More on how the funding projections work here: https://www.backerkit.com/blog/85-backertracker-predictions-...

If it gives a good ballpark estimate it's a game changer for these platforms.

thanks, we've found that it does take about 5-6 days before there is enough data, but from there they are a good approximation. One cool thing is that since it is a historical model, if some major event happens during your campaign (big media hit, some successful paid adverts etc), the model will adapt accordingly.

I don't mean to be negative, but this looks like a version of Kicktraq[1] with less information.

[1] https://www.kicktraq.com/

What we are trying to do is focus a bit more on building a model that is based on how projects have done in the past. Our model learns from each completed project, and helps create better predictions.

I am no Kicktraq expert, but I think their graphs boil down to something like rate of funding today * number of remaining days. I think they published a great blog post to this effect a few years ago.

You mean this one?

http://blog.kicktraq.com/the-state-of-projections/

Don't be intellectually dishonest. You aren't naive enough to not know they had this exact same functionality as big as they are and as long as they've been around.

You copied their projection charts which use a weighted historical model too. Your charts even look EXACTLY like theirs. Their projection also takes a few days to build an accurate estimation. So why did you really release this?

If you think you can do better, or have some other piece of functionality, just say so. Don't pretend you didn't know something like this didn't already exist.

Predictive modeling of kickstarter http://sidekick.epfl.ch/