Creating good races

People watch sports for two reasons: To cheer for their favorite team/player, and to see the unexpected. These two things must exist to get people interested in the event and to have any chance of them coming back for the next one. And while it’s far from being a sport, the same rules apply to Hot Wheels racing.

Everyone likes the unexpected

Lets look at real car racing. Every NASCAR fan  has their favorite driver/car that they cheer for. They want that car to win. But you know everyone is also watching to see the wrecks, the crashes, the fireballs, the underdog, the photo finish…all the events you can’t exactly bet on. The only thing you can bet on is a bunch of left hand turns.

In diecast racing, all you can bet on is the cars going down a hill. Outside of that, things should be left to chance. Maybe…

Rules rarely keep things interesting

When I setup the Fantasy League, I used a rating system that ranked cars based on their performance. The result was weekly races that matched up cars that were similar in performance, hoping to avoid complete blow-out drags. At first this seemed great, but after a few weeks it kind of flipped because you could pretty much bank on winners without much worry. You can’t avoid sure bets in races, but you can try to keep them to a minimum.

The flip side of that is making everything random. Random cars, random match-ups, and so on. This sounds great because it eliminates any bias but this can also lead to a bunch of boring races because the chances of having fast cars racing against slow cars is pretty good. The solution is hand picking every race, right? That sounds great but that can be a lot of work to do for each race, especially if you have a lot of cars participating.

The best of both worlds

The solution? A limited population of hand-picked cars. Instead of making all cars in my collection available for play, about 200 cars, I’m dropping the pool down to around 50 cars. And thanks to the rating data, I can hand pick those 50 cars so they are all comparable within reason. This way I can still apply random selection and get some interesting, unbiased match-ups. And believe me, a 16-car bracket pulling from a population of 50 cars offers up plenty of possible race combinations. There’s lots of room for surprise without being completely unmanaged, which is just what you want.

All of this theory works in context to the Redline Derby Fantasy League, which pulls from a known set of cars. But if you’re running an open tournament where people can bring their own cars it gets a little hairier. You just have to rely on knowledge and whatever ratings data you can find to create the match-ups. Or you create separate tournaments by classes of cars to limit entry.

A usual, you have to find a balance for things to work out for the best. You don’t want too much accuracy or people get bored, but you don’t want too much random or the fun is lost.