BuzzerBeater Forums

Hi everyone

Thought id make a bit of a thread about the statistical analysis that im doing, that some of you might be aware of.

The Premise

Basically that the Box Score that is given to you when you scout a draftee actually means something! Maybe not definitively, but maybe on average, it can indicate something. There has been many a post about this in the past, and a little light bulb went off in my head one day whereby I thought "Hang on, BB doesn't really throw really big red herrings into this game, maybe Box Score actually DOES mean something with a draft pick". AND that means BB are sitting there laughing at all of us for ignoring it. Sods! Right! can't have that!

So the best way to find this out and whether there is a relation?
LOTS OF STATS WOOOOOOOOT.

I have been able to take a number of stats from various divisions, in various manners and, without giving too much away, have been able to gather together box scores from various leagues. I will tally this data together using my fancy new method, and then once the draftee's are drafted, I rely on the users from these leagues to - not fire - but sell their draft picks, such that i can see their stats, and grab them.

One i have Player Box Score and Player Stats, i can compare the two.

48 picks over 4 draft's that i have, is a lot. I dont have all the box scores, maybe about 100. So even if i get 50% of those player stats, im hoping that a player pool of 50 players, will be enough to do some analysis.

If not, then I can keep the stats, and do it all again next season! Woot!

So thats a premise. Its a bit of work on my end, and ive done it before (but i think i cocked it up).

So ill see what happens. Ive already got some interesting stats just out of the box scores, which I will share in future posts.

I will also provide the results of my analysis in the coming weeks as I trawl through the data.

Thanks to those anonymous peeps who have helped thus far. You know who you are, and your identities will be revealed in the near future, so that I can lay blame on you for not doing enough scouting. NO im joking.

So some stats for you to tantalize your taste buds.

In the 2 league's i have collated thus far, the minimum and maximum number of minutes a player plays, is the same, 20 and 40 respectively.

In league 1, the Ave PTS = 8.5, whilst in league 2, its 10.5

Everybody, in both of the leagues, turns the ball over AT LEAST once.

The maximum number of total rebounds of either league, is 7 - which, is quite low IMO. ESPECIALLY considering that the FG% of the players, is around 35%

Nobody has dominated a game yet offensively, maximum points of 24 and 27 respectively

These stats actually made me think of a few questions - which i might pose to BB.

- Do all the draftee's play in the same game? There are 48 players, is there perhaps, 3 games, with teams of 8? that could accommodate minutes of 20-40.

- Is the game played like an arcade game? eg: Like the all-star game?

- Perhaps the box scores are derived from a game against a team of Lucky Fans? A bit like a walk-over?

anyways - enough for now! Any questions let me know.

If anyone is sitting on 200+ scouting points that you will never use for just a quarter of that you can scout an entire league and help this worthwhile project. Maybe check with coach to make sure your league hasn't been done yet before you blow the points although two people scouting the same league and getting different results would be interesting in itself.

well actually i just need the box score, so 200 points is a bit excessive! haha

either 48 points to 'scout' all the players. or you could do the Group Combine, get approximately 20 of them (but only spend 10 points) and then scout the remainder, so estimated 38 points.

but yer, essentially the info im after is -

Box Score, age, height, and star rating.

The biggest challange is, that, once drafted, identifying who is who - which will be done through a combination of age, height and then star/grade comparison. So the more ages and heights and star ratings i get, the easier it will make for determining 'who is who'.
That is - in the draft list, if there are three 18 year olds, at 6'1, the only way to tell the difference once drafted will be their salary, which will be related to their star rating. If two of them have the same star rating, then im unable to determine which guy is which, and subsequently, their box score data becomes invalid.

so yer, there is some complexity it what im doing, and its very interesting. If there are people out there who have 40 points to spare, id love to hear from you. as yoda said, msg me first because your league may already have been 'drafted'.

I'm afraid I can't contribute this season, but if you continue the analysis next season, I'll be more than happy to help, I only have 34 points at the moment, but by next season, I should have plenty ! :)

Even without scouting the info can help the project

How?

The Box scores can provide info on the players. There is a simple way for all that info to be submitted. Then if and when any players are listed for sale the box scores can be used to see if it has any correlation to the skills. but if no one submits info from the league it is excluded from the data pool.

dont you need to scout to see the box scores though? ;)

Awesome work, Coach. From the 27 box scores I will be giving you 2 of the players have scored more than the 27pts you have as a highest game so far. I have another 2 players who have scored 9 and 12 assists respectively...pretty decent. I can also confirm that the lowest and highest minutes played from my 27 box scores is 20 and 40 respectively...clearly that is a fact. I have also no-one who has scored more than 6 rebounds...

...some interesting in formation and I think you might be spoton about it played as an allstar/or arcade type match...would seem to reflect data so far.

I will type up box scores/star rating/height/position for you and text it to you today sometime