Wednesday, 18 June 2014

Preview of Euro Draw Systems

One of my big pieces of work since the close season has been to develop a couple of new rating algorithms to allow me to build a set of Draw systems for the European leagues. I suspect if it wasn’t for the fact my UK draw systems were my best performing systems last season alongside the fact my outright European systems ended the season strongly, I’m not sure I would be bothering looking at another set of systems. Given I already have 50 systems live for next season, adding another 11 systems might sound a bit crazy but as always, I’m always looking for another edge at this game to allow me to diversify within my betting portfolio.

Last season in my own portfolio, I followed UK outright bets, UK draw bets and Euro outright bets. The missing piece of the jigsaw for me was Euro draws and as far as I am aware, there are no standout Euro draw systems available in the market. Being honest, even if there was a standout tipster or system in the market, it would have to have a long-term proven track record to get me to pay for it and follow it myself.

I’ve been here before though and my solution is always the same….build something to follow myself in future once it is proven to work. Therefore, it seems a natural progression to move onto looking at European Draw systems.

I actually planned to do this later in the Summer but I realised when I was writing the reviews of the systems from last season that I needed a break from writing and therefore, I started playing about with the data for the European leagues. As always happens, once I get started with data and ratings, I struggle to stop and I managed to build these ratings fairly quickly since I had all the European data to hand as I was refreshing the Euro ratings for my other systems this Summer too.

I’ll do as I usually do and share the backfitted and backtested results but as always, these are not meant to provide a guide of what we can expect going forward. Depending on how good my ratings are, the true profitability tends to be around 25%-35% of the backtested results although it varies by rating algorithm. All I can guarantee is that the results next season live won’t look anywhere near as impressive as the results I’m about to show!

I tend to share the historical results so people can see how the systems look in terms of bet numbers and average odds. Knowing that the results tend to be lower when live, we can use the backtested/backfitted results to help us set some targets for the season forward although for things like betting banks, we need to be really careful as the drawdowns historically won’t be near as bad as the drawdowns we can expect in future.

Anyway, caveats aside, let’s see what I’ve come up with for these systems. I’ve spent a little bit of time thinking through which seasons to backfit and which seasons to backtest and I’ve gone down the route this time of the backfitting the last 3 seasons worth of data and backtesting the previous two seasons. I could have went back further and included more data but I think I’m of the opinion now that the best data to backfit on is the most recent data and as long as the results look OK for the previous seasons, everything should be OK. I used to think that it was better to build the ratings with historical data and backtest over the recent seasons but I think the issue with this is that the ratings then get out of date fairly quickly as I’m using data that is so old. Hence, when I’m refreshing all the rating algorithms this season on the other systems, I’ll be flushing through more recent data and tweaking the ratings.

The best place to start is probably the overall results to get a flavour of what the returns look like for all systems and to check the filtering works as I intended to. Here’s the results by system for the 5 seasons combined:

I always love this picture when I’ve developed a new set of ratings as it is my first check that I’ve built systems that could be useful. The filtering on the systems works like a dream (as we would expect based on my other systems I’ve built) and I think the overall ROI of 24.4% is in line with most of the sets of systems I’ve built. I think if this was below 15%, I would probably be worried and likewise, if it was above 35%, I would be worried that I’ve fitted the data too much during the backfitting.

Of course, the table above doesn’t help us too much since it has 3 seasons of backfitting and 2 seasons of backtesting included. Here’s the results by season:

I quite like this picture too. I think we can see quite clearly which seasons are backfitted (ROI is circa. 10% points higher) but the two seasons which are backtested still show a very strong return of 18%. Basing it on historical performance, I think the backtested return can convert to something like a 5% or 6% return as a minimum live return and this would be acceptable I think over a large number of bets.

Again, we can see that for any season, the filtering may not work as perfectly as hoped but over a large sample of bets as we’ve already seen, the filtering can work well.

One place I have struggled a little with the other Euro ratings has been managing to find something that works across all 4 Euro leagues. I do tend to find France and Italy quite difficult leagues when looking at Euro ratings as what works in these leagues tends to not work so well in the other two leagues. Interestingly though, for these new Draw systems, I didn’t seem to have too many issues with France as the draw strike rate tends to be fairly high in France. Here’s the results split by league:

I find this quite pleasing that there appears to be an even split of returns across the leagues historically. The Italian league has a lower ROI but it does have the highest turnover by a good bit (lower goal expectations in Italy relative to the draw odds I found) and therefore, the returns and turnover look acceptable.

It’s interesting to look at the returns by league and season as over smaller samples of bets, we can always see some crazy returns for some leagues. Here’s the results:

I do find some of these returns quite amusing! It’s quite interesting how 2 of the last 3 seasons in France has saw low returns even though this data is part of the backfitting process but the first two seasons have great returns! Likewise, Spain has really flat returns over the last 5 seasons which probably points to Spain being the most consistent league out of the 4 which fits in with my findings for the other Euro systems. Spain is the only league I seem to have a good handle on with the other Euro ratings.

I think that’s enough of an introduction to the European Draw systems for the moment.

As always, the best way to end these posts is with an overview of the targets for the upcoming season. With it being the first live season, it’s difficult to set accurate targets but I always try to err on the side of caution. None of us know how good (or bad!) these systems may perform next season but it’s always worth setting out targets so we have a benchmark to judge the season against.

Here’s the targets for this season:

I think the system that stands out is DE1-DE6 as being the one with the highest potential ROC but this is usually the case for the combined systems. My only concern with following this system is the high turnover and the fact that if anything goes wrong with these systems next season, this system is likely to lose more money than some of the higher combined systems. However, as we saw with the UK draws last season, if following the highest combined systems, you run the risk of filtering out too many winners and making a loss even though the ratings have an OK season! It’s a tough balance following a set of new systems as you don’t want to push all in which leads you to filtering the bets someway but if you do this, you run some other risks,

If I was giving my own honest advice of how to play these systems, it would be to give them a miss this season completely or to follow enough bets to ensure any edge shines through but to play trial stakes. Trial stakes is something I’ve been doing for the last few seasons now on my unproven systems and it sort of works for me. Yes, the money won or lost doesn’t actually make that big a difference to my overall P&L but I get a feel for what the systems are like and how the losing runs are etc. This means that if the systems perform as I hope, then when I migrate these systems to my main betting portfolio, I already have a feel for how these systems perform and what they are like to follow.

Overall then, I think this concludes the preview of the new European Draw systems. As always with my systems and ratings, things look very promising but we’ll need to see what happens when the bets go live. I think for next season, my hope would be to see the systems make a profit in the first instance. I’m still not as confident with the Euro leagues as I am with the UK leagues and therefore, I’m not wanting to set the bar too high in the first instance. If all systems could make a profit and hopefully have a fairly stable season, then this would be a good first season I think. I guess I want something to build on in future and it’s always easier tweaking ratings to improve them rather than build something new from scratch!

We’ll see what happens…..

Tuesday, 17 June 2014

Positives and Negatives from Season 2013/14

This post is going to be a more generic review of the season across all systems. I feel like I’ve spent a while looking at all the various systems and giving my view on them all individually and what went well and what didn’t go well but I’m going to look at the systems and service as a whole now for last season. I’ll start with the things that went well and in a really tough season, there were a number of things that went pretty well I think. Obviously, at the same time, there were a number of things that didn’t go well and it’s important I highlight these too. I’ll pick 5 things for each circumstance. I’ll end the review with a look at some splits of data and see where money was won and lost this season at an overall level.

Positives from the season:

All 8 algorithms made a profit – If someone had said to me at the start of the season that all 8 algorithms would make a profit over the season, I would have been over the moon with this. Coming into the season, I had 4 proven algorithms (6,21,31 & 41) and 4 100% unproven algorithms (D1,D6,E1 & E6).  Here’s the results of all 8 for last season:

Without trying to filter any bets or using any value ratings, simply putting 1pt on every bet that appeared on the algorithm turned a profit on every algorithm. May not sound too impressive but last season wasn’t easy for many football tipsters or systems and therefore, this level of consistency shows the strength of my rating algorithms last season was as good as ever.

TFA Value Ratings – This was the first season where I decided to release the value % on every game for the 4 UK algorithms and ultimately, the ratings performed in line with expectations. Going into the season, the observation I had made from the past few seasons was that the low value bets (less than 5% value) hadn’t made too much profit at all on the algorithms and therefore, removing these from the algorithms was one option.  Here’s the results of the 4 algorithms restated from above with any bets of less than 5% value removed:

Really pleased looking at these results and by applying a very simple filter to my current rating algorithms, we can increase the ROI on every algorithm. Next season, I expect more people within the service (including myself) will be paying much more attention to the value ratings.

Draw Combined Systems – The highlight of the season for me was probably the performance of the Draw systems. Going into the season, I didn’t have any great expectations for these systems as it was more of an experiment to see if I could build Draw ratings but the results were very good. I think the thing that excites me the most is the fact this season saw the lowest strike rate for the last 6 seasons in these UK leagues overall and yet, some of my draw systems had an amazing season. Here’s the results for the Draw systems this season:

System D2 stands out to me with a profit of 19.6% from 201 bets. With the suggested betting bank, it’s a ROC of 112% over the season. System D2-D7 also had a great season although the turnover was much lower. A profit of 38.3% from 104 bets which equates to a 114% ROC.

I think the issue with filtering the bets may be that we don’t end up with enough turnover on the Draw combined systems. However, system D1-D6 does have enough turnover and although the ROI and ROC isn’t as impressive as the two aforementioned systems, it is still pretty good. An 11% ROI from 390 bets and a 73% ROC.

Next season is a huge season for the UK Draw systems as they will need to build on their debut season.  They were largely under the radar last season but next season, I expect many people will be following a Draw system or two in their portfolio.  Let’s hope they can at least match this season!

System 31 – I have already touched on this system when I discussed the fact the 8 algorithms above made a profit but system 31 has definitely elevated itself to another level after the season past. Last season was the 3rd live season for this system and it was another great season. I loved the fact that when all my systems were making massive losses in September at the start of the season, system 31 managed to make a profit! With the high turnover and decent ROI, this system has become the cornerstone of the whole service and next season, I expect everyone will have a piece of system 31 in their TFA portfolio.  Here’s the results for this system since inception:

Over 2,157 bets across 3 seasons and an ROI of 7.3%. An ROC of 197% which equates to an average ROC per season of nearly 66%. A stunning performance from a football system and remarkable consistency.  I think if TFA only had one system, this would be the system.

League Two Aways – Long-term blog readers will be aware of the issues my ratings have had with League Two Aways and ultimately, it is the one subset of bets that before last season, I was really doubting if I had an edge with my ratings. I know a few people were less than patient than me and decided to drop League Two Aways last season from the TFA systems and being honest, based on the first 3 seasons, I couldn’t really blame them.  Here’s the updated results of all League Two Away bets:

After losing 258pts in the first two seasons from only 1,596 bets, things improved a little the season before last season but even so, going into the season past, League Two Aways had lost 168.8pts from 3,439 bets. A fairly big sample of bets and I can see why people would maybe think there was no edge here.  However, the season past was an amazing season for these bets with a profit of 568.2pts from 2,104 bets.

Looking at the overall results now, an ROI of 7.2% from 5,543 system bets would appear to show quite a significant edge I think. Therefore, what we have seen over the first 3 seasons is simply variance and being honest, the results last season have benefited from variance as there is no way the ratings will have a 27% ROI too often on any bets, never mind League Two Aways!

Overall then, I think that’s the 5 positive things to take from the season past. As I said in the intro, not everything went to plan this season and now let’s see what the disappointments were this season.

Negatives from the season

Filtering of the systems – Historically, one of the great things about the TFA systems and the service was the fact that if you didn’t want to follow all bets on an algorithm, the systems seemed to do a pretty good job at filtering the bets and therefore, you could play a higher combined system which had a lower turnover of bets but with the higher ROI expectation, you could still achieve an acceptable ROC. If we concentrate on the top 8 combined UK systems, we get the following picture for the live results:

I think Steve from the Daily 25 blog has taken a fair bit of stick this season for the way he set his portfolio up for this season from some commenters to this blog. Ultimately, the table above basically shows the issues Steve had this season. Going into the season, the 8 combined systems shown had generated 350pts profit at an ROI of 17.3%. An amazing performance.

However, the season past was a nightmare and the strike rate was over 12% points lower than the previous live results on these systems and unfortunately, this resulted in some large losses on some of these systems. All 8 systems suffered a loss and before this season, it was hard to imagine these systems could suffer a loss if my ratings were profitable but this is what has happened I’m afraid.

It’s easy with hindsight to say that following the higher combined systems carried a massive amount of risk as there aren’t too many unique bets on these systems overall. However, before the season past, we’d never seen these higher combined systems make a loss and therefore, it was an easy trap to fall into.

Home Bets Coming into the season, Home bets had been very consistent over the first 3 seasons and the fact that the season past was the best season for underlying Home bets surely translates into another great season for Home bets you’d think…..Err, no. Far from it. The season past was the toughest season my systems have had for Home bets unfortunately. Here’s the results over the 4 seasons:

Historically, my Home system bets have never been as profitable as my Away system bets but even so, seeing a loss last season is still a massive shock. Before the season past, a profit of 446pts from 5,579 Home bets. An ROI of 8%.  To go from this sort of ROI over so many bets to seeing a loss of 5.3% over the next 2,419 bets is very disappointing. The fact it happened in a season when Home bets generally had a higher strike makes it slightly more confusing but as I’ve shown with the Underlying results already this Summer, the type of Home bets winning (low odds) wasn’t caught by many of my systems.

I think going into next season, Home bets now carry a fair bit of risk and it will be interesting to see if they bounce back. I played higher stakes on Homes than Aways last season and it has cost me a great deal of money. With hindsight, introducing another variable into my portfolio (e.g. staking) may not have been the best idea and I think the season past shows that no matter what the historical results show for Home/Aways or League performance, there is no guarantee we will see the same again!

Bury & Partick I suspect if you asked all my subscribers which team lost us the most money last season, Blackburn and Middlesboro would appear high on the list. Blackburn lost 69pts from 194 system bets and Middlesboro lost 103pts from 146 system bets. However, the most disappointing teams of the season and the teams my ratings really struggled to get a grip on were Bury and Partick!

It’s actually quite scary how much I lost on both these teams personally (nearly 5 figures ffs!) but in terms of the system results, it wasn’t great. Bury lost 137pts from 191 bets which is an amazing loss! Partick wasn’t much better and they lost 123pts from 153 bets.

Trying to break down why this happened, if we look at where Bury finished, they finished 12th place in the league. However, in 26 games against the top 14 teams last season, Bury won 3 games out of 26! Yes, 3 wins out of 26 and yet, they finished in 12th place in the league. My ratings continually threw them up as value against teams above them in the league and unfortunately, they couldn’t win.

Looking at Partick, it’s quite easy to understand why my ratings continually threw them up as value at Home against bottom half sides and why they were continually thrown up as value Away from home against top half sides. Away from Home last season, Partick lost 1 in 10 games against bottom half sides. An amazing performance. At Home, Partick only won 1 of 10 against bottom half sides! I think you will struggle to find too many teams with that sort of record Home v Away and ulfortunately, my ratings fell into this trap Partick set.

Away from Home, the fact Partick were so strong meant they were thrown up as value against teams in the top half Away from home. Unfortunately, in 9 away games against top half sides, they only won once! Very frustrating.

I could probably go through most teams that accounted for large profits/losses and justify the bets we had over the season but at the end of the day, this is part and parcel of the game. Next season, we will bump into another Bury or Partick or even a Blackburn or Middlesboro and unfortunately, there is nothing we can do about it!

Low Priced Aways/Odds on Homes – I think I’m including this as a negative as quite simply, there were big profits to be made backing these type of bets last season and my systems missed them all.

As all followers know, I exclude any value Away bets where the odds aren’t greater than 2.75 (apart from the odd occasion) and I don’t give out too many Home bets where the odds are less than 1.80 (apart from the odd occasion).  Historically, my ratings didn’t perform well during backtesting with these odds ranges and during the first season of live results, there was no real edge for these type of bets.

My other issue is that I am not comfortable myself backing these sort of low odds teams and therefore, I deliberately go out of my way to exclude these bets from the systems. Over the first 3 seasons, I didn’t have too many regrets and for 2 of the 3 seasons, these bets I exclude from the rating algorithms were loss making. One season saw a small profit but the season past saw a very large profit on these bets (over 10% ROI) and therefore, during a season when it was very difficult to make money following the systems, it was very frustrating to see these bets perform so well.

My initial reaction at the end of the season was to adjust the filters on the bets and bring in some more of these low odds bets but I think I maybe overreacted. When I’m refreshing the algorithms this Summer, I will look closely at the profitability of this particular subset of bets and take it on a case by case basis. If I can find a clear edge and a decent ROI from the lower odds bets, I will tweak the odds filters on the algorithms. If I can’t find an edge, I will continue to exclude these bets going forward.

BSq Prem Aways & SPL Aways – To say it was a nightmare season for these bets is probably an understatement! The Underlying results for Aways in these leagues were very difficult but even so, the losses my ratings and systems suffered with this subset of bets were almost impossible to believe!

Here’s the results from last season for these bets:

A 513pt loss from 1,163 bets. An ROI of -44%! It’s safe to say that if we didn’t have these bets last season, the season would have been a much different season to the one many of us experienced.

If we look at the longer term picture, we see a much different trend. Here’s the results for the first 3 seasons for the same subset of bets:

Based on the first 3 seasons, there was nothing to highlight that these bets were about to have a disaster of a season. I think we have to look at the Underlying results and ultimately, my ratings were not near good enough in these leagues to overcome the bias against Away bets. Bsq Prem Aways now have a massive question mark against them next season but then again, League Two Aways had some big questions to answer last season and answered the questions emphatically!

I think this summarises the best and worst aspects of the season for my systems.

To end this post, here’s a look at some overall results for all UK outright systems. Again, we can’t really draw too many conclusions from these results but seeing some of the profits and losses will bring back some memories for all of us from last season!

Here are the results by month for all these systems:

September 2013 and March 2014 will always be remembered as a nightmare and April 2014 was as good a month as I’ve ever seen for my systems. What it does show me is that with so many systems in live play now, the losses and profits for any month could be substantial!

Here’s the results by league:

I think the thing I like here is that my systems not only had the best ROI for League Two but they also had the most bets in this league. This is not always easy to do but quite clearly, my ratings had a better handle on this league than the odds compilers last season. Bsq Prem has the same number of games as League Two and yet, my systems had less than 50% of the bets in this league compared to League Two. Again, my ratings struggled but at least they minimised the damage by having fewer bets.

Here’s the 10 teams who cost me the largest losses last season:

Already touched on Bury, Partick, Middlesboro and Blackburn as big losers for us but you can see the other main culprits here! Barnet irritate me as they had a great season but they always let me down when I backed them.

Here’s the 10 teams that won us the most points last season:

Newcastle were a massive winner for me last season. Seemed to continually get thrown up as good value most weeks and pulled off some great wins for us, especially at Home. Braintree pulled off an amazing win away at Luton at massive odds after taking a big lead and having to hold on. Interestingly, 6 of the top 10 teams were in League Two. If only every league behaved like this league!

I think that will do as a review of the season. Roll on the 2014/15 season…….

Tuesday, 10 June 2014

Still on track.....

It feels like a good time to take stock of where we are now I’ve finished reviewing the systems from last season.

Most of those who have subscribed to the service next season were sent a diary detailing how my workload was planned for this Summer. The eagle-eyed amongst you keeping check on my progress may think I’ve slipped by a week but in reality, I have stuck to the diary like it’s my bible! What I did do though was start looking at Euro ratings and the new Euro Draw systems soon after the season ended as quite simply, I was getting bored writing the Est Systems and New Systems reviews!

I know it’s hard to believe at times given how much I write on the blog and how long my blog posts are at times but I’m much comfortable playing with data in Excel or SAS rather than writing words down on a page.  Therefore, spending 3 weeks writing reviews of the season past isn’t really my idea of fun! To try to break this up for me (other ‘normal’ people would go for a walk or watch TV!), I decided to start refreshing the Euro ratings and start building the new Euro Draw ratings for next season. Therefore, I have adjusted the diary to take this into account:

As you will see from the above diary, I am exactly where I hoped I would be at this time of the Summer. If anything, I’m a couple of days ahead of schedule which may mean I get a weekend off this weekend to get on with building some furniture for the nursery! I’m terrible at DIY, so I suspect I’ll probably try to put this off for even longer and throw myself into more football data but my wife is starting to keep as close an eye on the TFA diary as I am. :)

I have refreshed the Euro ratings and being honest, there isn’t too much to share in terms of changes. I intend on doing a summary post once I’ve refreshed all rating algorithms but I won’t be restating historical results or going into any detailed results.

The next post should be a summary post which I see is a more general review of the season we’ve had and my own thoughts on what we learned from it and how we use this information going forward. I’ll try to do some data analysis on funky things (P&L by league, by team) for a bit of interest as I’m sure we’d all like to know which teams cost us last season even though it’s only for a bit of fun. I’ll try to keep this post fairly light hearted as I’m sure we’ve all got a little fed up reading reviews of last season!

After this, it will be a post on the European Draw systems. Really keen to see how these look and although I’m not quite there with the system building (I’ll finish it this week), the results look very good. Of course, the results ALWAYS look very good for my systems before they go live, so we won’t read too much into these results but it will be interesting to see what they are anyway.

I think after that, things will quieten down on the blog and I’ll be busy with data manipulation for 3 weeks solid. I’ll be taking some time off my day job to basically work solidly on the ratings as quite simply, doing this in 2 hour blocks is painful. If I can sit down and spend 8-10 hours doing this, I can get so much further than trying to do it every evening for a few hours. That’s my aim anyway and I’ll dedicate a day a week to doing these ratings as well as the other time in the evenings.

Anyway, that’s a quick overview of where I am and how things are coming along. At the moment, I’m on track…….

Under-Over Systems 2013/14 Season Review

I’m going to keep this review very, very short. Both these systems have now been retired. I should have probably retired them after the 2012/13 season as I think it was clear then was no edge there any longer and I think for a market such as the Under/Over goals market, systems which were built a few years ago on data from before that won’t maintain an edge.

I think you would need to have pretty good ratings to maintain an edge in this competitive market and although these systems started off OK, they flattered to deceive and probably needed a lot more time investment in them to make them viable long-term. Given I only looked at these systems as a favour to someone, I wasn’t going to put the necessary work in to maintain an edge on systems that no one followed!

Here’s the results for the 2013/14 season:

Not a pretty sight and in a market which is not too far away from a toss of a coin, to only hit a 44% strike rate is poor. The season before saw a small loss and overall, the systems clearly have no edge after a promising first season. Over 3 seasons, a profit of 12.6pts from just over 1,061 bets probably tells the story and therefore, both these systems will now be retired.

Phew, that was a short review! 

Monday, 9 June 2014

Euro Systems 2013/14 Season Review

Before I start reviewing the performance of the Euro systems this season, I should touch on the results the previous season and the changes implemented as a result. Although this is the 2nd season I’ve had Euro ratings running live, the rating algorithms this season were not the same rating algorithms as last season. Here’s a quick recap of the results for the Euro systems in the 2012/13 season:

A loss of 127.7pts from 2,880 bets. Both rating algorithms lost badly and the only saving grace was that the filtering worked fairly well on the systems and this helped minimise the losses. It was actually very encouraging that systems E2-E7, E3-E6 and E3-E7 made a profit when the ratings were so poor as it meant that that the filtering idea I applied on the UK systems worked on the Euro systems too.

If we look at the breakdown between Home and Aways, we can see the following:

There were slightly more Away bets than Home bets and the overall average odds were 3.84. Homes were actually the main reason for the loss but overall, I was uncomfortable with this split of Home/Away bets in the Euro leagues and I thought the average odds of 3.84 were a little too high for my liking.

At the end of the 2012/13 season, I basically decided that I should scrap these rating algorithms completely and start again for last season. I decided I wanted a set of Euro systems that were Home based rather than an even split between Homes and Aways. This would reduce the average odds and hopefully reduce the losing runs the Euro systems would experience.

Last Summer, I developed a couple of new rating algorithms and from backtesting, the results looked more in line with what I was aiming for. 70%-75% of bets would be Home bets and I was much more comfortable with this idea for Euro ratings. I hoped that by learning from my experience of building the ratings the previous season and seeing them fail, I’d be able to build better ratings the second time round. Here’s the results from 2013/14 season:

I guess the first point I want to make is that if you fail at something once, it doesn’t make you a failure! These results are much, much better than what I achieved with the first set of Euro ratings I built and therefore, I’m really pleased with the season these Euro systems have had. A 7.5% return over 2,075 bets isn’t an amazing return by any means but considering the mess I made in the 2012/13 season with these bets, it at least gets the Euro bets into profit over the first two seasons.

I’ve said this before but I’m quite proud of my ability to build profitable ratings and systems and therefore, my ego did take a bit of a knock after the first season with the Euro bets. However, seeing both algorithms make a profit of 5.8% and 6.7% respectively over a season is a great result and a boost to my confidence in these Euro ratings.

Looking at the results in detail then, I think the disappointing aspect is that as we move down the systems on the first algorithm, the results slightly deteriorate. Obviously, the opposite was true in the first season although I’d rather the filtering didn’t work great and the systems be profitable than the filtering working great but the systems being unprofitable!

On algorithm two, the filtering did work well and system E7 made a profit of 11.6% from 184 bets. A really solid return and given the fact the 5 single systems were profitable, it isn’t surprising to see the combined systems do well.

The 6 combined systems all had a decent season and profits of 6.2% to 10.7% were achieved across all 6 systems. The bet number this season was lower than the first season (deliberately so as I tried to focus on quality rather than quantity) and I think this worked well. Of course, the knock-on impact is seen on the Return On Capital figures and these Euro systems in their current guise are never going to make anyone rich but I’m just pleased they had a profitable season.

Here are the results by month:

December was a nightmare month and at the end of December, I think I was seriously contemplating giving up on looking to find an edge in the Euro leagues. The systems bounced back almost straightaway though in January and I guess the reason I’m so upbeat in this review is because they ended the season well. I think how systems end the season swings my general mood when writing a review and these systems ended the season strongly.

The results by league are always interesting to look at and here they are:

If I look back at the 2012/13 season, France and Spain were the two profitable leagues (especially Spain) whereas Italy and Germany had massive losses.  Spain again ranks highest although France is a losing season this season and Italy doesn’t look great either. I know we can’t really sum season’s together as it’s different systems for each season but Spain has now created a profit of over 200pts profit from over 1,300 bets. A really good performance.

Looking at the split by Home and Aways, we see this:

Aways had a tremendous season and achieved an ROI of 13.6% from 517 bets. Homes made a 5.5% ROI from 1,558 bets which in itself is a great performance and overall, really pleased with the split of bets and their respective profits. All things being equal, Aways should achieve a higher ROI as the average odds are much higher.

I still think it’s too early to draw too many conclusions and after one bad season and one good season, I don’t think I’d be shouting that these systems have an edge in the Euro leagues. Clearly, the ratings now are much better than what I had originally but just how good they are, I’m not too sure.

I suggest the next season is the key season for these systems and if we get another positive season, I think we can start trying to draw conclusions about what sort of edge the Euro systems may have and whether or not they are worth following seriously going forward. I’ve now followed a couple of Euro systems for two seasons and managed to make a profit each season. My stakes have been very low and even after the season past, I’m not confident enough to increase stakes to anything other than trial stakes. I’ll apply the same staking level next season too.

To end the review, here’s the updated targets for these systems going forward:

One thing that strikes me from looking at this table is the low average Return on Capital figures. Unfortunately, with the systems having a lower turnover than the UK systems to try to improve the ROI, it really impacts the ROC and therefore, when refreshing the rating algorithm this Summer with updated data, I’ll be looking to see if we can increase the bet number slightly whilst at the same time, maintaining the same level of ROI. 

That concludes the review of the Euro systems. I think we need to see more live results before drawing any concrete conclusions and I’m really keen to start tracking the Underlying Euro results next season which will help me to put this season into some sort of context. I know the draw strike rate was lower this season than normal and therefore, it’s likely my ratings have capitalised on this to an extent although I also think there was an Away bias this season and 75% of my bets were Homes, so it’s hard to read too much into the Underlying results without doing the necessary work like I did for the UK Underlying results.

Last review for the 2013/14 season will be the Under/Over systems which won’t take too long as these systems are now in their retirement home!

Saturday, 7 June 2014

Draw Systems 2013/14 Season Review

I think in many ways, the review of the Draw systems this Summer is maybe one of the most important reviews this Summer. At the start of the season, I was really keen to see how my draw ratings and systems would do. I put a lot of work into developing these systems last Summer and like anything in life, when you develop something, you hope to see them go on to do well.

After the first month of the season when the Draw systems made a loss, I probably jumped the gun by saying I had little confidence in the systems. I was struggling a little with the higher average odds I think and seeing a strike rate that’s a little lower than I’m used to. I guess my other systems have Home bets which increase the strike rates on them but with Draws, the average odds are always going to be around 3.45 and therefore, your strike rate isn’t going to be too high at these odds.

Once the systems had recovered the losses and followed it up with a very solid run of monthly results, my confidence grew and grew and I think it was after February that I really started to sit up and take notice of these Draw systems. One thing I started doing then was tracking the underlying results of the UK leagues and it started to hit me that it was a very difficult season for backing the Draw and yet, my systems were doing pretty good.

Unfortunately, as always happens in life, just when you think you have it cracked, the systems had a poor April and May and ultimately, they made a loss over the last 20% of bets last season. Hence, they started poorly, ended poorly but the in between wasn’t too bad at all!

Before we get on to looking at the results, let’s remind ourselves of the difficulties those backing the draw faced this season. Here’s a recap of the underlying results:

As you can see, the overall strike rate was only 26% and this is actually the lowest strike rate of the past 4 years I’ve been keeping ratings. The Premiership and SPL were both very difficult leagues with strike rates of 20.6% and 22.3% respectively whereas League Two was a dream for draws with a strike rate of 32% and an ROI of 12.4%.

Looking at the results by month, September and April were the toughest months by far.

OK, here’s the results of all the Draw systems this season:

At a high level, a profit of 229.6pts from 2,535 is a great result and an ROI of 9.1%. However, it wasn’t all good news and there are some interesting trends to comment on.

Firstly, the algorithms didn’t actually end up with that impressive results. The first draw algorithm made a profit of 3.5% if backing all bets and the second draw algorithm made a profit of only 0.6%. It’s amazing how good the overall profits were on these systems considering the ratings didn’t perform that well but of course, we need to take the results in context of how difficult the season was generally. In addition, although you can’t see it here, backing 1pt on every draw bet throw up on either algorithm was a loss making strategy over the season. Hence, it wasn’t the best season for my draw ratings by any means.

The other issue was the filtering of the ratings on the first draw algorithm. It worked like a dream from D1 to D2 and D2 made an unbelievable return of 19.6%. However, D2 to D3 didn’t work at all and D3 made a loss. Quite a confusing picture really although as always, the sample size on D3 isn’t huge and there weren’t enough bets to show the edge the system has I suspect.

On the second draw algorithm, the filtering worked like a dream. 12.3% ROI on system D7 is a great result. 

As always, when looking at the combined systems, we need to analyse one main thing, did the cross referring of the algorithms improve the results?

System D1-D6 did an amazing job this season. An 11.3% ROI from 390 bets. Being honest, the top 4 combined systems all did an amazing job this season. D1-D7 made a profit of 15% from 188 bets. The best performing systems were D2-D6 and D2-D7.  A special mention to system D2-D7 I think which made a profit of 38.3% from 104 bets. This is the brother of system 7-22 (which of course has lost its halo this season!) and system 32-42 (which had a losing season too).

The issue on the Draw combined systems not surprisingly comes from systems D3-D6 and D3-D7. Given D3 was loss making, it is difficult for these two combined systems to make a profit. The sample sizes are very small and therefore, I’m not sure these top two combined systems are worth following in future as they’ll need some luck to make a profit over the season with such a small sample size.

Overall then, a really strong performance from some of the Draw systems but of course, if anyone chose to follow systems D3, D3-D6 or D3-D7 (of if you are Steve, follow two of these draw systems!), then you may not have had such a high opinion of the draw systems. If you chose to follow one or more of the top 4 Draw combined systems, then I suspect your opinion of the Draw systems is a little bit different to Steve.

Let’s take a quick look at some other breakdowns of the Draw returns for this season. Here’s the results by month:

If you look back at the Underlying results by month, you will recall that the two toughest months by far were September and April.  These were the only two months my Draw systems suffered a loss.  No coincidence I’m sure and it again highlights what I’ve been saying for a while around how much my results are correlated to the Underlying results.

Here’s the results by league:

Again, if you look back at the Underlying results by league, you will see that the Championship and League Two were the two leagues where you could have made a profit from backing all games to be a Draw. Not surprisingly, my returns in these two leagues were astronomical and a 28% ROI in the Championship is an unreal return I think. League Two wasn’t too far behind with a return of 15.3% as the sample size was much larger in League Two.

The Underlying results pointed to the Premiership and SPL as being the toughest two leagues last season for Draws and this is backed up by my returns in these leagues. My systems only managed a 17.8% strike rate in the Premiership and SPL  and losses of 38% were achieved which is far worse than the Underlying results.

I guess one thing I should make clear is that I don’t believe that the results by league or month mean anything. Every season, trends will change and underlying results will vary across leagues and my returns will therefore vary across leagues but as long as my systems have enough bets across the various leagues, the edge should show through and a profit be achieved. I would steer anyone away from trying to draw conclusions from this data over one season as it really doesn’t mean much.

Think back to League Two Aways and the fact that for the first 3 seasons, I couldn’t make any profit from these bets at all. Last season,  League Two Aways accounted for a huge chunk of my profits on the outright systems as the strike rate for backing all League Two Aways was higher than the last few seasons.

Lastly, let’s take a quick look at the results by odds ranges:

I find this a fascinating picture tbh. Quite simply, if you were trying to back bigger priced draws last season, you had no chance! My systems only hit a 10.6% strike rate on bets where the average odds were 3.96 for all draws greater than 3.71. Crazy really. All the profits on the Draw systems came from bets priced 3.31 to 3.50 and the other 40% of bets made a large loss.

Let’s look at the targets for next season:

Before I discuss the future targets, a quick comment on just how good the season past was. System D2-D7 managed to achieve a ROC of 114% and system D2 achieved a ROC of 112%! Unbelievable returns and shows how special a season it was for these draw systems. On average across the 11 draw systems, an ROC of 46% was achieved. I think this is as good as I’ve had for a set of 11 systems over a season.

When setting the targets for next season, I have deliberately tried to play on the pessimistic side. I don’t believe one live season of amazing results should influence the future targets too much and therefore, I have stuck with a longer term view. D2-D6 and D2-D7 are probably the two draw systems to take from the season although given what I have said in other reviews, we may have to ask ourselves what we gain from cross referring other systems with system D2. If we had another season like this one for system D2, I suspect we wouldn’t need to worry too much about other draw systems!

That completes the review of the draw systems. I do think there is an amazing amount of potential here for these Draw systems but as always, one season isn’t long enough to make too many conclusions. I think the thing that excites me the most is that the season past is the toughest season anyone backing draws would have had for years in these leagues and yet, a couple of my systems doubled their betting bank! If they can do this during a really tough season, what will they do during a more normal season or what happens if we have a season where the draw strike rate is higher than normal……the mind boggles!

Thursday, 5 June 2014

Misc Systems 2013/14 Season Review

Here’s the results for the 2013/14 season:

An interesting set of results and depending on which system or systems were followed, results could have been quite variable.

The reason the Misc systems have this particular name is due to the fact that the systems are a bit of a mixed bunch. 3 systems are created from cross referring other systems on other algorithms and 3 systems try to filter the bets on 3 other systems by using a different method to my other systems. This season was their 3rd live season. The first season was poor, the second season was great and unfortunately, the 3rd season has been poor again.

It’s very disappointing that system 6-32 made a loss over the season.  Given the fact system 32 was profitable (3.7%) and system 6 was profitable (3%), then to see this system not make a profit is very poor. It’s a common theme in these reviews at the moment I think but what is the point of system 6-32 if it can’t improve on the returns of system 6 or 32?

System 21-31 made a profit of 6% from 405 bets. Looks OK on the surface and a decent season for any football system I suspect. However, when you see system 31 made a profit of 7.6% from 753 bets, you have to question what is the point of cross referring the system with system 21 if it reduces the ROI and reduces the turnover?

System 6-21-31 made a profit of 6.7% from 382 bets. Same comment applies really to system 21-31. A good return this season but when you see the returns system 31 have achieved, you have to ask what is the point of this system if it can’t improve on system 31?

System TOX is basically a filter of system 21. The system made a profit of 4% from 292 bets. System 21 only made a profit of 2.6% from 462 bets, so I guess in a way, this system at least managed to beat the system it is filtered from. However, 4% return isn’t really acceptable for this system given the historical drawdowns and therefore,  it’s a poor season for this system.

System STOY is a filter on system 6-21. Clearly, it is very highly correlated with system TOX and therefore, following TOX and STOY in a portfolio provides no diversification at all. This season, STOY made a profit of 7.5% from 284 bets. A profit of 21.4pts. Considering system 6-21 only made 12pts profit from 410 bets, STOY at least managed to do its job this season. 7.5% is a decent return and STOY was the best performing system out of the 6 Misc systems this season.

That only leaves STOZ which is a filter on system 31. Last season, STOZ made a profit of 21% from 336 bets and going into this season, STOZ would have been high on the list of TFA systems to follow this season. Unfortunately, this season was a nightmare. The system did an amazing job for most of the season of picking the worst bets on system 31! The system crept into profit by season end and ended up with a profit of 2.3% from 388 bets. When you consider system 31 made a 7.6% profit from 753 bets, then to see STOZ made this sort of return is very poor.  A nightmare season for the system.

If we look at the results by Home and Away, we can see where the issues were this season. Here’s the results:

Aways made a profit of 9.6% from 1,369 bets but Homes had a nightmare season. A loss of 10.1% from 483 bets. Considering how difficult it was to make profits from Aways and how ‘easy’ it should have been on Homes, to see these results is a real headscratcher!

Coming into the season, TOX, STOY and STOZ had the best results for Home bets out of all my systems, so it’s surprising to see these systems make a loss on Homes. The other 3 systems were never that great with Home bets anyway before this season, so seeing losses on these systems is less surprising.

Let’s take a look at the results for the last 3 seasons in total to see if these systems add anything to the service at all:

I think looking at these results in total, things look much better than the last season in isolation.

I think it’s fair to say that system 6-32 isn’t really a system worth following. 20.7pts profit from 641 bets is a poor return and when you consider system 32 has made 25pts profit from 512 bets, I think this highlights the fact system 6-32 does nothing for the service. I’ll keep it going for another season but I would look to retire this system at the end of next season if the same pattern remains.

System 21-31 has made a profit of 7.7% from 1,280 bets. Clearly, the system has an edge and it’s not a bad return for a football system but as always, we need to ask ourselves what this system brings to the service. System 21 has made 7.2% from 2,111 bets and system 31 has made a profit of 7.3% from 2,157 bets. Clearly, system 21-31 improves the returns on system 21 and 31 slightly but also reduces the turnover greatly. For those that want a lower turnover system than 21 or 31, they could do worse than follow system 21-31 I think.

System 6-21-31 has made a profit of 7.2% from 1,191 bets. Clearly, this system is not as good as system 21-31 and therefore, I don’t think it adds anything to the service that systems 6,21, 31 or 21-31 already bring to the party.  I’ll give it another season but I would look to retire this system after this season.

TOX has made a profit 9.9% from 877 bets and is actually the best system out of all the Misc Systems. Given TOX is correlated to system 21, it’s surprising that this is the case as system 6-21 and system 31 are stronger systems than system 21. System 21 has a return of 7.2% from 2,111 bets and therefore, TOX is a great alternative to system 21 I think for those that want to follow the 2nd algorithm but don’t want to follow every bet.

STOY has made a profit of 9.4% from 872 bets. Given this is correlated to system 6-21, it’s pleasing to see that 6-21 has made a profit of 7.7% from 1,751 bets and therefore, STOY improves on this also.

STOZ also improves on system 31. A profit of 9.4% from 996 bets is a great return as system 31 has made a profit of 7.3% from 2,157 bets. It’s an interesting one as I think there is an argument to say that it’s much better to have a lower ROI by a couple of points and a turnover over 100% higher and I would agree with this. System 31 is a better system than STOZ IMO but I can see why some people may not want to follow all the bets system 31 has and therefore, STOZ provides a good alternative to system 31.

Overall then, I think 4 of the 6 Misc systems add something to the service. 6-32 and 6-21-31 don’t really improve on the returns of the single systems and therefore, I wouldn’t suggest people follow these systems.

Here’s the live results split by Home and Away bets:

For the first 3 systems, it is clear that the Home bets are fairly weak. 6-32, 21-31 and 6-21-31 actually make no profit in total from Home bets. However, the Away bets on these systems do OK although 6-32 lags behind.

For the 3 SGM systems, the Home returns continue to outperform the Aways even though the season past has put a big dent in the Home returns.  I think it’s clear that over the first 3 seasons, the Home bets on these systems have been my best Home bets and therefore, it’s not surprising to see many other systems struggle with Home bets last season. If these 3 systems can’t make a profit backing Homes, I doubt many of my other systems can.

Here’s the targets for next season for these systems:

I think the ROC each of these systems have achieved on average each season tells the story of these systems and 5 of the 6 systems appear to make around 50% ROC each season which is a very good return. The targets for the next season are pretty much in line for each of the 5 systems too and therefore, it’s easy to see that system 6-32 isn’t worth following and should probably be dropped from the service.

I think I’ve maybe come across as a little harsh towards the Misc Systems overall but at the end of the day, the systems don’t appear to improve too much on the other systems and therefore, we have to question if it’s worth following any of these systems. I do think there should be a place in everyone’s portfolio for a Misc System but I think if you start following too many of these systems, you basically end up with a very correlated portfolio of bets which isn’t a great idea as last season showed.

I personally find it very difficult to separate the 5 systems. Before last season, I would have gone with STOZ as the strongest of the 5 but after a poor season, it now lags behind the other 4 systems! I think any of the 5 is probably worth following in a portfolio although if you are following systems 21,31 or 6-21 already, you have to ask yourself what is the point of following one of these 5 Misc systems……