Information from TBC for the 2011 Fall/House season
(below information is geared towards coaches, but we encourage everyone to read)

Coaches are Critical:
We're counting on coaches to be our "eyes" and help us analyze what's going on "on the pitch".

Inform TBC of non-competitive games:
If you're in a one-sided game (regardless of if you're dominating, or being dominated), we want to hear about it.   Send an email to TBC@phmsa.org.   Let us know the date, division, teams, and how bad the discrepancy between the teams was.   Please add any other relevant information.
The score isn't necessarily the main indicator   ...   Is ball-control even between the teams, or is one team frustrating the other ?

Keep track of the strength of teams that you play (or see):
At about 2/3 of the way through the season, we will send out a "team-competitiveness" questionnaire to all coaches (head & assistant).   We will list all of the teams in your division and ask you to rate the strength/competitiveness of the teams you are familiar with (ones you've played or seen play in a game situation).   This information will help determine how well we did with team-balancing. It will also help us analyze if opportunities for improving team-balancing further exist.

Player Ratings:
Quality player-ratings cannot be stressed enough.   This drives team-balance.
From day one, as you coach during the week and during the games, start formulating ratings for your players.
We encourage you to discuss players briefly with other coaches so that we as a league rate to one standard (as opposed to every coach having their own standard).   You could discuss players with other coaches on the practice fields, before/after games, scrimmages, playoffs/tournaments, etc.   Use this as an "ice-breaker" to start up conversations with other coaches.

10-50 Range:
We want to fully utilize the 10-50 range.   By making better use of the lower and higher ranges (roughly 10-18 and 42-50), we are able to better differentiate the bulk of players (who rate in the 18-42 range).

PHMSA Team-Balance changes in recent years:

• No-buddy policy.   Avoids 3 "star" non-related players ending up on the same team.
• Select coaches re-rate Select players after Select season (most players will improve over 6 months of playing).   TBC decides adjustmnet.
• Select coaches rate other players they know.   TBC uses this as a cross-check to see if any particular coach might be higher/lower with their players compared to the division's norm.   TBC makes any necessary adjustments.
• We ask parents/guardians of unrated players for estimates of their kid's skills (instead of applying a default "25" rating to all).
• Extra effort by PHMSA of communicating to its community the importance and philosophy of team-balance.   The top priority being to get coaches to take ratings seriously and drive towards a consistent standard.
• We pool late-registered players and distribute on specific distribution dates (versus a first-in first-out approach).   Parents/Guardians and Coaches will need to be patient in most cases as we wait for the best time and most-fitting team (for team-balance) to assign their child to.
• The creation of a Team-Balance Committee, and appointing its Coordinator to the PHMSA Board of Directors.
• Trying different team-creation algorithms.   This is a "work-in-progress" exercise.   For 2011, PHMSA has attempted an algorithm outside of eAYSO and will be continuing analysis of it's algorithm compared to eAYSO's.   The eAYSO algorithm first pairs up the head/assistant coach pairs and the sibling pairs.   Then it takes the older-age players & distributes them by "ratings" order.   Then it takes the younger players & distributes them the same way.   According to eAYSO, their algorithm is meant to be a first-cut with the expectation that there will be manual adjustments.   PHMSA has used this approach in the past to attempt to achieve team-balance.   PHMSA's 2011 algorithm and team-creation process differs from eAYSO's in three significant areas:
  Age:   during the team-generation process, the ratings of all players who just moved up to a new division are decremented by a standard value.
  Weighted team-average: during team-generation, a dynamic process continuously calculates team-strength by giving stronger players on a team more weight than average & below-average players.   The player distribution algorithm drives every team to approximately the same weighted average, while also maintaining rough age balance and unrated player balance.
  Manual intervention: the expectation that teams need manual adjustment is removed as we're counting on the computer algorithm to create the "final cut" with no manual adjustments. Unfortunately, late-registrations affect team-balance, so sometimes we are forced into making adjustments.