Hands on use of Odds Wizard including results

[pswirly]

Below is the results of using Odds Wizard from mid-December until now.

I've been a subscriber of Odds Wizard for a couple of years and my subscription is due for renewal in June. Should I resubscribe?

Here are my findings.

CRITERIA

1X2 only
2007 algorithm (as recommended)
Applying trend hints

BETS

TOTAL - 1762
WINS - 595
LOSSES - 1167

PROFIT/LOSS -128.68
PORIFT/LOSS % -7.3%

HOME WINS

TOTAL - 831
WINS - 350
LOSSES - 481

PROFIT/LOSS -57.32
PROFIT/LOSS % -6.89%

DRAWS

TOTAL - 201
WINS - 46
LOSSES - 155

PROFIT/LOSS -11.66
PROFIT/LOSS % -5.80%

AWAY WINS

TOTAL - 730
WINS - 199
LOSSES - 531

PROFIT/LOSS -59.7
PROFIT/LOSS % -8.17%

I will await any interest in a discussion before publishing some filtered results.

[push.button.bang]

i have done similar trials just on the American sports, basically as i have zero interest in soccer.

the USA sports usually show more positive results, however getting the odds to update the BWIN always seems to be the biggest issue.

do u have any stats on any USA sports?

[pswirly]

Unfortunately, no. I am the opposite of you in that I have no interest in US Sports!

Perhaps the 2007 algorithm is (as I suspect) not as effective as 2006.

The loss isn't heavy in percentage terms but how long a sample is required in order to conclude the software is not profitable?

[Alex]

PSWIRLY:
Dig in the previous posts to find out how successful can be using Odds Wizard. Otherwise it couldn't be so popular amonst the punters and so unpopular amongst the bookies. The latter are really afraid of this program: those who understand well how dangerous Odds Wizard is simply reject applications of our web site to be their affiliate! Recent denial from Eurobet is a graphic evidence.

However you might know that every betting strategy does imply that there is a chance of getting lost. In particular even the accurate use of Stake Wizard bet optimizer with the confidence probability 95% assumes that unhappy 100-95=5% may be yours.

Your claims about systematic losts require more detailed studies, and we naturally need your spreadsheet to look closer into your specific problem. Please send it to me asap.

[jazzfish]

Interesting topic. I've placed a large number of bets using OW (a number of thousand) and overall, results are profitable. However, it's important to note a number of findings:

1) The difference between OW odds and line odds is not a good predictor of actual profit.
2) The stake wizard (in my experience) does not improve profitability.
3) Trend hints do not improve profitability - I have done extensive analysis on this.
4) Results are very variable. You may go a year without making a profit if you bet blindly where OW odds are greater than line odds. So use a big bank.
5) I believe betting blindly has its issues as with any statistical system. Of course OW does not consider injuries, team news etc so you should research this yourself.
6) Where line odds and OW odds are extremely different beware. This is probably due to injuries, motivation etc. Matches with extreme differences are the least profitable in my analysis.

Overall I must stress that OW is a good program and I feel that Alex et al are very transparent and honest in the way they present it. However, it is a statistical model and therefore must be seen as such. It consequently has a number of blind spots that you should be aware of. I think my profits over two years showed 5% which is quite impressive for a statistical model over a long period of time. I should add that these results were statistically significant using chi square analysis, so you can be confident that the positive results were not a consequence of variance. However, for me the yield was too low to justify the time I was spending on selecting bets. If I did not have a job I might have another look at the possibilities.

One final point. No betting system can gaurantee positive results, however good it is. Variance means unexpected events occur all the time and this is what makes the world go round so make sure you bet with a conservative bank that allows long term losses.

Hope this is helpful.

JF

[pswirly]

No one is questioning anyone's integrity, let me just state that from the start.

From May 1st I have decided to continue the list without applying trend hints.

I will post more soon.

Alex, I will send the spreadsheet to you tomorrow.

[odds software]

Hi Pswirly

Interested in hearing more about your results so I can better understand them

eg

1 - What leagues does the above refer to?
2 - odds used to determine if value or not. Are you using Bwin only?
3 - Odds for actual betting.. are you using bwin only or shopping around?

Best wishes
Mick

[Alex]

Pswirly:

Thank you for sending me your XLS files with results of betting.

Upon extensive studies, eventually I have found the reason of your losts - wrong calculation of stakes and incompleteness. Sorry for the delay as it took a lot of time.

You might read the Odds Wizard manual, the topic "Stake Wizard Guide -> Math background". It clearily states that even though you have met the criterion Line > Fair, you can fall in red when the stake is either too small or too high. Just look at the convex curve (P vs x0).

By the way, why do your files miss many games with potential profit?

For example, you could add the following profitable game to Europa League
events of December 16-17, 2010:

NK Dinamo Zagreb vs PAOK Saloniki (2) F=3.49 L=4.5
NK Dinamo Zagreb vs PAOK Saloniki (2X) F=1.66 L=1.91

Your file "Stake.xls" is definitely incomplete with respect to events meeting criterion Fair < Line and other bet types, e.g. 1X, 2X that may be even more profitable.

Let's look closer at your results. The total sum of your 2565 stakes is GBP 8678.84, and the resulting balance is negative: GBP -513.76.

Applying the simplest Kelly strategy gives more favourable results which are presented in the last two columns highlighted by light green. Total amount of stakes for the same 2565 events is GBP 8423.05, and the balance is still negative GBP -380.13. I attach here modified version of your file Stake.xls.

Note that Kelly strategy is the simplest one. Since 2007, Stake Wizard 4 uses the theorem on the optimal distribution of the bettor's bankroll with more advanced calculations and much greater results.

I have also included a spreadsheet where the same results are filtered if they are above 25% and below 50% in terms of difference in fair odds and line odds and this in itself produces profitable returns.

Your another XLS file "Above 25-Below 50" shows positive balance +GBP 1581.21 even without applying Kelly or more advanced strategy. It is also attached here with an added TOTALS line at the bottom. Positive balance is well expected here as suspicious events with the great difference between Fair and Line odds were cut off (according to Generic tips http://www.newhavensoft.net/forum/viewtopic.php?f=5&t=6) while events with the small difference Fair/Line result in small stakes/wins/losts, and thus make no significant contribution in the total balance.

[jazzfish]

Hi Alex,

Just wonder why you chose under over 25% and under 50% as your benchmarks? The problem is that when done after selections it could be seen as back fitting and it does not necessarily follow that the trend would continue in future selections. Also, simple Kelly does not improve the results on the filtered selections but actually makes them worse which is worrying because it suggests a weak correlation between level of edge and level of profit.

Regards

JF

[Alex]

I don't use this benchmark. I just refer to the Pswirly's spreadsheet. Also, simple Kelly was never supposed to be something good. It was retained in Stake Wizard in order to compare results with those of regular mode. At the same time, SW regular mode uses numerical calculations which are hard to be reproduced in the XLS sheet for hundreds of events. Thus the major sense of XLS sample file is to show that use of certain edge range has some positive financial result due to the Odds Wizard's ability to predict events better than sportsbooks, presumably in that specific range.