Is Tendulkar's career now "complete"?

[Bun]

Reckon the data could do with some serious weighting. The raw innings/100 ratio is really prone to bias.

Agree, but any filtering of the same imho just increases the bias one way or the other.

 

[Top_Cat]

Yeah, as I've said many times, there's no defensible statistical basis for excluding weaker opponents. It's actually quite a complex issue truly understanding a dataset like a sports one. Just about all sports data, if you're to use the raw measures, should only be used as vague guides only. Otherwise, you need to sample, transform and/or weight appropriately.

 

[honestbharani]

Yeah, as I've said many times, there's no defensible statistical basis for excluding weaker opponents. It's actually quite a complex issue truly understanding a dataset like a sports one. Just about all sports data, if you're to use the raw measures, should only be used as vague guides only. Otherwise, you need to sample, transform and/or weight appropriately.

as I have said before, when it comes to "how to use stats in cricket", listen to the pro!!!

 

[Howe_zat]

Top_Cat speaks the truth. And speaks it better than me by being all clever and whatnot.

 

[Prince EWS]

Yeah, as I've said many times, there's no defensible statistical basis for excluding weaker opponents. It's actually quite a complex issue truly understanding a dataset like a sports one. Just about all sports data, if you're to use the raw measures, should only be used as vague guides only. Otherwise, you need to sample, transform and/or weight appropriately.

AWTA.
z