'02 Girls FBR Rankings - Week 5

Thanks for your work.

I have Fever Reds record as 4-1 in their last 5 games of spring. ???

RoidRage wrote:I have Fever Reds record as 4-1 in their last 5 games of spring. ???

The 1-2-1 comes from 3 games against '02's in the Jan. Winter Classic tourney + 1 game against an '02 team in U10 PTSAL.

Winter Classic:
0-3 v. FCD Grubb
1-1 v. DT Glotz
0-4 v. Rowdies

PTSAL
2-1 v. DT South '02

Unfortunately it is very difficult for me to include games against '01's from leagues where there are only 1 or 2 '02 teams because I don't get enough common games to get an accurate read on SoS. It's different for an '03 team playing up in a league that is majority '02 teams because I get several games against common '02 opponents in the DB to get a better SoS read.

So for Fever Red (and Fever White for that matter) it's going to be tough to get real accurate read on them until the hopefully play in some '02 tournaments later this Spring.

Hope that explains it.

bwgophers wrote:

RoidRage wrote:I have Fever Reds record as 4-1 in their last 5 games of spring. ???

The 1-2-1 comes from 3 games against '02's in the Jan. Winter Classic tourney + 1 game against an '02 team in U10 PTSAL.

Winter Classic:
0-3 v. FCD Grubb
1-1 v. DT Glotz
0-4 v. Rowdies

PTSAL
2-1 v. DT South '02

Unfortunately it is very difficult for me to include games against '01's from leagues where there are only 1 or 2 '02 teams because I don't get enough common games to get an accurate read on SoS. It's different for an '03 team playing up in a league that is majority '02 teams because I get several games against common '02 opponents in the DB to get a better SoS read.

So for Fever Red (and Fever White for that matter) it's going to be tough to get real accurate read on them until the hopefully play in some '02 tournaments later this Spring.

Hope that explains it.

Crystal clear....Ty.

Gophers i totally appreciate your effort and do not wish to derail it one bit. However, i have not even heard of a couple teams in the top ten (TFC Elite and Dynasty?). maybe as the data matures this will work itself out.

TFC Elite ranking looks good bc it is computer based and strength of schedule is a factor and so the win v dynamo helped. Plus the team has only played 4 games. I am sure it will all work itself out along the way, and I'd also like to think the team will keep winning. It is my understanding that this team will play in some spring tournaments. I'm sure there are those who will suggest that it shouldn't be ranked at all, since right now it is a "blend" from all TFC 02 teams, and that's fine too. We can merely imagine the team ranked in our best guess spot if they get kicked out of the rankings for currently being an "all star" team. I'm just enjoying watching the games. The 02s seem to have some good top teams who appear to be coached to play "real" soccer as opposed to kick and chase. Hope that continues and that those who aren't doing that figure it out sooner rather than later!

PS--never heard of Dynasty either. Are they playing SDL or elsewhere?

bobmac15 wrote:Gophers i totally appreciate your effort and do not wish to derail it one bit. However, i have not even heard of a couple teams in the top ten (TFC Elite and Dynasty?). maybe as the data matures this will work itself out.

TFC Elite is playing in EAL on Sundays with Solar Stav, Solar Dynamo and 2 weaker ASG teams. Lost to Solar Stav 5-1, beat Dynamo 2-0 and have beaten the ASG teams by similar goal margins (~6 goals/game) as both Stav and Dynamo have. I believe that I read somewhere on the forum that this team is sort of a TFC "All Star" team that includes some of the best players from the TFC Wells SDL team and a couple of the other '02 TFC teams. Given their results, I think that somewhere between Stav and Dynamo in the 5-7 range is an appropriate rank for this team.

Dynasty is a team playing in Arlington PAL. They are dominating that league to the tune of a 4-0 record with 16 GF and 2 GA, every game they have played they have won by a minimum of 3 goals.

As a point of reference, Arlington PAL includes Polaris 02. Polaris 02 played in Texas Winter Cup at the end of January. In that tourney, Polaris split 2 games with D'Feeters 02 (currently ranked #15 in FBR), and lost 1-0 to TFC Wells (currently ranked #17 in FBR).

Dynasty beat Polaris 3-0 in Arlington PAL on Feb 20. So, Dynasty beat a team by 3 goals that has been very competitive with well known SDL teams currently ranked in or near the top 15 (I believe Roid Rage even has Feet in his top 10).

I would dare suggest that Dynasty may be the best little unknown '02 independent out there. Hopefully they will show up in some Spring tourneys against the better known SDL teams and we will get an even better read on them. Something tells me that if you are a weaker SDL D1 team or SDL D2 team, I wouldn't take Dynasty lightly if they show up on you tourney schedule this Spring.

bwgophers wrote:

bobmac15 wrote:Gophers i totally appreciate your effort and do not wish to derail it one bit. However, i have not even heard of a couple teams in the top ten (TFC Elite and Dynasty?). maybe as the data matures this will work itself out.

TFC Elite is playing in EAL on Sundays with Solar Stav, Solar Dynamo and 2 weaker ASG teams. Lost to Solar Stav 5-1, beat Dynamo 2-0 and have beaten the ASG teams by similar goal margins (~6 goals/game) as both Stav and Dynamo have. I believe that I read somewhere on the forum that this team is sort of a TFC "All Star" team that includes some of the best players from the TFC Wells SDL team and a couple of the other '02 TFC teams. Given their results, I think that somewhere between Stav and Dynamo in the 5-7 range is an appropriate rank for this team.

Dynasty is a team playing in Arlington PAL. They are dominating that league to the tune of a 4-0 record with 16 GF and 2 GA, every game they have played they have won by a minimum of 3 goals.

As a point of reference, Arlington PAL includes Polaris 02. Polaris 02 played in Texas Winter Cup at the end of January. In that tourney, Polaris split 2 games with D'Feeters 02 (currently ranked #15 in FBR), and lost 1-0 to TFC Wells (currently ranked #17 in FBR).

Dynasty beat Polaris 3-0 in Arlington PAL on Feb 20. So, Dynasty beat a team by 3 goals that has been very competitive with well known SDL teams currently ranked in or near the top 15 (I believe Roid Rage even has Feet in his top 10).

I would dare suggest that Dynasty may be the best little unknown '02 independent out there. Hopefully they will show up in some Spring tourneys against the better known SDL teams and we will get an even better read on them. Something tells me that if you are a weaker SDL D1 team or SDL D2 team, I wouldn't take Dynasty lightly if they show up on you tourney schedule this Spring.

p.s. Go look up Dynasty's record on GotSoccer... Played in a December tourney. Lost twice to Solar Murin (0-3 & 1-3) in that tourney. I asked around and was told that Murin '02 was considered a top 10 team before they broke up, don't know if that's true or not. Dynasty also had some pretty good results in Arlington PAL last fall, though not quite as good as they've started out this Spring. Early results from the Spring suggest that they may have improved from the fall, and if so, I think they possibly belong in the discussion around the bottom of the top 10. If not, it still looks to me like they belong in the top 15 discussion. Again, hopefully they'll play in some tourneys this Spring with crossover to SDL teams so we get a better read on exactly where they should be.

Okay for scores and rankings you cannot compare different leagues as most of the top teams play in the SDL league. The PAL and CAL leagues are lower level leagues, but still could have a good team dominating. We will have to wait until they play other teams in tournaments. Other top leagues are the EAL League..I think that is second to the SDL. ALso primetime has a league which in my opinion is third with level of dificulty. It is unfortunate that the Primetime did not have a TGPL league for the U9 age group this Spring would have been nice.

Anyway there are top 01 teams playing maybe up...
#1. Rowdies 02( Dallas Texans Top Team)
2. FCD Grubb 02- not sure where they are playing?
3. Fever 02- playing up in primetime 01's

IN SDL there are three Divisions:
SDL Division 1- has 7 teams rank them with the top three.
then SDL Division 2- there are 12 teams rank them next
then the SDL Division 3- teams there are 13 teams.

The 02 group needs someone to do standings first with all the leagues before rankings can come out with all the different Leagues.

Polaris 02 played in Texas Winter Cup at the end of January. In that tourney, Polaris split 2 games with D'Feeters 02 (currently ranked #15 in FBR), and lost 1-0 to TFC Wells (currently ranked #17 in FBR).

JUST an FYI I Could be wrong about what tournament but one of the Tournaments that Polaris played over the Winter break they definately used close to 4-5 Rowdies 02 players and did very well made it to the Finals...so it was not the true Polaris Team. I know bc i watched some of the game as I was going to our game.

kickback49 wrote:

Okay for scores and rankings you cannot compare different leagues as most of the top teams play in the SDL league. The PAL and CAL leagues are lower level leagues, but still could have a good team dominating. We will have to wait until they play other teams in tournaments. Other top leagues are the EAL League..I think that is second to the SDL. ALso primetime has a league which in my opinion is third with level of dificulty. It is unfortunate that the Primetime did not have a TGPL league for the U9 age group this Spring would have been nice.
Anyway there are top 01 teams playing maybe up...
#1. Rowdies 02( Dallas Texans Top Team)
2. FCD Grubb 02- not sure where they are playing?
3. Fever 02- playing up in primetime 01's

IN SDL there are three Divisions:
SDL Division 1- has 7 teams rank them with the top three.
then SDL Division 2- there are 12 teams rank them next
then the SDL Division 3- teams there are 13 teams.

The 02 group needs someone to do standings first with all the leagues before rankings can come out with all the different Leagues.

They do! though it's a small league. 11v11

Quest FC Fireflies
Waco Lady Blast
Fever FC 02 Black
Southlake Saints 02G
SRSA 03
DTS Adames 03

First time poster, but always enjoy these forums and all the opinions and discussions. Have some information:

Dynasty is a new team/club from the SW FW suburbs consisting of players who would have a tough time getting to practices with the more urban clubs. They practice in Granbury in Hood county and consist of girls from the surrounding areas. Dynasty is currently independent and tries to compete well with the best teams in the metroplex. Had trouble with a good Solar Murin team in December, but I doubt any player (or parent) really wanted to be out playing in arctic weather with high crosswinds that weekend. Playing in the Lightning Cup and looking for other good spring/summer tournaments too. PAL is the closest thing, still it takes an hour to get to games. A Dynasty U8 team is practicing now too. Thanks Gophers for including everyone in your rankings!

kickback49 wrote:

Okay for scores and rankings you cannot compare different leagues as most of the top teams play in the SDL league. The PAL and CAL leagues are lower level leagues, but still could have a good team dominating. We will have to wait until they play other teams in tournaments. Other top leagues are the EAL League..I think that is second to the SDL. ALso primetime has a league which in my opinion is third with level of dificulty. It is unfortunate that the Primetime did not have a TGPL league for the U9 age group this Spring would have been nice.

Anyway there are top 01 teams playing maybe up...
#1. Rowdies 02( Dallas Texans Top Team)
2. FCD Grubb 02- not sure where they are playing?
3. Fever 02- playing up in primetime 01's

IN SDL there are three Divisions:
SDL Division 1- has 7 teams rank them with the top three.
then SDL Division 2- there are 12 teams rank them next
then the SDL Division 3- teams there are 13 teams.

The 02 group needs someone to do standings first with all the leagues before rankings can come out with all the different Leagues.

Hey Kickback,

First of all, thank you for caring enough to give your input.

Let me lay out my philosophy on rankings so that people know where I'm coming from. I've done this ad nauseum for the '01's, but probably need to post it here for the '02's...

First, I firmly believe that no single ranking system is "the best" and no system is anywhere near perfect.

Second, I firmly believe in having multiple types of rankings including both formula based "computer" rankings, as well as human opinion polls. Different types of rankings emphasize different factors, which are all valid. So, the more, different types of input you can get, the more accurate your final averaging will be. I by no means am under any delusion that FBR is better than any other system, just different.

The question is really whether or not the '02's want or care to get into this as deeply as the '01's have. I really don't care either way. I offered up my services to publish an FBR for '02's and got ~80% positive that said they'd like to see them, so I'll publish them. I've also offered to publish some form of a "BCS" that averages the FBR with one or more alternate ranking systems because it's fairly easy to do that using the FBR software package. However, it's up to the 02's to decide if and how they want to do it.

Now, specifically regarding formula based rankings, those type of ranking systems will be less "noisy" with more data (it a matter of statistics, pure and simple). Now, by more data, I mean not just more games in the system, but also, more interaction between teams of varying strength. In my initial '02 FBR post (http://www.txsoccer.net/t5555-initial-02-girls-fbr-rankings) I gave quite a bit of detail on how I am assigning teams to different ranking "Tiers" to help with getting an accurate SoS calcuation. I believe that the initial assignments do generally follow the league strength as you suggested, maybe not exactly, but I think they are pretty close. Now, frankly, there is no way around making some assumptions when doing initial placement of teams. However, as time goes on, FBR will use a performance based promotion/relegation system (which was also detailed in the initial ranking post) that will move teams in between the ranking tiers. So as time goes on, FBR ranking tiers will NOT necessarily align with league division assignments, but will rather be determined by a team's recent historical performance (past 7 games). This has proven to work very well over time with the '01's. Also, you will see that every week when I post FBR rankings, at the bottom of the post, I include a table that is a summary of the records of games played between teams in different ranking tiers. This is used as an overall sanity check to make sure that ranking tier assignments are doing their job in grouping teams of relatively similar strength together, so that the weighting factors we apply to the different tiers have their desired effect in the SoS and Weighted Win parts of the FBR formula.

Given that I've just started the FBR rankings for the '02's and only have data from January forward, FBR likely doesn't have a great read on the relative strength of these leagues. What FBR needs is what I call "crossover" data, which is data where a team or teams playing in a particular league/division have played some games against a team(s) from a different league/division. This comes from teams that play in multiple leagues, or via tournament results. So far, crossover data is somewhat limited. I have some from January tournaments and I also some data from teams like Cosmos who are playing SDL & CAL, DT South '02 who are playing SDL & PTSAL, Polaris who played a Jan tourney against SDL teams & APAL. That gives me a little better read on the relative strength of the various leagues (and thus the teams within those leagues) than if I didn't have it, but I will be the first to admit that I would certainly like to have more.

So, regarding a team like Dynasty... the fact of the matter is that we don't have a particularly good read on this team yet. However, there isn't any data that suggests unequivocally that they aren't a top 10 team at this point, only opinion. If they show up in some Spring tourneys and get hammered by other top 15-20 teams, then they will fall in the FBR rankings. Until then, as long as they keep dominating APAL, they will stay right about where they are, which is somewhere in the 9-11 range. They won't go much higher because their opponents in PAL are all from lower ranking tiers, so their SoS will always be much lower than the teams playing SDL D1 and any team that dominates SDL D2 in a manner similar to how Dynasty is currently dominating APAL.

Now, since I'm obviously on a roll, a few comments regarding what teams/games I include in the rankings, and how I approach "guest" playing...

My philosophy with FBR is that I want to include any and all '02 teams from NTX that are playing in recognized academy leagues and/or academy tournaments. Now, the reality is that academy is to some degree a free-for-all. Guest playing happens all of the time, both in league games and in tourneys. Roster sharing and "All-Star" teams (i.e. TFC Elite) happen all of the time. Some teams play under the same name in different leagues, but have significantly different rosters in each league. These are simple facts of NTX academy soccer life. I have no way of knowing who guested with who, who is sharing players, and frankly, how much impact did that really have on the game result. The bottom line is that if your team is signed up to play in a recognized league or tourney, I am going to try to include them in the FBR rankings without prejudice. The only requirement that I have is that I can get a sufficient number of games (minimum of 3) between that team and other teams that are in the '02 ranking database that I can start to get some kind of accurate SoS calculation for the rankings.

I include tourneys in FBR because I consider those games to be equally as valid as league games in terms of competitive assessment of a team's strength. Also, tourney are one of best sources of "crossover" data that help to ensure that I have the ranking tiers dialed in.

The place where guest playing, roster sharing, etc. should come into play is human opinion polls. That is a much more appropriate avenue for applying those kind of filter's, and as I've said before, I don't consider human polls to be superior or inferior to computer polls, just a different, equally valid viewpoint.

Also, at U9, I'm not going to make any differentiation between teams/games at 9v9 vs 11v11. Just way too hard to do. If I carry the '02 FBR rankings over to U10 next fall, I will switch to only including teams playing in 11 v 11 leagues/tourneys. This is what we did with the '01's and frankly makes sense because any team that will have any relevancy come U11 QT time is going to have to start playing 11 v 11 at some point in U10.

Now... how should people interpret the rankings...

The first reason for doing rankings is simply because they are fun to do, and provide fodder for endless forum discussion and debate. Nothing more, nothing less. At the end of the day, we are ranking 8-9 year old girls soccer teams, which is absolutely absurd. But, no one is getting hurt by doing so, as long as everyone keeps it in perspective.

In the grand scheme of things, it really doesn't matter whether Team A is ranked a couple of spots higher or lower than Team B. No one is getting a trophy or a t-shirt based on the rankings. No one is getting placed in Lake Highlands based on the rankings. While I try to make the FBR rankings be as accurate an assessment of the relative strengths of the teams as possible, in the end I do not get hung up over teams being out of place 2 or 3 or 4 spots from where anyone "thinks" they should be, unless when you look at the data, there is an obvious, no-brainer set of results that doesn't make sense (i.e. a team is ranked below 3 or 4 different teams that they have played and never lost to because we have one of the weighting factors in our formula out of whack). What I do try to do is get the relative rankings of groups of teams right. In other words, the group of teams I have at 1-5 should as a whole be better that the group at 6-10, should be better than 11-15, and so on down the line.

Where there is some possible tangible benefit to the rankings is for Coaches/Managers/TD's/LD's to identify and group teams of relatively equal strength together. For the most part, all of these people want to put their teams in game situations that are competitive. No one benefits from blow-outs on either side of the ball. If we've done our job correctly with the rankings (and historical data from last fall onward with the '01's suggests that we have), then you should get the following behavior...

Games between teams ranked within 5 spots of each other should generally be very competitive (i.e. goal differential 2 goals or less, wins by lower ranked teams aren't necessarily uncommon or a surprise).

Games between teams ranked 5-10 spots from each other should also be very competitive, (GD 2 goals or less), but the higher ranked team will get a W in a large majority of those games (>70% or so).

Games between teams ranked >10 different from each other will generally not be very competitive and it will be rare that the lower ranked team even manages a draw in those game.

O.K. I'll step down off the soapbox now. If you were able to make it through this entire message, hopefully you've got a little better perspective of where I'm coming from with this stuff. I also think that you will find that I am always willing to listen when somebody thinks something in the rankings is screwed up, but in the end, I am very data driven. Unless the data paints a clear picture that something is out of whack, I am going to be very hesitant to tweak things within the model.

In the end, none of this really matters and it's just for fun. My final words of wisdom (or stupidity, your choice) regarding the FBR rankings are:

Like 'em? Great! Enjoy them!
Hate 'em? Great! Ignore them!

bwgophers wrote:

kickback49 wrote:

Okay for scores and rankings you cannot compare different leagues as most of the top teams play in the SDL league. The PAL and CAL leagues are lower level leagues, but still could have a good team dominating. We will have to wait until they play other teams in tournaments. Other top leagues are the EAL League..I think that is second to the SDL. ALso primetime has a league which in my opinion is third with level of dificulty. It is unfortunate that the Primetime did not have a TGPL league for the U9 age group this Spring would have been nice.

Anyway there are top 01 teams playing maybe up...
#1. Rowdies 02( Dallas Texans Top Team)
2. FCD Grubb 02- not sure where they are playing?
3. Fever 02- playing up in primetime 01's

IN SDL there are three Divisions:
SDL Division 1- has 7 teams rank them with the top three.
then SDL Division 2- there are 12 teams rank them next
then the SDL Division 3- teams there are 13 teams.

The 02 group needs someone to do standings first with all the leagues before rankings can come out with all the different Leagues.

Hey Kickback,

First of all, thank you for caring enough to give your input.

Let me lay out my philosophy on rankings so that people know where I'm coming from. I've done this ad nauseum for the '01's, but probably need to post it here for the '02's...

First, I firmly believe that no single ranking system is "the best" and no system is anywhere near perfect.

Second, I firmly believe in having multiple types of rankings including both formula based "computer" rankings, as well as human opinion polls. Different types of rankings emphasize different factors, which are all valid. So, the more, different types of input you can get, the more accurate your final averaging will be. I by no means am under any delusion that FBR is better than any other system, just different.

The question is really whether or not the '02's want or care to get into this as deeply as the '01's have. I really don't care either way. I offered up my services to publish an FBR for '02's and got ~80% positive that said they'd like to see them, so I'll publish them. I've also offered to publish some form of a "BCS" that averages the FBR with one or more alternate ranking systems because it's fairly easy to do that using the FBR software package. However, it's up to the 02's to decide if and how they want to do it.

Now, specifically regarding formula based rankings, those type of ranking systems will be less "noisy" with more data (it a matter of statistics, pure and simple). Now, by more data, I mean not just more games in the system, but also, more interaction between teams of varying strength. In my initial '02 FBR post (http://www.txsoccer.net/t5555-initial-02-girls-fbr-rankings) I gave quite a bit of detail on how I am assigning teams to different ranking "Tiers" to help with getting an accurate SoS calcuation. I believe that the initial assignments do generally follow the league strength as you suggested, maybe not exactly, but I think they are pretty close. Now, frankly, there is no way around making some assumptions when doing initial placement of teams. However, as time goes on, FBR will use a performance based promotion/relegation system (which was also detailed in the initial ranking post) that will move teams in between the ranking tiers. So as time goes on, FBR ranking tiers will NOT necessarily align with league division assignments, but will rather be determined by a team's recent historical performance (past 7 games). This has proven to work very well over time with the '01's. Also, you will see that every week when I post FBR rankings, at the bottom of the post, I include a table that is a summary of the records of games played between teams in different ranking tiers. This is used as an overall sanity check to make sure that ranking tier assignments are doing their job in grouping teams of relatively similar strength together, so that the weighting factors we apply to the different tiers have their desired effect in the SoS and Weighted Win parts of the FBR formula.

Given that I've just started the FBR rankings for the '02's and only have data from January forward, FBR likely doesn't have a great read on the relative strength of these leagues. What FBR needs is what I call "crossover" data, which is data where a team or teams playing in a particular league/division have played some games against a team(s) from a different league/division. This comes from teams that play in multiple leagues, or via tournament results. So far, crossover data is somewhat limited. I have some from January tournaments and I also some data from teams like Cosmos who are playing SDL & CAL, DT South '02 who are playing SDL & PTSAL, Polaris who played a Jan tourney against SDL teams & APAL. That gives me a little better read on the relative strength of the various leagues (and thus the teams within those leagues) than if I didn't have it, but I will be the first to admit that I would certainly like to have more.

So, regarding a team like Dynasty... the fact of the matter is that we don't have a particularly good read on this team yet. However, there isn't any data that suggests unequivocally that they aren't a top 10 team at this point, only opinion. If they show up in some Spring tourneys and get hammered by other top 15-20 teams, then they will fall in the FBR rankings. Until then, as long as they keep dominating APAL, they will stay right about where they are, which is somewhere in the 9-11 range. They won't go much higher because their opponents in PAL are all from lower ranking tiers, so their SoS will always be much lower than the teams playing SDL D1 and any team that dominates SDL D2 in a manner similar to how Dynasty is currently dominating APAL.

Now, since I'm obviously on a roll, a few comments regarding what teams/games I include in the rankings, and how I approach "guest" playing...

My philosophy with FBR is that I want to include any and all '02 teams from NTX that are playing in recognized academy leagues and/or academy tournaments. Now, the reality is that academy is to some degree a free-for-all. Guest playing happens all of the time, both in league games and in tourneys. Roster sharing and "All-Star" teams (i.e. TFC Elite) happen all of the time. Some teams play under the same name in different leagues, but have significantly different rosters in each league. These are simple facts of NTX academy soccer life. I have no way of knowing who guested with who, who is sharing players, and frankly, how much impact did that really have on the game result. The bottom line is that if your team is signed up to play in a recognized league or tourney, I am going to try to include them in the FBR rankings without prejudice. The only requirement that I have is that I can get a sufficient number of games (minimum of 3) between that team and other teams that are in the '02 ranking database that I can start to get some kind of accurate SoS calculation for the rankings.

I include tourneys in FBR because I consider those games to be equally as valid as league games in terms of competitive assessment of a team's strength. Also, tourney are one of best sources of "crossover" data that help to ensure that I have the ranking tiers dialed in.

The place where guest playing, roster sharing, etc. should come into play is human opinion polls. That is a much more appropriate avenue for applying those kind of filter's, and as I've said before, I don't consider human polls to be superior or inferior to computer polls, just a different, equally valid viewpoint.

Also, at U9, I'm not going to make any differentiation between teams/games at 9v9 vs 11v11. Just way too hard to do. If I carry the '02 FBR rankings over to U10 next fall, I will switch to only including teams playing in 11 v 11 leagues/tourneys. This is what we did with the '01's and frankly makes sense because any team that will have any relevancy come U11 QT time is going to have to start playing 11 v 11 at some point in U10.

Now... how should people interpret the rankings...

The first reason for doing rankings is simply because they are fun to do, and provide fodder for endless forum discussion and debate. Nothing more, nothing less. At the end of the day, we are ranking 8-9 year old girls soccer teams, which is absolutely absurd. But, no one is getting hurt by doing so, as long as everyone keeps it in perspective.

In the grand scheme of things, it really doesn't matter whether Team A is ranked a couple of spots higher or lower than Team B. No one is getting a trophy or a t-shirt based on the rankings. No one is getting placed in Lake Highlands based on the rankings. While I try to make the FBR rankings be as accurate an assessment of the relative strengths of the teams as possible, in the end I do not get hung up over teams being out of place 2 or 3 or 4 spots from where anyone "thinks" they should be, unless when you look at the data, there is an obvious, no-brainer set of results that doesn't make sense (i.e. a team is ranked below 3 or 4 different teams that they have played and never lost to because we have one of the weighting factors in our formula out of whack). What I do try to do is get the relative rankings of groups of teams right. In other words, the group of teams I have at 1-5 should as a whole be better that the group at 6-10, should be better than 11-15, and so on down the line.

Where there is some possible tangible benefit to the rankings is for Coaches/Managers/TD's/LD's to identify and group teams of relatively equal strength together. For the most part, all of these people want to put their teams in game situations that are competitive. No one benefits from blow-outs on either side of the ball. If we've done our job correctly with the rankings (and historical data from last fall onward with the '01's suggests that we have), then you should get the following behavior...

Games between teams ranked within 5 spots of each other should generally be very competitive (i.e. goal differential 2 goals or less, wins by lower ranked teams aren't necessarily uncommon or a surprise).

Games between teams ranked 5-10 spots from each other should also be very competitive, (GD 2 goals or less), but the higher ranked team will get a W in a large majority of those games (>70% or so).

Games between teams ranked >10 different from each other will generally not be very competitive and it will be rare that the lower ranked team even manages a draw in those game.

O.K. I'll step down off the soapbox now. If you were able to make it through this entire message, hopefully you've got a little better perspective of where I'm coming from with this stuff. I also think that you will find that I am always willing to listen when somebody thinks something in the rankings is screwed up, but in the end, I am very data driven. Unless the data paints a clear picture that something is out of whack, I am going to be very hesitant to tweak things within the model.

In the end, none of this really matters and it's just for fun. My final words of wisdom (or stupidity, your choice) regarding the FBR rankings are:

Like 'em? Great! Enjoy them!
Hate 'em? Great! Ignore them!

I like it, fun to read, adds a little twist and different look. As you have done with the 01's, I would like to see the human polls added in to average out the FBR computer ranking for the complete "BCS" ranking as you said. RoidRage has a regular poll each week that is updated. Several others have put their opinions up at different times each week. AP Pollster added a 2nd human poll last week, in an attempt to be used with RoidRage for this purpose. The three polls together should make for a good averaged result and ranking.

BW Gopher.

Wow, not my intention to get you that into a response...wow way too much work. The standings and or rankings need to be somewhat accurate and because the 02's have not responded like the 01's, it is way off this ranking computer system thing.

So all I am suggesting is start with the SDL divisions first to give you some kind of accurate base.

For example the #57 ranked team playing Division 2 SDL tied #11 ranked team in your ranking system??? and the #57 team was playing shorthanded a few sick and one had broken her leg...and they tied 7v9 against the #11 ranked team...( not that the team is really ranked #11)

I am applauding you for trying and all the effort, but you are right, the 02 group needs someone in 02's to do this. Any takers..I am sure BW will show you how.

Thanks and I know you do a great job on the 01's ranking computer based system as it is fairly accurate to the other ranking systems. Thanks for all your time.

kickback49 wrote: BW Gopher.

Wow, not my intention to get you that into a response...wow way too much work. The standings and or rankings need to be somewhat accurate and because the 02's have not responded like the 01's, it is way off this ranking computer system thing.

So all I am suggesting is start with the SDL divisions first to give you some kind of accurate base.

For example the #57 ranked team playing Division 2 SDL tied #11 ranked team in your ranking system??? and the #57 team was playing shorthanded a few sick and one had broken her leg...and they tied 7v9 against the #11 ranked team...( not that the team is really ranked #11)

I am applauding you for trying and all the effort, but you are right, the 02 group needs someone in 02's to do this. Any takers..I am sure BW will show you how.

Thanks and I know you do a great job on the 01's ranking computer based system as it is fairly accurate to the other ranking systems. Thanks for all your time.

I think the FBR ranking system is pretty good. I would say the top teams are in the top bracket of his rankings. As more games and tournaments are played some of the other teams will fall better into place.

kickback49 wrote: BW Gopher.

Wow, not my intention to get you that into a response...wow way too much work. The standings and or rankings need to be somewhat accurate and because the 02's have not responded like the 01's, it is way off this ranking computer system thing.

So all I am suggesting is start with the SDL divisions first to give you some kind of accurate base.

For example the #57 ranked team playing Division 2 SDL tied #11 ranked team in your ranking system??? and the #57 team was playing shorthanded a few sick and one had broken her leg...and they tied 7v9 against the #11 ranked team...( not that the team is really ranked #11)

I am applauding you for trying and all the effort, but you are right, the 02 group needs someone in 02's to do this. Any takers..I am sure BW will show you how.

Thanks and I know you do a great job on the 01's ranking computer based system as it is fairly accurate to the other ranking systems. Thanks for all your time.

Kickback,

Sorry, I got myself on a bit of a roll when responding to you, and the response wasn't aimed directly at you
Rather it was aimed at the '02 parents in general to try and give a bulk answer to several different questions I've been getting about the rankings. Now that it's out there and typed up, I can pretty much point any parent in that direction and say "please read this" so that they can get an idea of where I'm coming from.

Regarding "Team #57"... The issue here is data integrity:

1) Until I got your PM last night, I only had 3 of the 5 SDL game scores for Fusion. SDL doesn't post scores or standings anywhere, so the only way for me to get the scores is for someone to post them or send them to me via PM. Thanks for doing that!
2) At least 2 of the scores that I did have don't match the scores that you sent me. Let's you and I try to get to the bottom of this via PM and then I'll get the FBR database updated.
3) The scores you sent me don't mesh with your statements above. In the scores you sent me, you have Fusion Losing 1-0 to the team ranked #11 (I actually have a 5-0 score for that game currently in FBR - again let's get this clarified via PM). You sent me a score that has Fusion getting a draw against the #52 team in FBR.
4) Sorry that your DD's team had illness and injuries that forced it to play 7 v 9, but honestly, there is no computer ranking system that can take that into account in any manner. Again, that would be one where you would have to plead you case to the forum masses (or at least to RoidRage for the '02's) and let that be factored into the human poll.

I'll PM you the scores that I have, let's get those verified and squared away, and then I'll get FBR updated. However, I will say that I am confident that I have Fusion and all of your opponents properly placed in the ranking tiers based on playing D2 SDL. Let's get the scores/opponents properly lined up and see where things fall.

kickback49 wrote: BW Gopher.

Wow, not my intention to get you that into a response...wow way too much work. The standings and or rankings need to be somewhat accurate and because the 02's have not responded like the 01's, it is way off this ranking computer system thing.

So all I am suggesting is start with the SDL divisions first to give you some kind of accurate base.

For example the #57 ranked team playing Division 2 SDL tied #11 ranked team in your ranking system??? and the #57 team was playing shorthanded a few sick and one had broken her leg...and they tied 7v9 against the #11 ranked team...( not that the team is really ranked #11)

I am applauding you for trying and all the effort, but you are right, the 02 group needs someone in 02's to do this. Any takers..I am sure BW will show you how.

Thanks and I know you do a great job on the 01's ranking computer based system as it is fairly accurate to the other ranking systems. Thanks for all your time.

Kickback - I think his program is enter data, spit out results. Will an 02 parent make it more accurate? It seems the program will be more accurate when more games and data are input. It seems to have worked well for the 01's. Also, in the 01's they avg in some human polls to account for the other nuances not related to the pure score of a game. BWGophers has offered to avg in some human polls for the 02's as well to help balance and average it out. I suggest RoidRage, as he has been publishing a poll each week, and then another rankings list or two. Several have offered rankings here and there each week, we just need someone to publish it regularly as RoidRage does each week.

anothercrazysoccerdad wrote:

kickback49 wrote: BW Gopher.

Wow, not my intention to get you that into a response...wow way too much work. The standings and or rankings need to be somewhat accurate and because the 02's have not responded like the 01's, it is way off this ranking computer system thing.

So all I am suggesting is start with the SDL divisions first to give you some kind of accurate base.

For example the #57 ranked team playing Division 2 SDL tied #11 ranked team in your ranking system??? and the #57 team was playing shorthanded a few sick and one had broken her leg...and they tied 7v9 against the #11 ranked team...( not that the team is really ranked #11)

I am applauding you for trying and all the effort, but you are right, the 02 group needs someone in 02's to do this. Any takers..I am sure BW will show you how.

Thanks and I know you do a great job on the 01's ranking computer based system as it is fairly accurate to the other ranking systems. Thanks for all your time.

Kickback - I think his program is enter data, spit out results. Will an 02 parent make it more accurate? It seems the program will be more accurate when more games and data are input. It seems to have worked well for the 01's. Also, in the 01's they avg in some human polls to account for the other nuances not related to the pure score of a game. BWGophers has offered to avg in some human polls for the 02's as well to help balance and average it out. I suggest RoidRage, as he has been publishing a poll each week, and then another rankings list or two. Several have offered rankings here and there each week, we just need someone to publish it regularly as RoidRage does each week.

I will do some consistent human polls with RoidRage. I just tried it last week and I think it sent a few people to the hospital with heart attacks. I just don't want that to happen again

AP Pollster wrote:

anothercrazysoccerdad wrote:

kickback49 wrote: BW Gopher.

Wow, not my intention to get you that into a response...wow way too much work. The standings and or rankings need to be somewhat accurate and because the 02's have not responded like the 01's, it is way off this ranking computer system thing.

So all I am suggesting is start with the SDL divisions first to give you some kind of accurate base.

For example the #57 ranked team playing Division 2 SDL tied #11 ranked team in your ranking system??? and the #57 team was playing shorthanded a few sick and one had broken her leg...and they tied 7v9 against the #11 ranked team...( not that the team is really ranked #11)

I am applauding you for trying and all the effort, but you are right, the 02 group needs someone in 02's to do this. Any takers..I am sure BW will show you how.

Thanks and I know you do a great job on the 01's ranking computer based system as it is fairly accurate to the other ranking systems. Thanks for all your time.

Kickback - I think his program is enter data, spit out results. Will an 02 parent make it more accurate? It seems the program will be more accurate when more games and data are input. It seems to have worked well for the 01's. Also, in the 01's they avg in some human polls to account for the other nuances not related to the pure score of a game. BWGophers has offered to avg in some human polls for the 02's as well to help balance and average it out. I suggest RoidRage, as he has been publishing a poll each week, and then another rankings list or two. Several have offered rankings here and there each week, we just need someone to publish it regularly as RoidRage does each week.

I will do some consistent human polls with RoidRage. I just tried it last week and I think it sent a few people to the hospital with heart attacks. I just don't want that to happen again

I want you to post more, so your avatar shows up more.

You can use my rankings for a BCS. I try to stay unbiased in my rankings and I will post them every week. I hope my rankings appear unbiased and I would bet most people don't even know which team I root for! I try to stay out of all the catfighting. I know my rankings aren't perfect, but I try to use my personal opinion on talent combined with actual results! I watch a bunch of games every week(I have multiple dd's) but I will admit there are some teams I've never seen in person....so I have to go off head to head results and like opponents in those cases! I'm sure most don't care what I think, but if you do care I just wanted to let you know where I was coming from!

PS....please don't pm me your guesses on who I root for! LOL

I say we try a BCS. But, I have a question....does it matter how high we go..ie..top20, top30? Reason I ask is because after about top 20-25 My knowledge of teams starts going to guessing! I suppose as more data and scores start coming in it will be easier! I believe as of now myself and AP pollster are going to do at least a top 20 every week. Anyone else want in? Also another question for Gophers....how many polls do u need to compute the BCS?

RoidRage wrote:I say we try a BCS. But, I have a question....does it matter how high we go..ie..top20, top30? Reason I ask is because after about top 20-25 My knowledge of teams starts going to guessing! I suppose as more data and scores start coming in it will be easier! I believe as of now myself and AP pollster are going to do at least a top 20 every week. Anyone else want in? Also another question for Gophers....how many polls do u need to compute the BCS?

Sounds great.