Thursday, 9 April 2015

March 2015 Monthly Review

I should just paste up the link to the March 2014 review and tell you to read this alongside the results for March 2015. It is quite scary how similar the month of March we’ve just experienced was to March 2014 and it was another nightmare month for the systems.

My conclusion last year  is quite similar to where I am now. “Mar-14 will now join the months of Feb-11, Apr-11 and Sep-13 in my memory bank. For me personally, this is the biggest loss I’ve ever suffered from betting and it has been a wake up call for me. I’ve discussed this before on the blog but I appreciate the way I play this game carries lots of risks but I have always felt like the returns were appropriate for the level of risk I was taking. Before this month, I was on course for a pretty good season which considering the start I had this season, would have been amazing. 4 weeks later and any thoughts of doubling my betting bank this season for the 4th straight season are gone. I’d take break-even again which is what I said a few months ago before things started going well.”

March 2015 goes down as my 2nd worst month of betting ever, not surprisingly, behind March 2014. Amazing how history repeats itself at times although my actual £ loss this year was a bit smaller. I’m pretty much break-even for the season. I said the same when I wrote the monthly review last year but I would happily take break-even for the season now. The current trends all point to me making a loss though, so I’m hoping for a swift turnaround in fortune.

Last year, it was a bit of a miraculous recovery but I followed my worst ever betting month with my best ever month and therefore, I actually ended the season in an OK place. Wasn’t a vintage season by any means and was well down on the previous 3 seasons in terms of ROI and ROC but at least I made a profit that was acceptable to me. Roll on 12 months and unless I see another miraculous recovery, I could be facing up to my first ever loss as a football bettor over a season.

In terms of the service and the systems, I think it now looks odds on that the systems will have their first ever loss over a season and therefore, it’s looking like a very disappointing season. Ironically, at the start of February, the systems were pretty much break-even after a great January but they’ve given back all these profits over the next couple of months to be in a hole again.

As promised in the last post, here’s the updated P&L for all systems to the end of March. I have also included the graph of the drawdowns.

As you can see, the systems are now on their 2nd worst drawdown since inception although they still have a bit to go before matching the drawdown we witnessed 12 months ago.

Anyway, enough doom and gloom. This is meant to be a review of the month, not a funeral for the systems and the service. Onto the review…..

Est Systems

Here’s the results for algorithms 1 & 2:

An 11.1pts loss from 102 bets on algorithm 1. Only 2 winners from 16 bets with 10%+ value shows the major issue this month on this algorithm.

Algorithm 2 did much better and managed a profit of 4.3pts from 118 bets. Again, when you look at the value bands, you are left scratching your head. 11.1pts profit for bets with less than 10% value and a loss of 6.8pts for bets with more than 10% value. Not really the pattern of results we’re looking for.

You would expect the systems to make a loss if I’m honest with these results. The wrong bets were winning and one algorithm had a substantial loss too. Here’s the results for the 11 systems:

A loss of 38.8pts from 553 bets. The filtering on the first algorithm made things worse which is hardly surprising given only 2 winners had 10% value. Systems 7 and 8 really struggled again but they had little chance given the poor results on system 6.  System 22 performed in line with system 21.

The combined systems lost 15.8pts from 224 bets which was disappointing and the combined systems are on course for another poor season this season I think. Given the results on system 6 and 21, it’s actually very disappointing that the combined systems are going to post a loss for the season but I said the exact same last season. The filtering on systems 6 and 21 just doesn’t work any longer or at least, nowhere near like it worked in the first couple of seasons.

Overall, a disappointing month for these systems but as we’ll see, this loss wasn’t too bad in the context of the rest of the TFA systems this month. There is much worse to come!

New Systems

Here’s the results for algorithms 3 & 4:

A 14.2pts loss from 134 bets on algorithm 3. Considering this is my best algorithm, the results have dropped off a cliff from mid January and we have now gone past the historical worse drawdown for this system. Very disappointing and after 3 very good seasons, there is a chance we’ll see this algorithm have its first losing season.

Algorithm 4 was more predictable with a loss of 10.5pts from 109 bets. Difficult to know what to do with this algorithm. It clearly has no edge and I’m tempted to keep it going to use it as a filter for other systems!

A couple of highlights from the value bands. Low value bets on system 31 lost 16.9pts, so if you exclude these bets, the algorithm managed a profit. Of course, the issue historically is that these low value bets on algorithm 3 are profitable, so you wouldn’t be excluding these. You’d be excluding the low value bets on the other algorithms and this season, they’ve made all the profits!

To highlight this again, low value bets on algorithm 4 made a profit of 15.3pts. A stunning return again from these low value bets on this algorithm but unfortunately, the bets with a value of 5%+ lost 25.8pts! You couldn’t make this up. It’s like the complete reversal of what we have seen historically on this algorithm. Goes back to a point I made above, the wrong bets are winning on the algorithms!

Overall then, with two algorithms doing really badly, the system results aren’t going to look pretty. Here’s the results from the 11 systems:

An 80pt loss from 627 bets. Disaster. On the plus side, the systems lost over 200pts last March from fewer bets, so it’s not as bad as last season’s performance in March.

Hard to take any positives from the results on these systems. System 33-42 somehow made a profit which is good going considering the rubbish performance on system 42!

Overall, a shocking month and these systems have dug themselves a hole for the season (again!). Could do with a quick recovery in April.

Misc Systems

Here’s the results from the 6 systems:

A loss of 31pts from 374 bets. A shocking month for STOZ (again!) and amazingly, STOZ found an additional 18 bets that STOY didn’t have and found…1 winner from these 18 bets. Amazing really and considering STOZ was potentially the best Misc System coming into the season, it definitely won’t be the best system leaving the season.

System 21-31 did pretty well to pretty much break-even and looking at the results since inception on this system, they aren’t bad. A 7% ROI over 1,714 bets so far and the system is on course for it’s 4th solid season in a row. Given the mess other systems have made of this season, this would be a pretty good achievement.
Overall, a poor month for these systems and in particular, the 3 SGM systems are really struggling this season. STOZ in particular is having a nightmare this season after a really poor season last season. Whatever has changed in the last 2 seasons has really impacted this system as much as any other I think. Shame really as a few seasons back, it looked like one of the best systems!

Draw Systems

Here’s the results for the systems:

A loss of 106.8pts from 304 bets. Easily the worst month these systems have had since they went live and nothing during backtesting was anywhere near as bad as this for a single month.

The filtering was a disaster again on these systems and system D2-D7 hit a perfect (or not so perfect!) 0 winners from 15 bets in the month.

Difficult to explain what has happened apart from blame it on the underlying results. The draw strike rate is the lowest it has been since I started betting on footie and therefore, I would expect the Draw systems to make a loss over the season when the strike rate is so low but maybe not to this extent.

The same issue is apparent here about the wrong teams winning. Here, it is the wrong teams drawing. Many of the draws are coming when I’ve backed the home or away team and therefore, they are not appearing as draw bets as often as I’d like. The games where I’ve backed the Away and Draw are tending to be won by the Home team and therefore, it’s just a nightmare scenario where I can’t catch the right games.

One midweek saw a plethora of draws in most leagues and I barely had a draw bet! I then had a ton of draws and there was barely a draw in any of the leagues!  Just want this season over and hopefully these systems can regroup and come back stronger next season.

Euro Systems

Here’s the results for the systems:

A loss of 9.4pts from 333 bets. Starting to sound like a broken record but the draw was a nuisance on these systems during the month of March. Of course, the games where the draw was a nuisance is games where I didn’t have an Euro Draw bet as we’ll see shortly!

Overall, not too despondent and although it’s a losing month, given the season these systems have had so far, I can handle a losing month.

Euro Draw Systems

Here’s the results for the systems:

Another massive loss of 88.5pts from 353 bets.

I’m actually in two minds about these systems now. One piece of work I’m going to do in the Summer is calculate the returns from laying the draw (I’ll look at ditching the H/A or staking 0.5pts on H/A) in these systems this season. Of course, it could just be the case that these systems have suffered from enormous variance and we’ll see a correction next season but I’m obviously not going to have these systems in the service next season in their present guise.

If I decide I wanted another shot at building Euro Draw systems (highly unlikely given the underlying results I calculated after I built these draw systems!) but if I did, I would obviously bin these systems and start again from scratch. In reality, I don’t think I do want another go at following Draw systems (lost a small fortune this season between the UK bets and Euro bets) and therefore, I might actually consider laying the Euro Draws next season in their current guise.

Anyway, we’ll keep this one for the Summer and I’ve no idea what the results look like by doing this but given I’ve never seen such bad ratings and systems before, I would think these could be pretty decent lay the draw systems to follow!

Overall then, a nightmare month for the systems and a combined loss of 354.5pts from 2,544 bets. Draws alone account for 195pts of this loss and therefore, anyone not backing the draw this season won’t be having quite the same level of nightmare as others.

Well, that sums up March. In a really tough season, we’ve just had the worst month the systems have ever had which only makes this season even tougher. I keep telling myself that if this is as bad as it gets, then we’re OK but unfortunately, things keep deteriorating and therefore, we may not be anywhere near the bottom of the P&L yet. If not, this could potentially get very painful for all of us following but at the end of the day, I don’t believe the ratings have suddenly lost any edge they had overnight. Yes, it’s clear the systems are nowhere near as profitable as they were a few seasons back but nothing indicated that they were about to fall off a cliff. It has to be variance and hopefully, this will correct itself over the next 6 weeks or so.

Let’s see what April brings…….