e-mail

About Overall Scoring and Ranking

The point-for-place system also emphasizes head-to-head competition with the athletes at hand rather than a comparison to an outside standard. We want the athletes to support and push each other, rather than watching several individuals each trying to internally reach their personal goal.

Another reason why I like the points-for-place is it is the scoring system that best lends itself to an overall scoring system that allows athletes to not necessarily compete in every event. I believe that we should allow athletes to compete in individual events only, if they prefer. The scoring for an athlete who does not compete in a particular event is simply one more than last place.

I also think there should be less events (six is ideal), and only one of each event type (only one stone, one weight for distance, one hammer, etc. per athlete). If there are more events, only one of each event type should be an overall-scoring event.

In contrast to my preference to overall scoring of a particular games, I think the decathlon system is ideal in athlete rankings. Athletes can be ranked according to two methods: they can be ranked according to the best-scoring actual overall performace, or they can be ranked by a hypothetical overall score based on their season best performances from multiple games.

In the first case, athletes can also be ranked by the total or average of their top few (3-5) competitions. The formula used in calculating the overall score may include events that were nonscoring for that competition's overall, and may exclude events that were scoring for that competition.

My recommended formula for overall scoring is designed to properly rank athletes and be compatible for events worldwide. It is a sum of six decathlon scores in the following six event groups:

1) Best stone or shot
2) Best weight for distance
3) Best hammer 
4) caber
5) weight for height*
6) best of sheaf or another event not already used in 1-3
   (a second stone, weight, or hammer).

This particular formula is designed to be compatible with almost all competitions around the world, including Scotland, Australia, New Zealand, and all parts of North America. The composition of events can vary from games to games, and there are regional and association preferences for events.

For instance, the sheaf toss is usually not contested on the U.S. West coast. The Braemar stone is usually not contested in the East coast. The heavy weight is usually not thrown for distance in Scotland (only the light weight is thrown for distance and the heavy weight is only tossed for height). I don't want a particular formula to put any pressure to standardize any particular set of events.

This formula requires the athlete to be versatile in every event group. It requires the caber and the weight for height; two core events that are a part of almost all games. The sheaf is not required as athletes can substitute a second stone, weight, or hammer.

I have an Excel file depicting the overall scoring. The fourth spreadsheet shows formulas of how an overall ranking might be calculated.

The decathlon scoring system is simply to divide the athlete's measurement by the world record in the event (as of January 01 of that year--do not change the standards in the middle of the season), then multiply by 1000.

This works for every event except the caber. For the caber, you have to use a formula that gives you maximum points for a 12:00 turn and less points for lesser turns. For more on caber formulas and scoring, see my related article.

Another thing I would recommend for viable rankings is to weigh every implement. I think it is especially important to weight the sheaves and the stones so that the performances can be handicapped. However, I would like to even weight all implements to assure that they are not grossly over or underweight.

To encourage the weights to be accurate, I would give a slight advantage for oveweight and short implements. However, to discourage, but preserve lightweight marks, I would penalize underweight and long implements. The handicap has to be conservative for heavy/short implements because we don't want to advantage athletes who might score higher with a heavy implement. We are saying that the athlete would have thrown at least this far if he had thrown an implement that was exactly the minimum weight.

On the other hand, we want to punish performances that are underweight or long, yet preserve as many performances as possible. This can be accomplished by a more severe handicap. We are thus saying that the athlete would have thrown at most this distance had he used an implement that was at least the minimum weight.

The first spreadsheet of the Excel file shows an example of world best distances of the Braemar stone by stone weight. The linear best fit has a slope of 1.8 feet per pound. Therefore, a handicap of 1 foot per pound is a conservative adjustment favoring the lighter stones, and 2 feet per pound is a sufficient penalty for underweight stones because the penalty is greater than the average atvantage.

This system is designed to preserve and properly handicap the maximum number of performances for overall comparison purposes. It also encourages equipment manufactureres to make implements that are close to but not below the minimum weight. Here is a recommended table for weight and length handicapping:

Stones: 1'/lb over minimum weight, -2'/lb under minimum weight.
Sheaves: 1'/lb over minimum weight, -3'/lb under minimum weight.
Weight for height: 1"/lb over, 4"/lb under.  No length handicapping.
Heavy weight: 1'/lb over, -4'/lb under; 1'/inch short, -4'/inch long.
Light weight: 2'/lb over, -8'/lb under; 2'/inch short, -8'/inch long.
Heavy hammer: 3'/lb over, -12'/lb under; 3'/inch short, -12'/inch long.
Light hammer: 4'/lb over, -16'/lb under; 4'/inch short, -16'/inch long.

Finally, I recommend a conversion factor for masters and lightweights who throw a #42 weight in a competition instead of a #56.

In contrast to the conservative overweight handicapping/underweight penalizing, the #42-#56 conversion must be the mean ("most likely") best fit. This is because we don't want to favor either weight. Many athletes throw both the #42 in masters or lightweight competition as well as the #56 in open competition.

The Excel file shows that a simple 3' handicap is a good fit for the weight for height. For the distance event, the linear 0-intercept best fit is to multiply the #42 performance by 0.71 to get the equivalent #56 distance. These were based on the NASGA database, and included all 2002 performances by athletes that had at least one performance with both the #42 and #56.

In both cases, if the #42 weight is overweight or underweight, the weight is first handicapped to #42, then converted to #56.