A Method for “Gaming” Predictions for the D6 Generation

D6_badge_transparent_472_CroppedI have been listening to the D6 Generation for a few months now and enjoy the podcast. I highly recommend all gamers to listen to it. It’s long so great for listening to in parts during your commute or between classes or during your performance review. At the end of last year apparently they did gaming industry predictions. This year they did the same and scored the previous years results. They have a loose set of rules they use. It’s amusing. As a data scientist in my real life job I’d like to suggest an alternative system to them that will make it more of a game and more strategic.

Rule 1: Each prediction must have a quantitative final measure. This can be true or false or a verifiable number. “FFG will bring out a new line of miniatures in 2013.” That is good, it is either true or false. “FFG will bring out another full-of-fail Silverline game.” No, that doesn’t work because its not testable. The “full-of-fail” part of it is subjective. “FFG will do really well in 2013.” No, that is also subjective. “FFG sales will double in 2013” That follows the rules but unless FFG releases the numbers it is unverifiable. “Games Workshop’s stock price (GAW: LONDON) will be greater than 740GBP by the time of the next end of year show.” That one is good – It can be researched (and is adjustable in the case of stock splits.) It is resolved in time for the follow on show.

Rule 2: Each question has a value of 1 point, +1 points for each hosts that doubts it will come true, for the predictor. If the prediction fails to come true then the predictor looses that many points. Hosts may agree with the prediction and may win or lose 1 point if it comes to pass or fails to. Neutral votes are an automatic deduction of 1 point.

Example 1: Prediction “Paizo will introduce a D7 die in 2013!” Hosts, 2 and 3 say nay. Value of the prediction is 3. If it comes to pass the predictor will get 3 points. If it fails, they will lose 3 points.

Example 2: Prediction “Games Workshop will introduce a Tau Titan in 2013” Host 2 is all for it, Host 3 doubts it. Prediction is worth 2 points to the Predictor. It is worth 1 point to Host 2.

Example 3: Prediction “Dark Future will make a comeback in 2013 due to a Kickstarter project!” Host 2 is neutral. Host 3 is vehemently against such a possibility. The Predictor risks 2 points and Host 2 will automatically lose a point.

Rule 3: All information is public. If a host has inside information they must reveal source and all relevant data. This is an honor system rule.

This form of the game puts a consequence on every action. In order to score well you have to make outlandish predictions but they also have to come true. You can disagree with a prediction but that indicates it is risky and thus there should be payoff. You can agree with a prediction but you share some of its inherent risk.

It is unlikely that anyone will cast a neutral vote but the option is there to limit the gains another player/host might make. Having a penalty for risk avoidance means that the hosts have to make a priority of understanding the market and making informed decisions over just “winging it.”

I think these rules can make the little game of chance interesting. It would be really interesting to see them go back over the December 2012 predictions and rework them into the framework – rejecting soft predictions or modifying them and seeing how it plays out.

Speak Your Mind