BuzzerBeater Forums

There are a couple of ok draftees in the ABBL draft this season, low potential but good starting skills, but also some of the best were the 19 year olds. We shall see who I get.

There are a couple of ok draftees in the ABBL draft this season, low potential but good starting skills, but also some of the best were the 19 year olds. We shall see who I get.

Or you can just give me a cool Mil for my ATG, Master Yoda :D

only 1 million?

ill raise that to

one

hundred

BILLION

dollars

...and SOLD!

THe BB community witnessed your pledge. See to it that you uphold your end when the time comes, Coach ;-)

hahah

see if you can list him at that, that would be funny to see.

hahah

see if you can list him at that, that would be funny to see.

I wonder how quickly I'd be reported ;-)

HOLD THE PHONE

*********BREAKING NEWS BREAKING NEWS***********

No seriously, this is ground breaking stuff. And the draft lists are removed so I can reveal this with safety.

Ive checked it all thrice now, (thrice?... hmmm)..... and im not wrong.

So ive managed to collect box score data from several sources, some of which are from the same league. ie: two users from same league give me data

guess... effin.... what?!?!?!

Player X - you know that 7'4" 5 star A+ grade 5-potential player that you scouted and Interviewed? So, two bits of news. Yes im afraid someone else scouted them. but the other piece of news?

His box score's are different.

Yep - you heard me. The box score that is given to Player X is different across users.

WHAT
THE
FUUUUUUUUUUUUUUUUUUUDGE

ok ok calm down, its alright. Deep breaths. in out in out.

So - for the moment, please continue to collect user data. From what I can tell, the box scores are SIMILAR, but definately not THE SAME. So this study is still valid I believe.

I think BB have implimented this deliberately, to help prevent collusion between players. (ie: Box score allows players to be identified easily). So if the box scores where the same, then people would be able to spend less points in order to get and collude data with other users. Kinda clever really.
Does it suggest that its more likely the box scores are randomly generated? Not sure.

It means ill have 2 box scores to use for some of the players, which is fine, it will give me extra data. which is nice.

but how's about that!!!!!!!!!

PS: First person who says 'Oh I already knew that' - I will be hunting you down and sticking a fork up ur pee pee.

i knew that..... youd find something worthwhile ;)

Refresh......



Refresh........




Refresh.

Last edited by whitewind at 3/31/2014 9:44:51 AM

Very interesting, Coach...very interesting. Kind of suggests that this data analysis of box scores might be less effective as first thought. I mean, if we are seeing the same core stats (ie. 5/5 C A+ 7'4" etc) but box scores are different...how relevant are two different box scores for the same player? How can they really help us better scout a player? Exactly how 'similar' are the two box scores?

Initial looks were quite similar, as in 37 mins vs 36 mins, 17 points vs 16 points

but i only checked a few. there might be some marked difference with some of them. Just had a box score where the guy scored 54 points and 12 assists - in the other guys draft, 32 points and 9 assists.

So it looks like it might even be 'scalar', in that where there are differences, they are 'proportional'
which - suggests MAYBE there is a relationship.... to something LOL

even now, im just looking at the data. ive only got about 8 box scores for each 'star' level, but the average points and rebounds are as follows:

Star - Ave Pts - Ave Rbs
1 - N/A
2 - 5.6 - 2.6
3 - 8.4 - 2.6
4 - 9.4 - 2.2
5 - 16.75 - 3.5

So it seeeems like it is trending upwards, but outliers are probably playing a part here.

So yer, ive got another 3 lots of box scores to enter. With the doubling of box scores, ill now have 7 sets of box score data, which is great.

Last edited by Coach Regan at 3/31/2014 10:17:21 AM

Top job, Coach. +1 Hopefully you can add a crap load more next season to widen the data and narrow the trends :)