Monday, 11 June 2012

First look at System E2

As I’ve discussed before, we need to be careful when we think about how we get from system E1 to E2. Although I talk about filtering the bets, in a way, there is no filtering going on here in the way that others would filter a system.  The filtering process is inherent in the way the systems are built using the preparation systems and therefore, the filtering process carries much less risk than if someone was taking a system with the results shown on system E1 and then applying filters to get system E2.  (Now I’ve written that paragraph, it reads like gobblygook to me and I understand what the hell I’m saying but hopefully those reading this get it. If not, read this…. http://thefootballanalyst.com/tfa-systems/ )

So, what does system E2 look like?

Well, not surprisingly, it looks better than E1!

Here’s the results by season:


Well, the 17% return on system E1 has become a 20% return here.  The number of bets has dropped by 40%. Don’t know about you lot reading this but I was disappointed with seeing this level of return on this system. I sort of assumed that when you cut out 40% of the bets, you’d end up with a better return than a 17% increase in ROI? Maybe I want too much!

It’s funny as I remember having the exact same thoughts about system 7 after I looked at that compared to system 6(partly why I’ve never liked system 7!).  Anyway, I fell in love with system 8 back when I produced the first systems, so I expect system E3 might look a bit better. 


Here's the results if I strip out the 3 seasons which may be affected by backfitting of some sort:


Similar comment again here and these results looks don't really set my pulse racing. The drop off in bets is also slightly worrying but then again, it's nearly identical to what happened with the first set of systems for the UK leagues. They reduced in bets over time until the systems went live.


I think the split between of Homes and Aways will be interesting here. It was 60% Home bets on system E1, so I'm curious to see if it changes at all as we filter the bets.  


Here's the split:



Well, the split has gone a little more in favour of Homes (60% to 62%) but the returns are interesting.  The 15.7% ROI on system E1 has become 17.4% here.  The Aways have gone from 19.2% to 25.3%.  


I'm a little perplexed by this as during the backfitting process, as I've already said, it was very easy to get something to work with Homes and achieve a very high ROI but Aways was much more difficult. However, the system results aren't really backing this up. 


I don't want to draw conclusions too early but I personally will be wary of Aways on the European systems compared to Homes I think.  Not sure I trust the Away bets as much as the Home bets but the results look very good nonetheless.


Lastly, here's the results by League:


Spain is interesting here. The 16.1% ROI on system E1 has only become 16.2% on System E2. Italy sees a big improvement as we move from E1 to E2 but the other two leagues see small improvements too.  


It's hard to draw too many conclusions from this at the moment. Like I did with the UK systems before they went live, I'll produce a detailed spreadsheet which analyses all the Euro systems. This sheet will look like what I've done with the UK system reviews this Summer.  I'll use this sheet to help us decide what sort of target ROI/ROC we'll be aiming for with these Euro systems and will help us set appropriate betting banks.


One thing that's important for me when I develop new systems is that I take readers/followers on a journey so people can see the thought that goes into this and what things I'm analysing when I see the initial results from systems.  I didn't get to where I'm at with the UK systems by just developing a system and seeing how it performs when it goes live. I work hard to ensure that I understand as much about the system as I can and therefore, when it goes live, I know pretty quickly whether it is performing in line with expectations but more importantly, whether I think the system is worth following going forward.  


I'll be back with system E3 next.  Looking forward to this system as I still remember the feelings I had when I analysed system 8 a few years ago! Hopefully E3 will give me the same feeling........

No comments:

Post a Comment