Check Your Score
The market usually views an error of 20,000 jobs as pretty significant. If a gain of 140,000 jobs is expected, and the result is only 120,000, the pundits will call it a "weak number" or a "big disappointment" etc.
In our game you start with a balance of 100,000 units. We give you a point award of 3500 on a forecast if you really nail the number. Your award declines as your prediction gets worse, but you always get something for playing.
You may well ask, "How do you know the right answer?" Good question!
In theory, one could actually count the number of jobs that existed in one month and compare it to the number in another month. The difference would be a specific and finite number, representing the change in non-farm payrolls. Let's call that number, The Truth.
The problem is that no one knows The Truth and no one will ever know it. It exists in the abstract, but it is difficult to measure.
For the purposes of the game, we use a very good regression model based upon the data series we provide. It has an excellent fit with the reported job change, and it makes good theoretical sense. We shall call that number, The Forecast.
The market reaction, however, is based upon the measurement of job growth from the BLS survey. We describe their methodology at some length. They do a good job, but there is still a lot of error. It is a survey with a confidence interval. We call that number, The BLS Report.
When you play the game, your prediction is scored based upon the idea that The Forecast accurately predicted The Truth. That truth is then reflected in a simulated BLS Report. That is why the same problem, with the same guess, will get different score awards. It all depends upon the BLS survey. Part of the lesson of the game is to see that even if one actually knows the truth, a forecast often differs dramatically from the reported BLS result.
