BuzzerBeater Forums

BB Global (English) > Power Ranking Analysis

Power Ranking Analysis

Set priority
Show messages by
This Post:
11
159625.1
Date: 10/7/2010 12:19:34 AM
Overall Posts Rated:
5555
The most common indicator right now of how good a team is, it is by looking at the ranking that BB gives us. However, say you sold all your players and for a while you had alot of money but no good players, eventually your team would rank low. But if you got bunch of studs finally, the ranking of your team would still be the same, which would be poor indicator of how good your team is.

Therefore I thought to myself, this would be a great way to directly determine how good your squad is based on your current players. I like to show you first a hypothetical squad which shows the player's position, the salary of each of those players, and the power value of each which I will explain soon what it means:

Position..............Salary.............Power_Value

C1....................$10,000...............$7,500
PF1...................$15,000...............$11,250
SF1...................$15,000...............$11,250
SG1...................$25,000...............$18,750
PG1...................$7,000.................$5,250

C2....................$6,000.................$1,250
PF2...................$7,000.................$1,458
SF2...................$10,000................$2,083
SG2...................$8,000.................$1,667
PG2...................$5,000.................$1,042

C3....................$2,500.................$104
SG3...................$3,500.................$146..
.......................................Sum=$61,750

Here are explanation of what it all means:

1. Position - The number next to the position of each of these players show their role in the team with each number indicating the following:
* 1 - Starter
* 2 - Backup
* 3 - Reserve
When a team plays a game, they can fill up to 12 players and therefore i use 12 players, this is an example so I randomly
chose Center and Shooting Guard for reserve.
2. Power Value - Probably what you were most wondering. The power value of each player is determined based on the role of the player in the team. I think it was best to use what BB uses when there is a "walkover" or when a team must forfeit and the forfeited team loses 25-0. When this happens, The winning team's 5 Starters get 36 minutes, the backups 10 minutes, and the reserve 2 minutes played in that game, you can learn more on walkover on this thread (144856.43).
Given that a full game has 48 minutes, if a starter plays 36 minutes then he plays 75% of the total minutes, backup plays 21% (10 out of 48 min), and reserve play 4%. Since Salary is best indicator of how good a player is based on the information available to us, the formula for the power value is the following
Power Value = Salary * (% Minutes played out of 48 given player's role in the team)
C1 Power Value would be his salary $10K * 75% = $7.5K C2 would be $6K * 21% = $1.25K, and so on. The sum of the team's power value of each of the 12 players came down to $61,750. So the power value of this team is $61,750. Using this a team ranking would be determined based on who has more Total Power Value

I'm not stating that this is the perfect indicator of how good a team is, nothing can be a perfect indicator. For now this is more to show you an idea of a way that works, and I can tell you that one of the more important things to consider is that if one team has 3 players of salary $200K it is worse than another team that has 1 player with $500K and 2 with $50K, both summing to equal salary. There is variation with the worse team so using the standard deviation it can be used as a small reduction factor into the power value. But I decided to keep it simple for now, just to show you an idea.

Last edited by Coach_Gil at 10/7/2010 12:38:15 AM

This Post:
00
159625.2 in reply to 159625.1
Date: 10/7/2010 12:39:40 AM
Overall Posts Rated:
5555
I would love to hear your comments, opinions, and suggestions on this. So feel more than free to express what you think and anything you would like to say

This Post:
00
159625.3 in reply to 159625.2
Date: 10/7/2010 12:51:10 AM
Overall Posts Rated:
5555
Here is the power value of my team

Position..............Salary.............Power_Value

C1....................$28,618...............$7,500
PF1...................$11,123...............$11,250
SF1...................$17,058...............$11,250
SG1...................$26,573...............$18,750
PG1...................$6,253.................$5,250

C2....................$3,258.................$1,250
PF2...................$2,896.................$1,458
SF2...................$5.926................$2,083
SG2...................$5,317.................$1,667
PG2...................$5,192.................$1,042

RS.....................$4,259.................$177
RS.....................$3,530.................$147..
.......................................Sum=$72,249

Looks like my team is better than the example I used, if you want to show what's the power ranking of your team feel free to show it, to make it easier for you just copy/paste from this thread and just replace the salary of your players and their power value with mine. And by the way RS stands for Reserve this isn't really important but in case you were wondering

This Post:
00
159625.4 in reply to 159625.3
Date: 10/7/2010 2:48:23 AM
Overall Posts Rated:
457457
In this you must consider some factors

- Not all the teams have 12+ players (mine has 11 players).
- I have not a fixed line-up, as a lot of teams here.

However, my basic line-up would give me something like this:
Position Salary  Power_Value
PG 30565 22924
SG 33531 25148
SF 11788 8841
PF 53103 39827
C 26561 19921

PG 14712 3089
SG 25746 5047
SF 6838 1436
PF 18447 3874
C 8126 1706

RES 6151 246
_________
TOTAL 132059

This Post:
00
159625.5 in reply to 159625.1
Date: 10/7/2010 4:52:53 AM
Overall Posts Rated:
404404

I'm not stating that this is the perfect indicator of how good a team is, nothing can be a perfect indicator. For now this is more to show you an idea of a way that works, and I can tell you that one of the more important things to consider is that if one team has 3 players of salary $200K it is worse than another team that has 1 player with $500K and 2 with $50K, both summing to equal salary. There is variation with the worse team so using the standard deviation it can be used as a small reduction factor into the power value. But I decided to keep it simple for now, just to show you an idea.

I still don't understand what your index should indicate
And I strongly disagree on the bold phrase,you should inserted a some kind of coefficient based on the role because at the time more inside oriented is your time,higher is your value,while the game surely doesn't work in this way

This Post:
11
159625.6 in reply to 159625.1
Date: 10/7/2010 7:39:37 AM
Overall Posts Rated:
522522
There is a reason I don't like this suggestion.
Consider this:
Do you know what snooker is? Or 8 ball? etc
Imagine someone is the best snooker player (or 8 ball player) in the world. They are winning every tournament etc.
But then they break their arm in an accident.
Doctors say that he will not be able to play for 4 weeks, and then he will be completely back to normal.

This is similar to a team in div 1 constantly winning the championship, and then selling all his players for 4 weeks.
Using your system, he would have a ranking of 0 during those 4 weeks. And so would be the worst team in the country according to rankings.

I don't think rankings should be able to fluctuate so much so quickly.
Plus (not sure about this one) but I think in the BBB the lower ranked team gets Home court? Or something like that. If ranking is based on this then teams might sell a player to get a lower ranking than the other team, and then get home court, and then buy a player again.

This Post:
00
159625.7 in reply to 159625.6
Date: 10/7/2010 12:58:01 PM
Overall Posts Rated:
409409
I would be surprised to know that HCA is determined that way in B3.

I think the main problem of this idea is salary. Salary is not very good at discriminating the quality of equally paid players. For example, a SF with high IS would be charing almost the same as one with a much lower level in that skill but, all of us would agree that the first one is better(like Passing for a SG). There many problems when salary is used to determine wheter a player is or isn't better than other.

Second, it is not so important who has the better players but how are those players being used. The Power Value approach doesn't take into account the performance of a team.

While I agree that current ranking system might be improved we don't know exactly how it works. We know is very valuable to advance in the National Tornament and that being invited to B3 gives you also some points but I'm not very sure how it works besides that.

Last edited by Zero, the Magi. at 10/7/2010 1:00:12 PM

This Post:
00
159625.8 in reply to 159625.7
Date: 10/7/2010 1:08:54 PM
Overall Posts Rated:
522522
I would be surprised to know that HCA is determined that way in B3.


I'm not sure that it is, I just remember reading something about Home court being decided using rankings, but I could be completely wrong here and just misremembering something so don't hold me to it :P

This Post:
00
159625.9 in reply to 159625.6
Date: 10/7/2010 4:16:13 PM
Overall Posts Rated:
5555
This is similar to a team in div 1 constantly winning the championship, and then selling all his players for 4 weeks.
Using your system, he would have a ranking of 0 during those 4 weeks. And so would be the worst team in the country according to rankings.


Yes I agree with alot of the things that you all said, as I said this was made to gesture an idea and receive comments.

I should have mentioned before on how I thought this would be beneficial and used. I don't think by all means it should replace the current ranking system. The current ranking is an indicator of the past accomplishments and we get good idea of their potent based on it, that ranking is the best and safest way to actually know how good that team is, and it is that ranking that say if you are div 1 team that sold all your players, will still indicate that you are very good.

The Power Value system is strictly made for one purpose and one purpose only. That is to measure how good is the team you are about to face, not a measure whatsoever of how well the team has done to that point and their situated position in accomplishment. If a team did sell all their players and you are about to face that team that has nothing, then the Power Value of 0 will accurately indicate how strong that team is when you will face them. In Conclusion, The Power Value is NOT made to measure the ranking of a team based on success but strictly an indicator of how good the team is you are facing based on their current roster.

While I agree that salary is not the best indicator of how good a player is, the information available to us on the other teams do not provide any better measure of how good a player is, with salary being the closest tool acknowledged. The salary is widely agreed within BB as best indicator of how good a player is from the limited information we have. If the skills of each player were available to us then determining how good a player is would be a completely different story.

As far as having less than 12 players in a roster, that's a very good point acknowledge I should have thought about. As I mention earlier I was only trying to show basic example as opposed to finalized proposal. That's one of the issues need to be addressed. I feel like if there are less than 12 players, adjusting the team's power value will definitely be the way to go about it. For now I'm just suggesting and how the adjustment will be made would require some work which for now i'll wait on figuring it. The final point i like to make is that i think with certainty adjusted power value can resolve this issue.

Other things that are involved such as enthusiasm level, home/away court, injuries, etc. Yes there are even more things to consider than this, as I mentioned earlier about the 3 players example and different salaries among them that's one thing to consider for example. With mentioning that in mind I tried to also send the message that yeah I did have few other things I knew should be taken into consideration. All of them should be considered, but the real point that I am trying to make is that they can be addressed and you can make it work.

Power Value of a team works pretty much exactly like we see in a video game, each team gets a value that indicates how good they are, and that's based on a formula that considers the players in the roster and their value. Power Value works exactly in that same way where values being considered are the salary. I think anyone who plays video game would like to know how good a team is, and that's exactly the case with Power Value it can provide us that exact info we would love to know about the other team. And like if for example in NBA Live you have 5 players on your roster and not 12, NBA Live would use some algorithm to consider the subsidized number of players, such can address the less than 12 players in BB


Last edited by Coach_Gil at 10/7/2010 4:29:54 PM

This Post:
00
159625.10 in reply to 159625.9
Date: 10/7/2010 4:32:08 PM
Overall Posts Rated:
522522
Gameshape is also an important factor to consider. As well as whether you use the salary shown or the "true" salary. The "true" salary is different to the salary shown because of training or drop in skills. So which one is used for the power value?

This Post:
00
159625.11 in reply to 159625.9
Date: 10/7/2010 4:33:03 PM
Overall Posts Rated:
959959
The Power Value system is strictly made for one purpose and one purpose only. That is to measure how good is the team you are about to face


i believe when you analyse your opponent, you could your time better then calculating with this formula ;) Even when choosing a probadly best formation isn't a bad move with it. And what brings you a indication of the ebst ten, when he plays with double backups etc.

And to emasure on salary only is very biased, salary differences on Centers are less important then salary differences on point guards etc. And also the skill contribution plays a role there ;)