Feb 23, 2011

Place Your Teams at the Right Level

Parity is every leagues dream.  The years of the Montreal Canadians only losing one game all season (1976-77) are over, and no team should go all season without being challenged or not winning a game.  It has been said that if you win between 30 and 70 percent of your games you were placed at the right level, anything more or less and you were in the wrong place.  Recent years New jersey Youth Hockey League has introduced Mite and Squirt Jamnborees.  The Jamboree is a mini tournament that takes place over 2 weekends, each team particpating in 2 days of a series of 30 minute games.  Teams are placed in catagories based on information provided by their organziation, but the final division’s will be determined by the league based on observations made at the Jamboree.  Better yet division are given names like Campbell or Adams as opposed to letters like A or B.  By all accounts the jamboree has been a huge success as greater parity has created a better experience for all involved.

The Cleveland Surburban League (CSL) has taken this a step further.  They have introduced an 8 game pre-season at all levels.  YouthHockey.com makes this far more practical, as YouthHockey.com allows you to change levels, divisions, team name and schedule anytime.  (so far CSL does it without us, but we hope to change that).  We would be happy to help your league implement a mini-season.  Details on CSL’s system is  available below:

Description of Preseason Ladder Schedule and Use of Relative Power Index (RPI) Rankings to Place Teams for the Regular Season.


In the 2004/5 season the CSHL instituted a formal program designed to place teams at the correct level of play (e.g. AA, A, B) within the various divisions of play (bantam, peewee, squirt, mite) the CSHL offered.

Previously, the League allowed programs to place their teams subject to several rules that attempted to force programs to put teams in the right level, including quotas on second year players, sandbag rules, and rules prohibiting a program from placing two teams in one level.  These crude rules worked occasionally, but oftentimes seemed to require arbitrary team placement, unrelated to the true level of play of that team.  Another consequence of this structure was that, although teams were required to play a preseason schedule, the results had no impact on final team placement.

Consequently, the League took action by deciding to use a Relative Power Index (RPI) to rank all teams in a division based on their performance in an 8 game preseason (6 games for mites).  Additionally, to increase the quality of the games, instead of random games, the League asked the programs to estimate where its teams fit on a scale from 1 (AA strong) to 15 (B2 weak), and has each team play an 8 game schedule against other teams that are estimated to be in the same strength range.  This ladder schedule has improved the quality of the preseason games by increasing the odds that the two teams are reasonably competitive with each other.

All the results of the preseason games are input into a computer program that takes into account the performance of an individual team and the performance of a team’s opponents and ranks all the teams relative to each other.  For example this year there are 40 bantam teams.  The League preliminarily ranked them from 1 through 40 based on their program strength estimates.  You can see the results of this preliminary ranking by the team number assigned to each team.  Bantam teams are numbers 401 through 440.  The CSHL website has information about preseason team placement under the 2005-6 Preseason Information button on the left side of the home page.

Only games scheduled at the preseason scheduling meeting count towards the RPI.  Tournament and scrimmage game are not included.

The preseason is 6 weekends long, typically running from mid-September to the end of October, and we post the first RPI rankings of teams after the 3rd weekend of play.   We do not post earlier because the RPI system needs a critical mass of data before its results start to make sense.  Approximately 40-50% of the preseason games will be done by then.  After that we will post new rankings after roughly every 150 games, with the final rankings posted late the last night of the preseason.

Because the CSHL schedule consists of a home and away series against all other teams in a level, there is a 14 week regular season from mid November to mid February and the League wants to limit level sizes to 7 (12 games), 8 (14 games) or max 9 (16 games) teams per level. 

Therefore, the League will determine the number of teams in a level primarily on where, within reason, a talent break exists between teams, but will start out with a framework as follows (number of teams from the 05/06 season):

Bantam (38 teams)       Peewees (48 teams)     Squirts (45 teams)        Mites (37 teams)

7 AA                                       8 AA                            7 AA                           7 AA

7 A1                                        8 A1                             7 A1                            8 A1

8 A2                                        8 A2                             7 A2                            8 A2

8 A3                                        8 A3                             8 A3                            7 A3

8 B                                          8 B1                              8 B1                            7 B

                                                8 B2                              8 B2    

Petition Process.

If a program believes there are facts causing one of its teams (or another program’s team) to be misplaced by the RPI system, a petition process exists to permit the program to raise the issue with the CSHL Board and have them review the facts and the preseason results and determine if a team is misplaced.  Last year out of approximately 180 teams, 10 teams petitioned and 5 were granted. 

The petition can only be submitted by a head of program (no coaches), it can only be submitted via email to the League Secretary, Terry Horonzy, and it must be submitted by 5pm on the Monday following the end of the preseason.  The petition should include FACTS about the team and why it is misplaced.  A request of the 10th ranked bantam team to play up to AA because “the coach wants to challenge his players” is not an acceptable reason; EVERY coach wants to challenge his players and the League has to also keep in mind the interests of the AA level teams whose coaches want to challenge their players by playing in the most competitive level the League can come up with.  A petition of the team ranked 10th to play up should also include reasons why the 9th ranked team should now be dropped to A1, since the League is indisposed to have a 10 team level this year.

An example from last year of a successful petition was the presence of impact players during the preseason that were unexpectedly not going to be there during the regular season (moving or high school), which allowed a team to drop a level; or vice versa, an impact player joining the team late in the preseason, coming back from an injury, moving into the area or a Barons player quitting the Barons and returning to his program late in the preseason.  The fact that some of the impact players on a team play football and therefore miss several games is not a good reason; the coach should rearrange his preseason schedule to account for the absences. 

Another successful petition was from a program that did not want two of its older teams in the same level; one at the top of the level and one at the bottom of the level, so either could move.

To summarize; a successful petition will avoid generic complaints about the accuracy of the RPI system (it has proven itself fairly reliable), and will concentrate on the FACTS causing the misplacement.

The League’s goal is to provide as competitive and fair an experience as possible for every team in the League, from AA to B2.  The RPI system is a big step forward towards this goal.