Dustys Aways

[lunatism]

Re: Dustys Aways would posting the "euro-selections" in this format be ok for you gooner and dusty?:

DATE	TIME	LEAGUE	MATCH			FORM POSITION	LEAGUE STRIKE RATE	ODDS
15/02/08	19:00	French Ligue 2	Chamois Niortais Football Club	v	Amiens Sporting Club Football	20 v 12	83%	3.35

that's what i'd be trying to have them coming out like when i do the form extractor thing.....unless there is any other preference

[happygooner]

Re: Dustys Aways

would posting the "euro-selections" in this format be ok for you gooner and dusty?:

DATE	TIME	LEAGUE	MATCH			FORM POSITION	LEAGUE STRIKE RATE	ODDS
15/02/08	19:00	French Ligue 2	Chamois Niortais Football Club	v	Amiens Sporting Club Football	20 v 12	83%	3.35

that's what i'd be trying to have them coming out like when i do the form extractor thing.....unless there is any other preference

Certainly looks fine to me but obviously it will have best odds added as well presumably (from what we discussed). Just one thing: why is it too big for my screen??

[lunatism]

Re: Dustys Aways

obviously it will have best odds added

so we say that ALL the odds are coming from betfair, but these are the best odds we can find for such a game? (when i said best odds before, i didn't mean 1X2, i meant the best odds the the away win)

just one thing: why is it too big for my screen??

erm......you have a small screen? :tongue2 no - it's what firefox does to my tables and i don't know how to remedy it, but i'll play around and figure it out before i finish the excel thing anyway

[happygooner]

Re: Dustys Aways

so we say that ALL the odds are coming from betfair' date= but these are the best odds we can find for such a game? (when i said best odds before, i didn't mean 1X2, i meant the best odds the the away win) erm......you have a small screen? :tongue2 no - it's what firefox does to my tables and i don't know how to remedy it, but i'll play around and figure it out before i finish the excel thing anyway

I only posted up the 1x2 as it was the easiest thing to do taking them all directly from odds comparison with copy and paste. Best odds for away win only is all good:ok

[lunatism]

Re: Dustys Aways

Best odds for away win only is all good:ok

so when i finish the form extractor (probably some time after i start it :unsure), it will produce nice little tables that are easy to post up here (a simple copy/past job), then i guess me/you/dusty can post these when we notice that it hasn't been done already....i'll go into the finer details of the DFE (saves calling it form extractor every time d obviously stands for dusty) with you two later on. For example....wouldn't even need any stats knowledge to post the stats if the spreadsheet calculated that all for you :D

[dustbin2]

Re: Dustys Aways Format for euros fine by me Loon:ok

[happygooner]

Re: Dustys Aways

so when i finish the form extractor (probably some time after i start it :unsure)' date= it will produce nice little tables that are easy to post up here (a simple copy/past job), then i guess me/you/dusty can post these when we notice that it hasn't been done already....i'll go into the finer details of the DFE (saves calling it form extractor every time d obviously stands for dusty) with you two later on. For example....wouldn't even need any stats knowledge to post the stats if the spreadsheet calculated that all for you :D

All sounds like taking candy from a baby (not that i've tried such a thing:unsure)...

[2good4u]

Re: Dustys Aways Just one question: Why the game Koblenz-St.Pauli is marked red when their form in the last six games is exactly the same (1 W, 4 D, 1 L)? Or I am missing something?

[lunatism]

Re: Dustys Aways

Just one question: Why the game Koblenz-St.Pauli is marked red when their form in the last six games is exactly the same (1 W, 4 D, 1 L)? Or I am missing something?

firstly, the selection process ONLY looks at the form position, and i gues syou noticed that even though same for WDL, the differing goal difference means that Koblenz are 13th, while St Pauli are 11th in the form table (meaning go for the away win - dusty specified that as long as they are ahead in the form table, you take them....that's it :D) but i think you misunderstand the reason for the highlighting of the matches, this is to do with the strike rate of that particular league, not really much to do with the individual form of those particular teams in those matches (i meant to put "league strike rate", but that made the table even wider :tongue2), but if you look at the sample table i posted that has the Amiens game in it, you will see that particular discrepancy has been "seen to" ;) sorry for the confusion

[happygooner]

Re: Dustys Aways

Just one question: Why the game Koblenz-St.Pauli is marked red when their form in the last six games is exactly the same (1 W, 4 D, 1 L)? Or I am missing something?

That reminds me of another question I had to put to you loon... Was speaking with another highly respected forum member 'meatman' yesterday who specialises in France. says that if the system had gone last week in France ligue 2 would have been a disaster as not 1 away win on 08/02/2008. Checked this and he's right. Do you think it's possible that the system just didn't highlight one French ligue 2 match last week or could there be a bug................? The floor is yours loon:dude

[lunatism]

[lunatism]

Re: Dustys Aways actually, that was for the 10th...bear with me, i'll get data for the 8th

[2good4u]

Re: Dustys Aways Thanks for the explanation. I misunderstood it (thought at least one point more is needed to fit the criteria, not the goal difference) but it is clear now. Speaking of the red part, I realized it the moment I clicked submit button :D and since I saw you were online did not bother to edit it cause I knew you will answer ASAP.

[lunatism]

Re: Dustys Aways in fact, for the 8th, there were 3 selections - they all lost :(

DATE	LEAGUE	MATCH			RESULT
08/02/08	French Ligue 2	Clermont Foot Auvergne	v	FC Libourne Saint-Seurin-sur-L'Isle	1	1	WRONG
08/02/08	French Ligue 2	La Berrichonne Chateauroux	v	Dijon Football Côte d'Or	2	0	WRONG
08/02/08	French Ligue 2	Sporting Club de Bastia	v	Espérance Sportive Troyes Aube Champagne	1	0	WRONG

now time for me to launch a super large scale investigation into the matter and produce my results as soon as possible (which may take a while because i've got a whole load on my plate for the next few days...not helped by A&E waiting times i might add!). At least we're still (or should be) in the paper trialling stage and so any issues that do arise can be dealt with before people start spending all their cash on it! another reason it will take a while is that when in a "rush" i like to spread my code all over the place...now i wish i hadn't for error checking is a no fun as it is ....my advice - watch this space

[happygooner]

Re: Dustys Aways

in fact' date= for the 8th, there were 3 selections - they all lost :(

DATE	LEAGUE	MATCH			RESULT
08/02/08	French Ligue 2	Clermont Foot Auvergne	v	FC Libourne Saint-Seurin-sur-L'Isle	1	1	WRONG
08/02/08	French Ligue 2	La Berrichonne Chateauroux	v	Dijon Football Côte d'Or	2	0	WRONG
08/02/08	French Ligue 2	Sporting Club de Bastia	v	Espérance Sportive Troyes Aube Champagne	1	0	WRONG

now time for me to launch a super large scale investigation into the matter and produce my results as soon as possible (which may take a while because i've got a whole load on my plate for the next few days...not helped by A&E waiting times i might add!). At least we're still (or should be) in the paper trialling stage and so any issues that do arise can be dealt with before people start spending all their cash on it! another reason it will take a while is that when in a "rush" i like to spread my code all over the place...now i wish i hadn't for error checking is a no fun as it is ....my advice - watch this space

Haha....you're nuts:lol Or crazy in the coconut as they say:tongue2

[lunatism]

Re: Dustys Aways ok guys - want the good news or the bad news first? i'll go for bad - so we can finish on good ad leave a nice taste in the mouth ;) first, my preliminary investigations showed a really silly error on my part, basically all the strike rates are opposite of what they should be (i was dividing number of losses against total number played instead of wins against games played), which means that instead of a strike rate, you are looking at a failure rate :$ and so the overall strike rate of 67% from 597 games, becomes 33% from 597 games....yield of course goes minus,etc that is a HUGE mistake by me, but not to worry (unless you put real cash on the selections this weekend :$) i have worked out a solution to that anyway (and would love your help on improving it) here it goes - in comparing my LSS system to dustys aways, if i add the following criteria dustys (from info generated by my system) we get around a 59% strike rate, with a yield of 46% (from league sample sizes of 15+ with an established strike rate of 50% or higher. based on stats from 36 games), or a strike rate of 60% and a yield of 51% (from just an established strike rate of 50% or higher - based on stats from 124 games)......and overall (all the "top" leagues) a strike rate of 45% with a yield of 13% from an overall sample size of 243 (110 winners). So as long as that pattern (especially the 59%SR / 46%Y) continues, there should be similar yield/strike rate as the weeks progress. So to go over it again - the bad news is that the original strike rates posted were backwards, which means leagues previously thought to be "crap", now appear to be a lot better, and vice versa and the GOOD NEWS is that i've spliced dustys with LSS and come out with a similar strike rate and yield, albeit from a sample size that is either 6% (the 36 selections) or 20% (the 124 selections) the size of the original sample size 13% yield from 243 selections is also not a terrible thing (of course, not as good as before, but still - makes you cash), and so the we should not lose faith in the system (or dusty).....just perhaps lose faith in my ability to see the difference between a win and loss when calculating strike rates :wall sorry lads, but at least the positives are good positives.......i still think that dustys system will bring in wads of cash for us all - BUT (and we'd all said this before anyway), this is paper trialling. this "investigation" has found the potential banana peel as far as i'm concerned, and now we should just watch as the sample size builds up before putting it "live"...... i put my hands up for this one...gross error on my part:cry

[dustbin2]

Re: Dustys Aways Loon not to worry m8, thats what paper trialing is all about. Would this be why the league I thought could be more profitable (league 2) showed not to be on your system? Knowing you , you will be now working on what leagues are the best. We must remeber that my simple theory of betting above evens and above, means that really just over half the selections need to come in every week to make a small consisitent profit. Still, lets see what happens tomorrow and review things then.;)

[happygooner]

Re: Dustys Aways oh shit, I just put my house up for a 2nd remortgage:puke :rollin:rollin:rollin:rollin:rollin:rollin :rollin :rollin :rollin Fairplay to fessing up though loon.....can I come live with you:unsure Let's have a look at those stats then Mr Statman....you know you're dieing to post a couple pages of graphs, spreadsheets and the like:lol

[lunatism]

Re: Dustys Aways

Let's have a look at those stats then Mr Statman....you know you're dieing to post a couple pages of graphs' date=' spreadsheets and the like:lol[/quote'] right now i'm fiddling and making them look as pretty as possible...there'll be here shortly :)

[lunatism]

Re: Dustys Aways

Loon not to worry m8, thats what paper trialing is all about. Would this be why the league I thought could be more profitable (league 2) showed not to be on your system? Knowing you , you will be now working on what leagues are the best. We must remeber that my simple theory of betting above evens and above, means that really just over half the selections need to come in every week to make a small consisitent profit. Still, lets see what happens tomorrow and review things then.;)

in short - YES! league two turns out to be one of the most profitable leagues, both when i noticed the original error (just invert the strike rate i initially posted), and also when i applied all those additional criteria (which i have forgotten to actually mention so i do it now....basically if my system says the most likely result will be an away win AND the rating for the match is 10 or higher) it turns out a 65% strike rate from 20 matches - which is the best performing league that reaches the sample size cutoff point. so time for a little dance from you....just shows your original research into L2 is turning out to be positive (for the 36 days that i backtested anyway ;))

[dustbin2]

Re: Dustys Aways League 2 rocks after all:dude Pity about selections Ive posted for this weekend but we will see how it goes then I may revert back to league two. Ive a feeling we will still at least break even:hope Still got a good feeling about this system tho

[happygooner]

Re: Dustys Aways

and these are the stats if you apply the selection criteria i mentioned

	bets	wins	losses	strike	staked	return	profit	yield	confidence
SR ABOVE 50%, SAMPLE 15+	36	21	15	0.5833	300	437.5	137.5	46%	99%
SR ABOVE 50%	124	75	49	0.6048	300	453.63	153.63	51%	100%

well dusty, now that you can see all those leagues and their stats - what do you think? i guess it would be too early (due to small sample sizes) to reevaluate the leagues you include, but it provides more talking material anyway

Hey loon, explain the difference in criteria that changes 36 selections to 124? Just a little confused right now:unsure Also....how are you staking 300 (identical amounts) on either 36 bets or 124?

[lunatism]

Re: Dustys Aways

Hey loon, explain the difference in criteria that changes 36 selections to 124? Just a little confused right now:unsure

basically, the 36 all have sample sizes of at least 15, 124 don't. i've sent you an email with the "proper" selections so you can find some odds anyway. basically, if you'd taken only those leagues and put in a criteria saying the league must reach 50% strike rate, then THAT is what the yield would turn out

[dustbin2]

Re: Dustys Aways It also shows us that scottish league three might be worth a look at in the future:loon . Anyway of out for a bit but still excited for tomorrow and about the system. Gr8 work by all involved, especially the mad stat man ( loon ):ok

[merlin]

Re: Dustys Aways ..... been watching this thread unfold with interest Dusty and wish you luck with selections. Your early rules indicate that only selections at evens and above will be chosen.Surely this will mean that the majority of selections will be selected as away odds tend to be higher anyway.The evens odds did apply to the aways, I have not got this wrong have I? I was also rechecking your selections for this weekend and thought that you might be interested to know that only three of these had higher rating figures from the RFO taking RFO Home rating v Away rating, indicating that perhaps the other selections did not look as solid:- Rochdale + 10 Notts Co + 28 York C + 46 Might mean absolutely nothing, but interesting nevertheless!:ok

[merlin]

Re: Dustys Aways .... As a matter of interest your above suggestion that Scottish Division 3 could be looked at, reveals that two were highlighted East Fife and Montrose, but both look as if they are ruled out by the low odds factor.;)

[lunatism]

Re: Dustys Aways

.... As a matter of interest your above suggestion that Scottish Division 3 could be looked at' date=' reveals that two were highlighted East Fife and Montrose, but both look as if they are ruled out by the low odds factor.;)[/quote'] is there any possibilty of providing (or getting) past rating figures for matches from the 6th january so i can do some backtesting to see how the strike rate and yield looks? perhaps a combination of that and dustys could mean a larger sample size too! let me know.....

[happygooner]

Re: Dustys Aways Ok first things first, lets get tonights selections out there!

15/02/08	Belgian Jupiler League	Koninklijke Atletiek Associatie Gent	v	Cercle Brugge KSV	7 v 1	67%
15/02/08	French Ligue 2	Football Club Gueugnonnais	v	Le Havre Athletic Club Football Associat	18 v 2	100%
15/02/08	German 2.Bundesliga	FC Carl Zeiss Jena e.V.	v	1. FC Köln	7 v 4	0%
15/02/08	Dutch Jupiler League	Stormvogels Telstar Combinatie	v	Stichting Helmond Sport	18 v 4	40%
15/02/08	Turkish Turkcell Süper Lig	Caykur Rizespor	v	Fenerbahçe Spor Kulübü	12 v 1	100%
15/02/08	Welsh League Of Wales	Llangefni Town FC	v	Airbus UK FC	18 v 13	43%

Odds for all except the Welsh game can be found earlier in the thread. Perhaps now the sample size we are working from is smaller we include games under evens? And from the lower Conference leagues as well as Welsh league?? Airbus odds 2.40

[happygooner]

Re: Dustys Aways Err....loon, we're using games from leagues with a 0% strike rate, is that right??:unsure Forgive me for being stupid but i'm still not clear on the 36 bets and the 124 bets either:unsure? 2 different sets of criteria being laid down??? How come both have equal amounts staked??

[happygooner]

Re: Dustys Aways And the selections you've sent me. Which camp do these fall under? The 124 bets with a sample size of less than 15? When you say sample size of less than 15 do you mean less than 15 selections per league? Is there any way at all you can break this down so it's easier to take in. Which leagues etc....how many selections for each...that type of thing....? And which leagues have sample size of 15+ Presumably if you are only getting 36 selections it can only be a maximum of 2 leagues?? Am I on the right track or so far out of leftfield it's untrue:lol???