Handicap races were particularly attractive to me for the simple reasons that I am no longer young and even if I were, my success in the race would be limited by a clear lack of talent. In recent years, I have raced in handicap racing series organized by different clubs and was as surprised by the popularity of these races as the anger and venom felt when people felt unfairly treated.
Among the skits I witnessed were a pair of five-mile races where the winner in the first exceeded his handicap time by four minutes and followed two weeks later by exceeding the goal by five minutes. Several people who ran big PBs in the second race were understandably unimpressed with the handicap goals. Worse still was the series of twelve races (for the disabled) in which the same person won 10 in a row. No adjustments had been made to the runners’ goals at any point during the series, although it was clear that some were easy to beat while others had a hard time approaching remotely.
The secretary of my current club (Beverley AC) asked me to find ways to improve the means of calculating handicap goals. These days where almost anything can be found on the Internet, I was surprised to find that a Google search for “running handicap” returned virtually nothing. No software. There is no discussion forum. No methods. There was plenty of horse racing related material and equal amounts to keep and record golf handicaps. But to execute it was a waste of information.
It seemed to me that any viable method of reaching the objectives had to meet a number of criteria …
1. Runners should be able to understand how you had reached your goal
2. The method must be applied equally to all
3. The objectives must be verifiable
4. Goals should reflect the rider’s current skill level.
And that if these criteria could be met, we would have the opportunity to please at least most people most of the time.
He had heard of a number of “methods” used to reach goals, some of which involved little more than a group of guys in a group (the handicap committee) trying to estimate (guess?) The arrival times of people and others based on the PB of the corridors. for an arbitrary distance, usually 10K. The problem with the PP approach is not the least: when do you consider that an individual time is no longer relevant? So what do you replace it with?
Most handicapping methods also used the Riegel formula as a means of adjusting the times of a race from one distance to another. This is a formula devised by Peter Riegel from researching the performances of elite and semi-elite athletes. It takes the form t2 = t1 * (d2 / d1) ^ 1.06 and, in simple language, says that if the distance traveled doubles, the speed decreases by 6%. This formula is widely used by the various working calculators available on the internet. A much more complex formula (Cameron’s formula) tends to give quite similar results even though the reasoning is very different. Predictions only start to differ from Riegel’s formula when, for example, you predict a marathon time using a much shorter run, such as 10K, as the base time. Under these circumstances, Cameron’s formula predicts slower times.
Another problem with trying to use PBs as the basis for calculating future goals is that you can guarantee that they would have been executed under a variety of different conditions. Something hot, something cold. Some windy, some still. Some rounds on a track, others point to a point where the effects of wind and elevation change would be even more pronounced (namely the Boston Marathon, which loses nearly 900 feet between start and finish).
I came up with a simple hypothesis: what if we took into account, for each runner, their last 3 recent races and then used Riegel’s formula (or a variant) to fit each race to an equivalent distance of 10K? So if there was also a way to account for the effects of elevation changes on the course, in effect producing a “flat 10km” time, we would have all runners on roughly the same basis. Taking the average of the 3 most recent runs as the base time for the target in the next race, it will adjust the base time first for distance and then for any known elevation on the course for the target race.
The hypothesis was easily tested using sample data from past seasons, and initial calculations with a Microsoft Excel spreadsheet were encouraging. The elevation change calculations were based on work reported by Dr. Tim Noakes (the author of the monster play Lore of Running) that demonstrated the notion that, on a course that goes up as well as down, most of the Runners definitely waste time when compared to a course that is completely flat (that is, you don’t go back on the downhill all the time you lost going up). Therefore, for meaningful time comparisons, you must have some compensation when all the runners’ base times do not come from exactly the same races.
When trying to apply the system to a real situation, several things became apparent …
1. For many brokers, the system worked well
2. You could make a spreadsheet work, but it was very easy to make major mistakes that affected the accuracy of your results and were difficult to prevent or even detect. This is a common feature in many spreadsheets.
3. Even with a spreadsheet, the system needed a lot of maintenance work once it had more than a few runners. Beverley AC had 160 members, at least half of whom were active in a 10-run handicap series.
4. You cannot just take into account the races of the handicap series, if you want an accurate assessment of the Current ability you need to record all the runs performed by each individual
5. There were some runners and circumstances where, in fairness to everyone, adjustments had to be made. The challenge is to find a way to do it that is not arbitrary or open to criticism if someone challenges what you have done.
The special circumstances that needed another look were …
Runners who had not run a race in a while, say, six months or more.
Runners who get injured or ill in the short term
Runners out of the ordinary
New runners with no racing history
Most clubs that organize a series of handicap races calculate the results of the series with some type of point system. We used to operate a sliding scale that was something like 4 minutes below the target to get 10 points, three minutes below it scored 9 points and so on. By using the program to do the work, we have modified it so that the score is better than 4% less than it gives you 10 points, 3% less that it earns you 9 points, etc. This percentage system is much better balanced between long and short races and also between high-fliers and medium runners. In the past, sprinters have found it more difficult to do well in a handicap competition, partly because they tend to be very consistent (hence it’s harder to beat the target by long) and also because it’s easier to be 2 minutes below target if you run 10km in 55 minutes if you are a regular 35 minute finisher.
We came up with the following solutions …
1. When a runner is new or hasn’t run a race in a while, we don’t try to predict the weather. For the first race, we simply assume that the target time is equal to the elapsed time and award points from the center of the table (that is, six points using our system). If the next race they run is not too far from the goal based on a single race, then we let the goal stand; otherwise, it is six points again until a sensible average base time is established.
2. For runners with a short-term injury or illness, it is usually easier to exclude the poor result if the race they run is true to the above form.
3. If the drop in performance is greater and looks like it will last longer, we treat the runner as a new entrant and set a new performance standard for that runner.
Sometimes the runners perform a very different performance than expected. If it is much better, it feeds back on their average for the next race and, effectively, they pay a penalty that is close to a third of the improvement (assuming the average is more than 3 races). Conversely, sometimes you will find someone who runs a race but is not trying to get it right. “Who would do that?” you might ask. Well, someone who is training for a major event and just uses the race as another training race, it happens quite often. In these circumstances, it is just a matter of asking the runner why the time was so slow and then excluding it from the next calculation on the basis that it is not representative of the current shape.