question about Performance of Odds Wizard ....

[andrix]

please excuse me for bad english language...
i have a question....
please take a look at this screenshoot....
in this i know that when a game have calculated OW probability of => 45% , have a success rate of 58.7%
the problem in this calculation is that => 45% include games with 60-70-80% success probability,
in this calculation thereisa possibility that in reality the games with calculated probability of ,ad example,
of 45.8% have a success rate of 40%.....
in this example, i don't know what success rate have a game with exact probabilty of 45% !!
in other world , there is a possibilty to know success rate of a range of calculated odds ?
i want,i like this:
when calculated odds are in range min 45%-max46% , success rate is 52%
or
when calculated odds are in range min 65%-max70% , success rate is 73%
thanks for any clarification,and again excuse me for bad english....

[Alex]

Presented table indicates an integral distribution of actual frequencies of wins and wins+draws which is absolutely different from differential distribution. Why integral? Because 1) it was requested to be presented first historically 2) it can be used even for the leagues with poor statistics (in the sense of available number of games).
Now you are asking for differential distribution. Here is requested figures for all leagues:
http://www.newhavensoft.net/cp/Allstats_dif.htm
and corresponding graphs:

There is quite definite answer on you question:

Differential success rate for wins ~=~ Computed probability

as it follows from the very tight statistical relationships.

[andrix]

it's possible to have this stats for all single national Championships ?
thanks in advance.....

[Alex]

As you can see, global stats for all leagues shows quite significant variance at the levels > 85% for home wins and > 70% for away wins. How do you think why? Because of the rather poor stats at those levels even for all leagues taken together. For a single championship, e.g. for ADEG Erste Liga (Austria), we would have the following graph:

The variance is too high to be able to extract any reliable or significant information. It might be very confusing as little changes in computed probability level may result in absolutely casual drastic changes in success rate simply due to poor statistics.

[andrix]

uhm....

i want stats on odds calculated and relative success rate....for single nations....too much difference from italy b and bulgaria....

and if amplify the range ?
example 40 to 43 ,44 to 47 , etc ?
or 40 to 45 , 46 to 50 ,etc ?
you have reilable stats in this case ?
thanks again for your great support....

[Alex]

We have introduced differential distribution of success frequencies for both 1X2 and U/O in the Tools -> Performance analysis -> More. Make sure to uncheck the box 'Integral' and recalculate stats to view differential values. New build 2.1.0.506 can be downloaded from http://www.newhavensoft.net/dload.htm.

[andrix]

We have introduced differential distribution of success frequencies for both 1X2 and U/O in the Tools -> Performance analysis -> More. Make sure to uncheck the box 'Integral' and recalculate stats to view differential values. New build 2.1.0.506 can be downloaded from http://www.newhavensoft.net/dload.htm.

dear alex,
where i can download build 2.1.0.506 ?
the last version at url http://www.newhavensoft.net/dload.htm is 2.1.0.493
and this version don't have the box 'Integral' to uncheck ...
thanks in advance

[Alex]

The last version/build at the mentioned URL is indeed 2.1.0.507. Could you please reload the file OW210.exe as 2.1.0.493 might be just web cache 'fantom'.

[andrix]

ok,i now have last version...
thanks for your support....