I’m much more likely to get more be matched when I’m wrong (offering up value to the layer) and when my ROI is certain to decrease. This along with Cheltenham being just around the corner and which on our site has historically been poor for value seekers, is why I’m reluctant to increase the stakes on these bets at the moment.
To make the most of an effect like this, you need to turn over a decent amount of money and hence need to make a lot of bets. Clearly this is a lot easier when betting using automation methods on betfair but it does show that you can still make very nice profits betting small amounts. Bet365 have a very fast interface and you can place large amounts of bets worth them very quickly. If you keep your stakes low on there, say £10, then you’ll also likely go under the radar and less likely to be banned/restricted. You can also now bet through most of the bookies through Oddschecker very quickly but this requires a little more effort in the way of managing your funds across several bookmakers.
How much of the improved returns I’ve seen is down to me getting much better odds than I was on the under 10/1 selections or whether it’s attributable to even more favourable odds on the outsiders, I’m not sure and this requires further analysis. It will certainly be interesting to see what this new approach brings next month.