Return to Oz and the Michael Wray Experiment.

[merlin]

Re: Return to Oz and the Michael Wray Experiment. ..... sure, no problem:- Barrow 0-0 Cambridge 2-0 Chester C 1-2 Ebbsfleet 0-2 Gateshead 1-2 Grays 1-3 Kettering 1-1 Kidderminster 2-0 Salisbury 0-1 Stevenage 2-1 Tamworth 1-2 York C 4-0 Hope this helps. ;)

[KThom]

Re: Return to Oz and the Michael Wray Experiment. Cheers, incidentally I have:

Barrow	1	0
Cambridge	2	0
Chester	2	2
Ebbsfleet	1	3
Gateshead	1	2
Grays	1	3
Kettering Town	2	1
Kidderminster	2	0
Salisbury	0	0
Stevenage	1	1
Tamworth	1	2
York	4	0

[KThom]

Re: Return to Oz and the Michael Wray Experiment. Incidentally - the odds on York are tumbling, I matched at 2.06 - but now trading at 1.92 on betfair

[merlin]

[gooner steve]

[merlin]

Re: Return to Oz and the Michael Wray Experiment. ..... thanks GS :welcome to the thread and may all your re visits be profitable ones! :ok

[clarky101]

[KThom]

Re: Return to Oz and the Michael Wray Experiment. Hi Merlin, I calculate 5/10 correct results for me and 3/10 for you - but you did nail the correct score in 1/10 compared to my 0/10. Make of that what you will??

[merlin]

Re: Return to Oz and the Michael Wray Experiment. ..... I wasn't aware that we had entered into any competition KT. These are purely the scores given by Michael's sheets and the only result I was concerned with was York City and not as a score but as a home win.This duly obliged so my score if any was only one correct from one selection.When I penned the predictions I thought I was doing you a favour and not for any comparison, after all this thread is really all about the Michael Wray spreadsheets. Thanks.;)

[merlin]

Re: Return to Oz and the Michael Wray Experiment. ...... OK it's time to catch up on any midweek selections for the 3-0 system, and these have been highlighted for Tuesday's round of Blue Square Conference matches:- Cambridge United (H) 1.58 Rushden & Diamonds (A) 2.18 Hoping to continue our opening good run. ;)

[KThom]

Re: Return to Oz and the Michael Wray Experiment.

..... I wasn't aware that we had entered into any competition KT. These are purely the scores given by Michael's sheets and the only result I was concerned with was York City and not as a score but as a home win.This duly obliged so my score if any was only one correct from one selection.When I penned the predictions I thought I was doing you a favour and not for any comparison, after all this thread is really all about the Michael Wray spreadsheets. Thanks.;)

Ok merlin - well we haven't entered into any comparison. I was just thinking aloud. As mentioned before i believe my working models are more accurate than the mw sheets, I can also use these for a wider range of leagues. Perhaps I could share any suggested 3-0+ games for France, Greece, Sweden or wherever.

[muppet77]

Re: Return to Oz and the Michael Wray Experiment. Kt. Would you care to share how your sheets work? Do you have any evidence that they are better than mws sheets? I dont doubt your claim its just that i have an inside knowledge into how they calculate scores and wondered how yours worked.

[KThom]

Re: Return to Oz and the Michael Wray Experiment.

Kt. Would you care to share how your sheets work? Do you have any evidence that they are better than mws sheets? I dont doubt your claim its just that i have an inside knowledge into how they calculate scores and wondered how yours worked.

HI muppet - I also have knowledge of how mw's sheets work. Mine are on a very similar way (using past 5 h2hs, recent form and league positioning) however the one glarring omission from mw's formula is a running goal rate. This is fixed so isn't league dependent whereas mine is. There is obviously a lot more complex formulae in both sheets but these are the most evident comparisons.

[muppet77]

Re: Return to Oz and the Michael Wray Experiment. Yes i knew that you knew kt judging by the similarities between your scores and wrays.

[Kanga]

Re: Return to Oz and the Michael Wray Experiment. Interesting stuff this guys ..... obviously 1 weeks results mean very little .... is there any chance of back testing one against the other for last season ???? If not a running comparison would be useful.

[KThom]

Re: Return to Oz and the Michael Wray Experiment. Kanga I could certainly test this for last season. The thing that needs to be established is - can any money be made from my (or anyone elses) score predictions? Betting on games over 3-0 would be one way, but I still don't like the level stakes regardless of odds approach. I'd like to apply some form of staking or confidence calculation to it. I am currently running a thread on 0-0 predictions as back tests showed good value in them (despite the obvious long loosing streaks). I wouldn't want to rock the boat here by making claims or comparisons to what is clearly a long succesful thread. Afterall were all in it for the same goal

[Kanga]

Re: Return to Oz and the Michael Wray Experiment. Hi KT, If you could backtest your last seasons 3-0 (+) predictions and work out if they are profitable to level stakes that would be awesome just as a comparison ..... Maybe you could check to see where MW has highlighted a 3-0 and yours hasn't and vice versa (if it's not too time consuming). Merlin has a number of contributers making additional use of the MW sheets posting on this thread, if you've managed to improve upon the sheets and can help improve the ROI then great in my book :loon As long as the chosen subset makes a profit to level stakes then I'm sure a better staking plan can be worked out. :ok

[KThom]

Re: Return to Oz and the Michael Wray Experiment. Kanga. I'll look into this vs premiership and see what results occur

[KThom]

Re: Return to Oz and the Michael Wray Experiment. Kanga, using the premiership alone - the last 6 seasons yield a negative return on both mine and the mw spreadsheet (betting on scorelines deemed 3-0 or greater and discounting the first 6 and last 3 games of the season). Last season also saw an overall loss on the 3-0 selections for both generated scorelines. I've looked along these lines before, and posed a similar question on a different thread which was unanswered - Assuming the 3-0 (or any other prediction method averages a 63% s/r - equivilant odds of 1.58) should you? a) You should only back selections over 1.58 with level stakes b) You should back all selections with level stakes and "hope" the average price is above 1.58 (afterall there is no way of knowing future selections odds) c) Introduce a staking plan which increases the stake the higher the odds (Ansell) d) Customise your own staking plan to return the same amount regardless of odds (For instance odds of 1.50 you'd stake £100 to win £50, odds of 2.00 you'd stake £50 to win £50) e) any other suggestions? What do you or anyone else think? What is the mathematically the correct procedure? I still think level stakes betting vs the bookies and all their betting models and statistical analysis available is dangerous.

[barrhead73]

Re: Return to Oz and the Michael Wray Experiment. Back to the EPL Unders which has got off to a slow start. Two games for this weekend, thought i'd get the early prices on them,

Date	Selection	Stake	Dec Odds	Win?	Return	Profit	Cum Profit
19-Aug	Liverpool Stoke	£ 2.50	2.05	n	0.00	-£2.50	- 3.41
12-Sep	Blackburn Wolves	£ 2.50	1.724
12-Sep	Pompey Bolton	£ 2.50	1.727

[muppet77]

Re: Return to Oz and the Michael Wray Experiment. good research KT - i think that the question of "do MW's sheets find value?" has been raised before. i am always uneasy about backing against something that i have no idea of the true price for. if you could relate prices to MW's picks then perhaps his yield would be even higher?

[Grex]

Re: Return to Oz and the Michael Wray Experiment. Hi Merlin..... you may remember some contributions from me to this thread a couple of seasons ago. Anyway, it's good to see the system is still going strong... and congratulations on a winning start:clap.... and good luck with your selections for tomorrow night.:hope Seems like you have a bit of friendly "competition" from Mr. KThom..... perhaps you should combine forces. You know what they say.... two heads :wall:wall are better than one !

[merlin]

Re: Return to Oz and the Michael Wray Experiment. .... Iv'e read the interesting debate above Gents on staking plans and odds etc and all ideas are most welcome. Over the years I believe we have had plans put forward, and I myself believe there is room for improvement on any profits made by the 3-0 system be it KT's or MW's.I personally think that some form of progressive staking would be benificial, where returns and stakes would be greater with each winning selection.I'm always open to positive suggestions which would be benificial to us all.All ideas :welcome

[Kanga]

Re: Return to Oz and the Michael Wray Experiment.

Kanga' date=' using the premiership alone - the last 6 seasons yield a negative return on both mine and the mw spreadsheet (betting on scorelines deemed 3-0 or greater and discounting the first 6 and last 3 games of the season). Last season also saw an overall loss on the 3-0 selections for both generated scorelines.[/quote'] KT ... just to clarify are you testing this as betting on a correct scoreline or just the plain win indicator as Merlins thread ? ... If the later as I suspect, then it's interesting that the Prem should show a negative return when the system has returned a positive value overall for the last few seasons. Personally in answer to this conundrum I would say that your average strike rate is 63% therefore your average odds has to be >1.58 to make a profit. a) You can't only stake on selections where the available odds are >1.58 as the odds available may not be value e.g a true price of 1.72 backed at 1.58 overtime will lose whereas a true price of 1.40 backed at 1.55 will. b) Not ideal as you say but Get the best price you possibly can, the past seasons have shown that there is an edge to the sheets slightly greater than average bookies prices. c) Staking plans need very careful research some popular recovery staking plans may even make you go bankrupt before recovery - a lot of data is needed. d) Customizing according the strike rate and odds is the way to go IMHO. e) I have access to a program called The Staking Machine if you can pm me I'll can tell you how I need the data and I can run it through ... the website is well worth a look I think you would find it interesting. :ok

[Kanga]

Re: Return to Oz and the Michael Wray Experiment.

"do MW's sheets find value?" has been raised before. i am always uneasy about backing against something that i have no idea of the true price for. if you could relate prices to MW's picks then perhaps his yield would be even higher?

Astute observation M :ok It would be very helpful to know if any value or probability could be placed next to a scoreline.

[KThom]

Re: Return to Oz and the Michael Wray Experiment.

KT ... just to clarify are you testing this as betting on a correct scoreline or just the plain win indicator as Merlins thread ? ... If the later as I suspect, then it's interesting that the Prem should show a negative return when the system has returned a positive value overall for the last few seasons.

That is correct, I was testing the result not the scoreline. I used the average odds as supplied by football-data so this may have varied slightly from actual odds taken at the time. It seems the deciding factor in all this is in the probability vs odds. I can align a probability of outcome to any scoreline (1.58 for over 3-0's etc) but by doing this you are throwing all scorelines into one basket and deeming their chance as equal. While this is all well and good, what is not taken into account by my por any other formula calculation sheet are - local derby's, injuries, manager changes, and what is at stake for both teams (This has somewhat been alleviated by avoiding the first 6 and last 3 games from each league). Therefore calculating a "true price" is not as easy as it sounds.

[clarky101]

Re: Return to Oz and the Michael Wray Experiment. Anyways..... nothing from me midweek.....see ya thursday..

[Kanga]

Re: Return to Oz and the Michael Wray Experiment.

That is correct' date=' I was testing the result not the scoreline. I used the average odds as supplied by football-data so this may have varied slightly from actual odds taken at the time.[/quote'] Interesting .... I know that Merlin looks around for the best price .... would it be possible to check against betfair or bet365 prices it may make all the difference. No you don't want to do that ... what we need is a probability for each event ..... In the MW formula's is there a way of calculating this ? Agreed finding a true price is going to be difficult to say the least ... a quest for the holy grail (See what i did there - :loon ) I had a thought today that if the overall roi of the system is 10% then the bookies price quoted less 10% could be taken as the true price and staking worked out on this basis ... unfortunately this isn't totally accurate but could be useful.

[KThom]

Re: Return to Oz and the Michael Wray Experiment.

Interesting .... I know that Merlin looks around for the best price .... would it be possible to check against betfair or bet365 prices it may make all the difference.

Ok, compared with the best prices (using football-data) and the PL still showed a loss last season and for the last 6 season's combined.

No you don't want to do that ... what we need is a probability for each event ..... In the MW formula's is there a way of calculating this ?

There is a way yes, would just need to calculate how to do it and what to base it on. I still feel (with my score system anyway) the best value is in the correct score betting (especially taking into account the inflated prices available on exchanges). Not All scorelines, but selective ones (see my correct score thread). The problem is back testing is not as easy due to the un-availability of previous correct score odds. Also with correct score betting - the loosing streaks will be longer, and strike rate lower - but I think ultimately over time (100 bets for instance) this is the more profitable way to use these sort of generated scorelines.

[Fast Eddie]

Re: Return to Oz and the Michael Wray Experiment. The problem with correct score betting, which I had a little go at last season for a short while, is as KThom says above, the long losing streaks and the higher risk you are exposed to, which in turn means having to use much smaller stakes to reduce the risk of bust to an acceptable level, so even if you were producing a much higher yield than the straightforward win bets of say the Merlin 3-0 system with it's odds-on prices, your overall profit wouldn't be necessarily higher, and for the amount of time you would be putting in, unless the starting bank was significant it may not be worth the effort and risk. I might have a little dabble this year using a mix of the 3-0s, a few correct scores 1-0s probably, and maybe the unders as well, but I haven't decided exactly what yet so I'll get back to you if I get going on that over the coming weeks. The Holy Grail would be to get a bet for every prediction but I think that may be a bit too hopeful. I will definitely be running with Merlin's 3-0s though, although at much lower staking levels, probably 3% of bank per bet, adjusting as the bank falls or rises in 5% increments.