"I had to play a big kid!"

An Introduction to Ratings
by Ken Larsen 

Probably the number one complaint at Southern Arizona Chess Association scholastic matches concerns players who are years apart. A first grader may face a sixth grader across the chess board. This can be daunting for the much younger player ... and sometimes also for the older player whose ego fears that he has lots to lose and little to gain.

Most games match players of roughly similar ages, but at SACA's League Matches occasional age disparities are unavoidable. That's because all the teams are bracketed against teams of similar overall ability based on ratings. Here's how it works:

Every player has a rating based on how well they've played thus far. This is like a career batting average, and SACA recalculates it after every tournament. Primary players in Grades K-3 start with a rating of 250, while older players start with a rating of 300. Each player's rating rises or falls depending not only on how well he has played, but also on the strength of his opponents.

For example, imagine that a player rated 368 beats a player rated 314. The winner will gain 11 points and the loser will drop 11 points based on the following formula:

16 plus or minus
1/10th of the difference in the initial ratings
equals the change in the ratings.

In this case the difference in the initial ratings amounts to 368 - 314 = 54, and 1/10th of 54 rounded down equals 5, so the change in the ratings is 16 - 5 = 11.

The stronger player won, so the initial ratings difference or ratings gap is deducted from 16 and the new ratings would be 379 and 303. If the weaker player had upset the stronger player, the ratings difference would be added to 16 and the new ratings would be 347 and 335, a swing of 21 points (16 + 5).

In short, defeating a higher-rated opponent helps your rating more than a win over a weaker opponent, while conversely a loss to a tougher opponent hurts your rating less than a loss to a weaker opponent.

Over time, some players have earned ratings as high as 700. That's a years-long achievement to be proud of, but coaches understand that in the short run, ratings are sometimes misleading. Accordingly, for SACA's league matches a coach may decide to seed one player higher than another on the school's team or teams despite their ratings. Also, a coach can always ask as a special request that a team be seeded up or down a level to better reflect the team's true ability.

Still, mere age is often a poor gauge of a player's talent. A third grader of ability is usually stronger than an inexperienced fifth grader or even sixth grader. Ratings usually confirm as much. In recent years we have had a kindergartner rated at 363 and a second grader nearing 500. A team of similar primary players may thus outrank a team of fifth and sixth grade beginners, and rightly so.

On the other hand, teams in the toughest divisions such as the Premier and Platinum divisions usually have experienced players who have built up their ratings over years of play. If such a team is missing a player or two, it will be seeded in a lower division than normal because the missing boards will count for a rating of just 150. The older players may find themselves matched against younger players, but with an automatic loss for every missing player.

Schools with too few players to field a full team, or players from schools without a chess program, or team players without an opponent in a given round will all play in the Individual Section. They are paired by grade, so age disparity isn't an issue.

One additional consideration: players always "bottom out" at their starting rating. A 4^th grader will never dip below 300 even if he loses every game. That means that an experienced player (albeit one with a poor record) will never have a lower rating than an outright beginner. This helps minimize age disparities in pairings while maintaining the advantages of a ratings-based competition.