I wrote a perl script (hey, I don't know python...) to calculate from
the xml file what seems to me like an obvious statistic showing how
much work your deck is going to be. I think of this as the "expected
cards per day" statistic, and It just adds up the reciprocal of each
cards "current interval". By current interval I just mean the interval
a card was given when it was last reviewed, e.g. a card that was
initially viewed 90 days ago and is now due in 10 days still has an
interval of 100.

The idea is that it tells roughly how many cards you expect to see
each day, smoothing out the randomness in the actual daily card
counts. In fact, you could implement a probabalistic pseudo-Leitner
scheme by picking cards randomly, so instead of giving a card an
interval of 100 you have a 1/100 chance of seeing it every day.  For
this scheme the statistic would actually tell you how many cards you'd
expect to see each day.

I find it fun to follow the evolution of the "expected cards"
statistic, and it also helps monitor how fast to add cards so that
things don't get away from you.

For example the current output from my Japanese deck is:

6234   62.54   1.861
<2      6      4.50
<10     78   10.84
<30    333 17.68
<100    887   14.91
<300    1595  9.25
<1000   2625   4.76
over     710    0.59

this shows that the deck has 6,234 cards, the "expected cards per day"
statistic is 62.54, and the average difficulty is 1.861.

There are 6 cards with 2 or less days to review which contribute 4.50
to the "expected cards" statistic, 78 cards with less than 10 days to
review which contribute 10.84, etc....

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"mnemosyne-proj-users" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/mnemosyne-proj-users?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to