In this article I want to try and articulate my reasoning about how I will try to articulate the Hallam Chase. I have written a technical article on the same topic but here, I will avoid discussing technical aspects.
What is a handicapped race?
One way of thinking of a handicapped running race is that it’s a normal ‘scratch’ race in reverse.
In a scratch race competitors start at the same time and finish at different times.
In a handicap race competitors start at different times, we want them to all finish at the same time!
Of course, in reality we can’t achieve this. We use the ‘expected’ performance of a runner to determine their handicap but it is an imprecise process. A runner’s performance can vary from day to day and race to race. Therefore an honourable objective of the race handicapper is to make sure that:
The winner is the runner that gives their best personal performance
A nice feature of the handicapped race is that technically, the winner could be anyone, male or female, young or old. If you give an outsized performance on the day you will do well.
Problems of prediction
Unfortunately our ability to predict the expected performance of a runner can vary widely. For example:
- Runner A has run lots of local races and performs very consistently, we can use each of their race results to make separate Hallam Chase predictions. Each prediction closely agrees.
- Runner B only runs parkrun and their time varies widely from week to week. We try to combine all of their parkrun results in a clever way to make a single Hallam Chase prediction.
Of course, in each of these cases we only know how good our prediction is after they run the race.
Let us pretend that our race consists of lots of runner As and lots of runner Bs:
- Our group of As roughly perform as we expect, some run slightly faster than expected, some run slightly slower than expected
- Our group of Bs vary widely, as it was difficult to make a prediction. Some run much faster than we expected, some much slower.
In the scenario above we would expect our winners to always come from the group of Bs.
The winner is not necessarily the person that gave the best personal performance, but someone that we struggled to predict. In addition, even if a runner from group A gives an amazing personal performance, they are very unlikely to win.
In this scenario runners are inadvertently penalised if:
- they have run a lot of races
- they have performed very predictably
- they have run the Hallam Chase in previous years
Accounting for uncertainty
When we create our handicaps we want to account for how certain we are of our predictions. If we have a lot of good data for a runner their handicap will be largely unaltered. However, if we have very little data, or the runner behaves unpredicably, we will account for this by altering the handicap. I.e. their handicap will be larger if we are less confident of their performance.
Getting a good handicap
If you’re a regular runner that has competed in the Chase before, we want to be able to make sure you get a good handicap. This means that runners we’re less certain off may start further back than they might expect. We have a lot of data for local runs that we can use to calculate handicaps and we’re going to try and keep adding more races as time goes on.
Things to do to help with handicapping:
- Do the Hallam Chase multiple years in a row!
- Run in local fell races, we’ll put a list of races we have on the website
- Give us as much information as you can when you enter the race. Our entry form should allow this