Monday 12 March 2012

Results Update

I get a feeling this could be a long post today as there is a little bit of explanation needed for the results tables and in addition, I want to have a little moan about the weekend and the fact we didn’t get many breaks go our way. However, before I get started, I want to mention another blog that has been added to my blog list on the sidebar.

Rowan has decided to return home to The Portfolio Investor blog and retire from his ‘A Punter's Year’ blog. I always enjoy reading Rowan’s blog(s) as the guy clearly has a sense of humour that comes through in his writing and isn’t afraid to wear his heart on his sleeve. He also isn’t afraid to give an honest opinion on services and therefore, I think blogs like Rowan’s help other punters know which services to follow and more importantly, which services should probably be avoided. I’ve added a link to his new blog at the side but I’ll keep the ‘A Punter's Year’ blog on my sidebar for a wee while in case people want to read back.

Before I get started on the weekend results, a quick explanation concerning the results tables that will appear on the blog from now until season end. As discussed on the blog a wee while ago, there is a lot of work going on in the background to improve the offering of The Football Analyst service. I call it a service but at the moment, it’s an email subscription list and a blog where the bets get posted free of charge, so not much of a service but then again, you get what you pay for in this game! :)

Firstly, the odds quoted in the results table now appear as 5.00 for a team that was 4/1 to win. This matches the way the bets are posted but for some reason (don’t ask me why!), I’ve always had my results database set up so that the odds recorded would be 4.00 for 4/1. It doesn’t change the P&L obviously as I just take 1 off the odds quoted to work out the profit on the bet now but something worth knowing in case you look back to previous posts and notice a difference!

One of the main pieces of work I’ve done is add some AH Line results to all my results data. I initially thought about going back and trying to backtest this live using odds portal etc. and choosing the most competitive AH line for each game but that would be far too much work! So, as a compromise, what I’ve done is use my current data to calculate the odds for 3 AH lines. As you know, I already monitor and track Draw No Bet (DNB) which is really just another name for AH0, so all I’ve done here is change all references in my sheet from DNB to AH0. Easy!

I’ve also added AH +0.25 and AH +0.50 to all my results data and tables also. For those not familiar with these AH lines, AH +0.25 means that you win half of your expected profit from backing outright if the game is a draw. AH + 0.5 means that you win if the game is a draw or if your team. AH +0.5 is also known as Double Chance and it is effectively a dutch bet on the draw and your selection to win.

Basically, it’s fairly easy (OK, it took me about 3 hours to suss out the AH +0.25 formula!) in Excel to calculate the odds for all these AH Lines as long you have the odds on your selection to win and the draw odds. Therefore, I now have the results for backing outright and AH0, AH +0.25 and AH +0.50 in my results database.

Every time I post the results up for the systems now, it is going to include the results for these additional AH lines. For the moment, these should be ignored!!!!!

OK, so I go to the trouble of adding all this data to my results and then I tell you to ignore it. Why?

Well, firstly, I don’t give the odds of AH +0.25 and AH +0.50 when I release the bets, so it seems a bit unfair to discuss the results. The weekend was a really good weekend for AH +0.25 and AH +0.50 but it would feel like I’m taking the pis* a little discussing the results on the blog considering I don’t even give this as an option to follow the bets!

More importantly though, AH betting will never improve the edge my systems have and the reason I’m introducing this option it isn’t linked to increasing the edge. I discussed it a little in this post here http://the-football-analyst.blogspot.com/2012/02/roi-for-show-roc-for-dough.html about comparing the returns of DNB (AH0) and backing outright but this can be extended further into other AH Lines.

However, AH betting has the ability to DESTROY the edge that my systems have and therefore, now isn’t the time to start pondering about how we use AH Lines. Quite simply, there are 2 months of the season left and I need to get us there in one piece. Therefore, all my time needs to be dedicated to keeping my ratings updated and finding the system bets and therefore, now isn’t the time to be discussing AH betting in detail.

Therefore, it is imperative that no one starts considering AH betting until I’ve done the necessary analysis on the systems to try to figure out where and when it may be best to use AH. Now that I have AH results in my results database, I can see quickly that using AH betting on all games will NEVER be a profitable method for my systems and therefore, it’s important that I do a fair amount of analysis and then spend a while educating people about how and when they may use AH betting in future. This will be done in the Summer, probably via this blog.

Anyway, I hope that makes sense and don’t get too drawn into looking at the AH returns on the results I post up. They don’t mean too much at the moment.

Onto the weekend then……

The base results weren’t a disaster and ultimately, a small loss was made, so not the worst weekend of the season (by a long, long way!). I try my best on the blog to not get too distracted by late goals, teams losing 2 goal leads, sending offs, missed penalties, goals not awarded etc. as at the end of the day, over the course of a season, most of my systems have enough bets to ensure that things should level out over the course of the season hopefully. Saturday was definitely one of THEM days though, if you know what I mean.

Chesterfield lost a 2 goal lead at odds of 5.5. Bristol Rovers lost a 2 goal lead at odds of 4.33, with the 2nd goal coming in injury time. Cheltenham missed a penalty to go into a lead against Swindon, lost 1-0. QPR scored a legitimate goal whilst on top against Bolton and went onto lost 2-1 (even though they scored a later goal that was offside!). As Mark Hughes suggested though, if QPR had been given their goal whilst on top in the game, would have been a totally different game from then on. We also had Rochdale at 7.00 and Accrington at 5.00 draw 2-2 against teams at the top of the leagues.

4 of the above mentioned games finished 2-2 and the other one was a defeat for QPR. I’m not saying that we deserved to win all these games but when you look at the odds of the teams we’ve backed in these games, it only takes one goal to go our way and it made it a totally different weekend for the systems.

No point crying over spilt milk though and in the long-run, if we keep finding these sort of bets, I’ve no doubt that we’ll win more than we’ll lose.

Algorithm 1 had 20 bets. 5 wins, 7 draws and 8 losers. A loss of 1.5pts for H/A betting and a profit of 0.2pts for DNB betting.

Algorithm 2 had 16 bets, 4 wins, 5 draws and 7 losers. A loss of 1.6pts for H/A betting and a loss of 0.7pts for DNB betting.

Algorithm 3 had 16 bets, 5 wins, 6 draws and 5 losers. A profit of 1.5pts for H/A betting and a profit of 2.5pts for DNB betting.

With the base ratings struggling, it’s a pleasant surprise that the systems results aren’t worse this weekend!

The established systems made a loss of 7.2pts for H/A betting and a profit of 0.4pts for DNB betting. The new systems made a loss of 0.1pts for H/A betting and a profit of 7.4pts for DNB.

Overall, a better weekend for DNB rather than backing outright which is always a sign that you’ve hit too many draws in the weekend. The fact the draws were mainly big priced away teams makes it all the more frustrating really.

I’ll be back tomorrow at 9am with the midweek bets. March is going to be a very long month as there are 5 Saturdays and all of the midweeks have games too, so a pivotal month in the season I suspect for my systems.



2 comments:

  1. Hi Graeme.

    Great to see you have added the AH results to your data. One thing that I was wondering is why you don't have lines like -0.5 or -0.25 etc. but only + lines ?

    You should also be aware that on the lower leagues (league one and lower), the big Asian firms will only have 1 or 2 lines. Therefore, the prices will not be very competitive for some of the AH lines.

    How do you decide which lines you will taking for any game?

    Mark

    ReplyDelete
  2. Hi Mark.

    I’ll reply to your comment on a new blog post as I think you ask some very good questions and I’d like as many people to see my responses mate.

    Cheers,

    Graeme

    ReplyDelete