BuzzerBeater Forums

your draft is relatively bad, the best draft i had seen is the one of my league this season:
(all salarys rounded)

3.200 18 4
5.000 18 10
4.000 18 10
5.800 18 6
3.600 18 10
4.400 18 9
5.500 18 6
3.500 18 7
4.700 19 1
4.300 18 4
4.000 19 5
4.000 18 5
4.000 19 6
4.100 18 6
3.900 18 4
4.400 18 4 (retired?!)

i don't think that you can top this soon ;)

In case someone is doing a study:

19 3 3744 PF
18 5 2088 PF
19 2 2089 PG
19 2 3175 SG
18 4 3200 SF
19 2 1934 C
18 5 2366 C
19 4 3563 SG
18 3 2067 SG
18 2 2888 PF
18 9 6241 PF
19 4 2768 PG
18 3 3171 PF
19 2 2011 C
18 4 2862 C
19 6 4732 C

i think for a study you should use the full list, because inactive "drafters" and different draft spending make a big different in draft behaviour.

So i could say that your pick 3 and 4 not get draftet in the first round, from active managers ;) And those guy with the high potential PF with 6200$ salary was lucky about this inactive opposition, i think^^

I just want to pose with our special draft result ;)

Last edited by CrazyEye at 12/15/2009 10:35:15 AM

Actually, I was the only human involved in this draft and I drafted last.

Additionally, I did not recognize the importance of scouting until about 3 weeks into the season so did not have info
on all players (that PF, for instance)

Allright then, I think I'm gonna do an analysis myself of American Div V drafts for season 10

Format would be based on the following that I got for my own draft:

# 18: 23
# 19: 25

Positions were dispersed as follows:

PG 11
SG 13
SF 3
PF 8
C 13

Potentials went from bench warmer to MVP with the following dispersion:

bench warmer 6
role player 8
6th man 4
starter 11
star 10
allstar 3
perennial allstar 3
superstar 2
MVP 1

Avg. Potential: 4.00

Salaries were dispersed thusly:

1-2 K: 4
2-3 K: 23
3-4 K: 12
4-5 K: 7
5-6 K: 1
>6 K: 1

HIGH SAL: $6241
LOW SAL: $1313
AVG SAL: $3038

and so on with potential and salary breakdowns for each age and position

2 Questions:

Is doing this worthwhile or just a way to kill some time?

If it is worth doing, would 10 drafts be enough or would I have to go 20?

I would only be analyzing Level V American Leagues and try to get a reasonable mix of bot-filled and active human leagues?

EDIT: Almost forgot. I would put all data on a publicly published Google Docs spreadsheet so anyone could check it out


Last edited by drencrom68 at 12/15/2009 1:35:36 PM

For CBBA was the worse draft by far. Usually we where able to get a decent player from there, at least the majority of the teams. This season is just awful. Looks like in Cyprus in general the draft is pretty poor compared to other seasons. I am closed to the NT managers so i have a clear idea from every seasons draft.

when the best team in a country recieve these very good players... no wonder Germany will have the best u21 team in 2 seasons..

three, we didn't scouted the other leagues yet, but normally you have that bunch of talent in 3-5 leagues not in one ;)

Hope this helps. I just like to crunch numbers and I was really curious about the relationships and how well our division drafted.

USA Div III.4

19x 18-yrs
29x 19-yrs

PG x15 (High-6 $4018, Low-48 $1360, Avg $2812)
SG x7 (High-5 $4196, Low-46 $1615, Avg $3175)
SF x9 (High-3 $4224, Low-43 $1780, Avg $2715)
PF x11 (High-1 $4724, Low-42 $1792, Avg $2990)
C x6 (High-2 $4699, Low-47 $1564, $Avg $2580)

Announcer x0
Bench Warmer x11
Role Player x7
6th Man x6
Starter x3
Star x10
AllStar x6
Perennial AllStar x2
Superstar x1
MVP x0
Hall of Famer x2
All Time Great x0

High Salary: $4,724
Low Salary: $1,360
Avg Salary: $2,859
Standard Dev: $889
Z-Scores: 2.10 to -1.69

USA Division III.4 is an Active All-Human Controlled Division. However, not all poured money into the draft as our league likes the transfer list.

Here's how our draft played out:

First Round:
None Retired
1st Pick: PF, $1984, Superstar
Best Pick: C, $4699, Hall of Famer, #2 Pick
Worst Pick: SF, $1780, 6th Man, #5 Pick
Avg Salary: $3620

13 of the 1st-Tier Salaried Players were drafted in the 1st Round (Avg Salary $3974)
1 of the 2nd-Tier Salaried Players was drafted in the 1st Round (Avg Salary $2504)
2 of the 3rd-Tier Salaried Players was drafted in the 1st Round (Avg Salary $1882)

Second Round:
4 Retired (1x Starter, 1x All Star, 2x Bench Warmers, Avg Salary $2853)
Best Pick: 19-yr SG, $3679, AllStar (RETIRED), #25 Pick
Worst Pick: 19-yr SG, $1615, Bench Warmer, #22 Pick
Avg Salary: $2848

3 of the 1st-Tier Salaried Players were drafted in the 2nd Round (Avg Salary $3583), 2 of which are retired
11 of the 2nd-Tier Salaried Players were drafted in the 2nd Round (Avg Salary $2859), 1 of which is retired
2 of the 3rd-Tier Salaried Players were drafted in the 2nd Round (Avg Salary $1682), 1 of which is retired

Third Round:
9 Retired (1x Perennial AllStar, 1x Star, 2x 6th Man, 2x Role Players, 3x Bench Warmers, Avg Salary $2163)
Best Pick: 18-yr PG, $2810, Role Player (RETIRED), #36 Pick
Worst Pick: 19-yr PG, $1360, Bench Warmer (RETIRED), #43 Pick
48th Pick: 19-yr PG, $2291, Star (RETIRED), 33rd highest salary
Avg Salary: $2107

0 of the 1st-Tier Salaried Players was drafted in the 3rd Round
4 of the 2nd-Tier Salaried Players was drafted in the 3rd Round (Avg Salary $2589), 2 of which are retired
12 of the 3rd-Tier Salaried Players was drafted in the 3rd Round (Avg Salary $1946), 7 of which are retired

For some reason, I was under the impression that higher level teams would get draft picks that were significantly better than those in lower leagues. Your post shows that this does not seem to be the case.

So, I have decided that I am going to go ahead and do a study and then put everything on an accessible to the public spreadsheet.

How many drafts should I survey?

What data should I include?

Should I gather data just from one country; at least one from every country or just make it completely random?

And, pardon the ignorance (actually more like senility), but how does one find a standard deviation?

I will be having a lot of free time in the next few months, so I swear I will get this done, but it may take a month or so.

This is my first time to actually record the draft results. I can go back and record but the data won't be accurate. I was also under the impression that drafts would be better the higher up you advance but I figured it would not be that dramatic of an increase in talent. I think the higher up you advance, however, the less significant the draft becomes as your rely more and more on the transfer list.

My theory is that lower level teams rely on the draft early as well as purchasing cheap trainable players off the transfer list. As the players progress and the teams progress along as well, then they sell them high in hopes of acquiring better talent that allows them to compete at their new levels.

With each draft consisting of 48 players, that would give you 480 subjects to study. Before you begin your study though, you need to clarify your objective of this study. After pulling the data from this draft, I've concluded that I really did not gather anything more than facts and at this point I have no idea why I am studying this data other than the fact that I was interested in the actual results of our draft and wondering who pulled the best draft class in my division.

I do agree that you would need to gather data from quite a few of the major countries and quite a few minor countries. Then I'd compare the two samples of data with each other and see if the major countries are pulling strong draft classes than the minor countries. That would be something I'm curious in.

As far as standard deviation, I only learned about it during my years as a maintenance analyst for the USAF. I've been out for nearly 6 years now so I am not current with any changes in statistical analysis. The only thing standard deviation does is determine the average deviation between data. In this case, I was comparing the salaries in the draft class and trying to determine the relationship between their actual draft position (which I've determined that draft position has no effect on the actual salary as they are apparently pre-determined ... which I'm pretty sure had already been figured out by a lot of others here).

To calculate standard deviation, you'll need no less than 30 samples. Thankfully there are 48 samples in a draft class. To calculate the Standard Deviation in Excel, it's actually quite simple:

=stdev(selection of cells) <--- Only 1 column, cannot be a group of cells
=stdevp(selection of cells) <--- Only 1 column, cannot be a group of cells

The two numbers will have a slight variance but it won't be by much. Honestly, I have a hard time determining when I should use STDEVP and STDEV.

To determine the Z-Score, subtract an individual sample from the average and divide by the standard deviation. To understand the importance of the Z-Score, I recommend reading up on Wikipedia.