Improvement in W', CP, or both?

[snarthelg]

Thanks for posting this algorithm. I have also used Xert and find it quite useful.

I'm not a programmer, so I don't understand the code here on Github. I suspect that the answer to my question lies in the code/mathematical formulas.

Q: How is it possible to differentiate if a change in fitness is due to changes in W' or CP?

[Berg0162]

When you exhaust your W-prime completely by cycling above your CP it is clear that the value(s) need to be adjusted... otherwise you would not have been capable to exhaust the W-prime! So the algorithm upgrades your CP and slightly adjusts your W-prime. The next effort needs to be long enough and above both new values to be adjusted again... The closer you get to your real "level" the smaller the adjustments become and the better the approximation is! Xert also adjusts your W-prime but does not make it visible. It needs the W-prime to determine how long your effort above CP should last before adjustment! Best results when you set CP and W-prime close to your real "level". The algorithm I coded makes some basic assumptions about the "average" values of CP and W-prime to start with when the user is ignorant about his/her level. During efforts one gets a fair approximation of the realistic values. Adjust the starting values and the next effort really gets seriously close to what one might expect to be your levels! Hope this helped you! Op do 12 mei 2022 14:31 schreef snarthelg ***@***.***>:

…

Thanks for posting this algorithm. I have also used Xert and find it quite useful. I'm not a programmer, so I don't understand the code here on Github. I suspect that the answer to my question lies in the code/mathematical formulas. Q: How is it possible to differentiate if a change in fitness is due to changes in W' or CP? — Reply to this email directly, view it on GitHub <#1>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ANS5LSXUEIFA64CQOHXMVSDVJT2ZHANCNFSM5VYDECLA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>