BuzzerBeater Forums

I think the Main thing holding Mina back would be his FG%.. IF you look at the list, many of those above him have shot better and piled stats other than scoring.. I sure Mina will climb as the season goes on

Don't take early season rankings as serious. You'll only give yourself a migraine. Early season rankings give players a baseline for measuring progress over a season.

Looking over Mina's stats, he was rated a 10.5 by BuzzerBeater's system but in terms of actual basketball statistics using John Hollinger's Game Score has the basis for measuring a player's success on a game-by-game basis, he scored a 15.2. When you compare Mina's HGS to the rest of the league without taking MINUTES into account, he scored the 14th highest score.

I tried to be as fair as possible when measuring a player's HGS to a fellow player and I decided to use minutes as the factor that grouped players.

Let me explain my ranking logic next.


Each game consists of 48 minutes that is divided among the 12 players that are allowed to suit up for the game.

I first determined that there are exactly 80 starters for each group of games we play (16 teams x 5 starting spots = 80 starters maximum). I tried to grade players based on starters but determined that the best method for grading players was to first determine the distribution of minutes.

Here were my findings for Game 1.1

146 players saw at least 1 minute of action for the first game, which includes two players that saw maybe less than 1 minute which they were labeled as 0 minutes so I grouped those two players with players that played 1 miute.

Maximum Minutes Played was 48
Minimum Minutes Played was 0
Range of Minutes Played was 48
12 Classes were created
Interval per class was 4

Here were the Classes

45 - 48 minutes :: 19 players
41 - 44 minutes :: 20 players
37 - 40 minutes :: 16 players
33 - 36 minutes :: 18 players
29 - 32 minutes :: 4 players
25 - 28 minutes :: 3 players
21 - 24 minutes :: 3 players
17 - 20 minutes :: 4 players
13 - 16 minutes :: 14 players
9 - 12 minutes :: 17 players
5 - 8 minutes :: 15 players
1 - 4 minutes :: 13 players

As you can see, it is heavy on both ends but light in the middle. I tried to group into three but there was a huge difference between a player that played 33 minutes to a player that played 48 minutes. So instead of 3 major groups, I broke them out into 6 major groups.

41 - 48 minutes :: Group A
33 - 40 minutes :: Group B
25 - 32 minutes :: Group C
17 - 24 minutes :: Group D
9 - 16 minutes :: Group E
1 - 8 minutes :: Group F

Group A & B: at least 68.75% of minutes, 73 of 146 players (50%)
Group C & D: at least 35.1% of minutes, 14 of 146 players (9.6%)
Group E & F: no more than 33.3% of minutes, 59 f 146 players (40.4%)

Group A & B were included for the Player Rankings and were ranked by HGS Largest to Smallest, Group A then Group B. For example, a Group A player with an HGS of 15.2 was ranked higher than a Group B player that scored an HGS of 15.2. At this point, I have considered normalizing each group based on that group's highest possible minutes but I have not decided if that will be the best method yet. With that considered, I did do the Normalizaton and Mina, since he played 40 minutes, his HGS did not rise and his ranking fell from 28 to 32.

Finally, keep in mind this is the first game.

I was just being a smart ass. ;-)

cool beans. still, I still had to explain how I came about my rankings.

The normalization of the players is tough but i think it is important. I guy that play 20 min ans score 20 points plus 8 rebounds 3 steals should have a highest impact than other that play 45 min score 25points 9 rembouds and 4 steals. On top of that the field goal percentage is a thing to be considered.

So good luck with it. It looks awesome for me.

The HGS does consider the FG% by counting missed shots as a negative and made shots as a positive.

I'm thinking highly of going back to the normalization but it will work differently than the other determining a player's HGS per 48.

The normalization process now only involves players within their own group.

For example, we will use Group B which consists of players who play 33 to 40 minutes. I determine a player's HGS per minute and then multiply that player by their group. So for example, a Group B player scores a 15.2 HGS in 33 minutes so I determine that he rates a 0.46 per minute and by that number alone, I then multiply that by 40 and determines he comes out to 18.4, +3.2 higher.

I could even go as far as using his subgroup which would basically add anywhere from 1 to 3 more minutes on top of their actual minutes. I haven't made that decision yet but I have decided to go ahead and revert back to normalizing by primary group.

I disagree that playing more minutes should hurt you or playing less minutes gives you any kind of bonus.

If we were talking about a 6th man award, then sure. Quick impact points/stats might be worth considering. However, we're not talking about that type of player. Your system is trying to evaluate who it is that a team would want on the floor to make an impact on a game. A player being capable of playing large chunks of minutes is a HUGE assist to the team as they can put their resources into other positions.

I think its awesome you have the time and energy to do something like this. I really did luck out with this forum. Hoepfully i stay here for a few seasons!

Man this is just the beginning. You haven't seen anything yet. This forum is hands down the best in BB. Pretty much 100% participation and insight from everyone on everything. The only forum I know of that keeps people who demoted and promoted coming back.

I disagree that playing more minutes should hurt you or playing less minutes gives you any kind of bonus.

If we were talking about a 6th man award, then sure. Quick impact points/stats might be worth considering. However, we're not talking about that type of player. Your system is trying to evaluate who it is that a team would want on the floor to make an impact on a game. A player being capable of playing large chunks of minutes is a HUGE assist to the team as they can put their resources into other positions.

It all about performance but somehow you have to normalize it in order to be realistic.

The only reason I would be against normalizing the data is that I ran a correlation test of Minutes to HGS and found that although the relationship between Minutes and HGS scores is positive, there really is no correlation in an HGS rising just because a player plays more minutes. In other words, minutes has no effect when measuring the population's HGS. However, we could easily take each player's data and run a correlation to their minutes played and their HGS scores and determine your player's target minutes where he should play more effectively.