On 11 July 2012 10:49, pharmtech <[email protected]> wrote: > On Jul 11, 12:02 am, Peter Bienstman <[email protected]> wrote: > > On Tuesday, July 10, 2012 01:11:33 PM pharmtech wrote: > > > > > I agree. But, isn't there a way to catch up in a more efficient manner > > > than just working through the backlog of cards in their order due? > > > > When I said 'most urgent first', I did not mean 'most overdue first', > but rather > > 'shortest interval first'. > > That's great. I thought urgent meant "most overdue." > > I never thought it was as simple as "shortest interval." I don't think > it's the same thing as "risk of forgetting" (overdue / interval). For > example: > > I've been thinking the same thing. The problem is meaningful in the situation where you've been either away for a long time, or you've been dealing with an almost never-ending backlog.
For example with my Chinese deck, I went away for a conference, then had too many reviews to deal with when I came back, got overwhelmed and soon had a backlog of 1000 cards (!!). That was several months ago, and I was doing just enough reviews to notch off a couple of cards per day from the overdue ones. There's still about 600 due. The problem with shortest interval scheduling in this case is that a card with a 5 day interval, due yesterday, will be prioritised over a card with a 5 month interval that's 4 months overdue. The cards you recently forgot (or added, although I won't add any until the backlog is gone, which will evidently take a while longer) tend to overwhelm the long-scheduled cards, even though you probably stand to lose more by neglecting them. OTOH, you could argue that quickly acting on forgotten cards is as important as maintaining the long-interval cards - I don't know which strategy is truly more productive. But scheduling cards based on an overdue ratio seems more faithful to the basic theory of SRS - we're trying to review cards shortly before we would have forgotten them. In the case where you have no backlog (just the cards scheduled for today), then the short interval cards would have a higher 'overdueness', so it would behave the same as the current method. Is it really more computationally difficult to do? Surely even on a deck with 100,000 cards, the arithmetic (even in Python... ;) would still take far less than a second. But perhaps the data structures containing the interval and due time are not optimised for this, I don't know. Oisín If you have a card with a 14-day interval come due on the first day of > your vacation, and another card with a 14-day interval come due on the > 14th (last) day of your vacation, obviously the first one is at much > greater risk of being forgotten. But, their intervals are the same. > When you begin catching up the next day, the one due yesterday is only > 7% overdue. Whereas the one which was due on your first day of > vacation is 100% overdue. > > I still think a "risk of forgetting" factor (overdue / interval) would > be a more meaningful way to prioritize the catchup effort. > > Am I wrong? > > -- > 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. > > -- 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.
