depreciation Period and performance analysis

[papadas]

Hello Alex,

I have give Odds wizard another change by searching the optimal depreciation period.

What I can see that of the majority of the leagues setting the depreciation period around 200 days and keeping the trees also to 200 days default creates very charming results in the Success Rate% and also the entropy reduces way below 1.

is there some kind of caveat in th eperformace analysis Tools that makes results look better than they actually are?

what is you opinion on this....ypou input her sis valuable....

[Alex]

First of all, it is neccessary to turn off Odds > Antibet filters > Rating/Intotal trends for optimizing DP.
Otherwise data will be distorted, in particular Entropy will be flipped so as lower values become larger and vice versa.
For optimizing DP, we use several parameters as shown on a screenshot for EPL.
Sure, sometimes reducing DP may have positive effect on performance figures, but it is not stable as only few statistics data is actually used.
For DP=200 days, all games older than 200 days are ignored. That's why I do not recommend reduce DP below 500 days in any case.

[papadas]

Hello Alex,

Thank you for the response.

I see also that when I keep the trends on, the success rate is getting a lot better .

So i the performance analysis results with the trend anti-bet filter “activated” are wrong ?

Actually when antibet trend filter is activated the success rate is getting very good !!!

Why disable the trend filtering ? Can you elaborate ?

[Alex]

Yes, antibet filter increases performance figures, and that was the reason of introducing it in OW.
However antibet filter must be turned off only when optimizing DP.

[papadas]

Alex
du you have any backrest information regarding the effect of the Antibes filters.

I mean in the performance analysis it shows improvement of the success rate from 110–> 130 most of the time.

Is this also happening in reality or this is just a fake indication?

[Alex]

Antibet filter is very useful option. It really improves performance figures and betting performance.

[papadas]

Why then it can be incorporated in the performance analysis in order to see how much better the results get.....?
Just to clarify that.....

if I put the check "tick" in the Account trends in the performance analysis I get false/fake results?
Yes or No?

[papadas]

Hello Alex,

Could you please explain if data in the performance analysis with the trend filter activated is wrong or not ?

And why ?

This is crucial for me because if the data is wrong the edge i think I found is not existent but if the results shown there are truthful tha means that the edge is there

[Alex]

if I put the check "tick" in the Account trends in the performance analysis I get false/fake results?
Yes or No?

No, results are correct. They clearly indicate that accounting trends improves performance.