This must be a odd question. My company asked me to design some games similar type of lumosity (http://www.lumosity.com/). I guess someone already played those games. I fight with them, saying , the algorithm behind is proprietor and gave a reference of below link:


They said, keep the algorithm in BLACK BOX and go ahead of designing the game. Now I have no idea how to start without that algorithm. So I need experts suggestions. Anyone have any idea about this, I mean how to think/design about the games? Should I simply tell them, it is not possible, you can fire me?


1 Answer 1


1) Collect small puzzles under these topics memory, attention, flexibility, speed of processing, and problem solving.

2) Give the user points based on their performance in each mini games. And finally give them a performance based points for each factors.

3) With all these points calculate a index. Find which factor they are falling back. If they are bad in memory, give them more training towards memory based games until they increase their points in those area.

The index algorithm is what is tough to form. The company should have done good research and statistics from lots of users before framing the algorithm. Otherwise it should be quite easy to roughly show a graph based on they score in different games.

  • \$\begingroup\$ Thank you so much codetiger. I have already realized that the algorithm would be tough to form. Also I think collecting/designing the games under topics memory, attention ... would be difficult too. Anyway, I got the draft idea, thanks again. \$\endgroup\$
    – samesky
    Sep 27, 2013 at 4:17
  • \$\begingroup\$ Hi Again, I have tried to understand what actually the BPI is. It is still confusing to me. For example, what will be the BPI, who played some brain training games for the first time (means without comparing other's BPI) ? BPI always depends on other's BPI to calculate ? I am so confused. \$\endgroup\$
    – samesky
    Oct 9, 2013 at 1:20
  • \$\begingroup\$ My guess: They first take a survey, calculate average/mean scores of 100s of people. Their points are based on average/mean value in each game. \$\endgroup\$
    – codetiger
    Oct 9, 2013 at 6:05
  • \$\begingroup\$ Umm..make sense. May be we cannot take a survey, so taking random score 100 or 200 times (which is not real) and calculate average/mean,then compare with it might be little fun. Thanks. \$\endgroup\$
    – samesky
    Oct 9, 2013 at 8:13
  • \$\begingroup\$ Yes, and later after you have enough players played the game you can take an average score and use it to improve your algorithm. \$\endgroup\$
    – codetiger
    Oct 9, 2013 at 10:31

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .