Floating Line Value

[hoopstats]

My name is Matias Barmat, from Buenos Aires Argentina. I developed a new linear weight measure called Floating Line Value.

http://english.worldhoopstats.com/news/ ... -line.html

Enjoy!

[hoopstats]

I have a new article from WorldHoopstats.com about basketball statistical criteria.

http://www.worldhoopstats.com/news/60-v ... teria.html

Enjoy

Matias Barmat
Buenos Aires, Argentina

[hoopstats]

This is another article from WorldHoopstats.com

http://english.worldhoopstats.com/news/ ... -ball.html

Enjoy!

[DSMok1]

Hola, Matias! I merged the three threads.

Thanks for the links.

A few thoughts:

Floating Line Value: I strongly recommend against a multiplicative approach like this--if a player has 0 assists, steals, blocks, etc--his value will be 0. That is not accurate at all. A multiplicative approach is neither stable or accurate.

Valuation of Statistical Criteria: How were they evaluated? A mathematical approach? I note that many of the better advanced stats are not included in the evaluation, most notably Win Shares/48 (the best commonly available stat in my opinion), the flavors of RAPM, and (not common) my own ASPM.

One Percenters: This article seems quite similar to the charting system outlined by Nathan Walker: http://www.thebasketballdistribution.co ... score.html

[v-zero]

I agree with DSMok1 on both counts. I would also like to take this opportunity to say that many of the ratings systems on that list are abysmal, and that Wins Produced is very unscientific both in its original formulation, and in its continued attempts to subvert proper scientific reasoning with strawmen, patronisation and hand waving .

A list like that would be more useful if it gave a true sense of what each metric does, and indeed where it falls down.