Friday, May 11, 2012

NPPL HB Stats Posted

Thanks to Chris from Virtue for passing the word (and the link) along. According to Chris the project continues to be a work in progress with Chicago stats promised nightly. Even so check out the HB data to see what they've managed so far and what's coming up next. Sometime soon VFTD will do a comparo of the PSP & NPPL systems. (But don't hold your breath 'cus you know how I feel about work.)

6 comments:

Anonymous said...

those really aren't that impressive compared to the psp...

Ken said...

my biggest problem with the statistics are a few are based completely on inferences. An is the assist.

An example

You get a point for every opponent that stops shooting their gun while you were shooting your gun.

You could be shooting at back center and a player who is not being shot at could step out of bounds while shooting causing him to stop shooting. You would get an assist for nothing.

Another example

Stamina is based on when is the last time you shot your gun in a game.

Theoretically you could fire it right in the beginning and get pinned and stop shooting the rest of the game. You could live the whole game, but the numbers will show you died early.

Admittedly these instances can be rare. However, unless Virtue has people on the ground watching the games and marking things down these statistics are really no more than a bunch of numbers where inferences are drawn inferences as to what happened.

Baca Loco said...

If you guys would like to get together between you it would save me the effort--and you're doing a great job, btw.

chris virtue said...

I don't disagree the power is in the inferences. Strangely Jason Edwards, et al. still rise to the top with only a handful of inferences ;)

But it will get better and I'm not attempting to claim what's online is ideal.

These are good points you make but there are important counter points. Rather than argue about it though I'd rather step back and see how the numbers pan out after a few events.

Ill just say the interesting about inferences in this case is there are (will be) thousands of them made and applied consistently across the board. This is where you'll see some strengths shine through.

I'm signing off until Chicago but I'd love the chance to have both suggestions and fair criticisms made and see if we can figure out a way to improve on what we've got developed. But so far those two legitimate thought experiences won't affect much except dilly round at the margins unless its common for a player to step out of bounds 50% of the time or sit in the snake and never fire a shot most of the time either. In either case there is a way around if you put your mind to it :)

But a critical take will only serve to make it more polished in the end if we can figure out how to address the serious issues - which I'm confident we will in several areas.

Anonymous said...

Sorry Chris, it comes off as a gimmick. Just save us the inevitable question and compare PBA stats/Virtue stats at a single event. Or at least just combine them...

Nick Brockdorff said...

Why at all compare them? - shouldn't the ultimate goal be to combine them?

The PBA system has just as many potential flaws, as some of it is based on an individual believing he saw something and making a note of it.

The combination of what PBA and Virtue is doing, would be an awesome tool to apply to the same league and both would complement eachother very nicely indeed.