A language learners’ forum

rdearman wrote:Based on a statement I found from an "anki guru" who believes that with 10 cards per day you'd end up with 200 reviews per day. I've gone back and looked at my current decks and based on 20 cards per day, I'm currently getting 100 reviews on those decks. So the figure of 200-250 is probably more correct than the statement I used earlier. So does 200 reviews per day actually hit the critieral I outlined? Will it let you see every card in a 10,000 deck, in 618 days time?

Sorry, I made a mistake in my calculations. Not 244 reps but 195 reps. So,

10,000 words
learned evenly over 600 days
6 initial reps
0 to 10.3222 revision reps (10.3222 reps for the 1st batch of words, 0 reps for the last batch)
About 10% error rate
====> average 195 reps per day over 600 days

Yes to your question - if you use Memrise.

I guess it's hard to predict the total number of reps of Anki users, because in Anki, the initial learning stage is flexible and totally up to the individual learner whether s/he wants to see a card 5 times or 50 times. Unlike in Memrise everyone is fed 6 times and most people just stick to 6. And in Anki you get to rate Easy/Good/Hard so there's a lot of room for variation. Confident people may review much less than timid or prudent people...

I must say I highly dislike the Memrise challenges as they have quite nothing to do with motivating regular users. The winner, in order to get the money (the 2nd gets nothing) usually spends unreasonable amounts of time on it (winning is incompatible with having a job, studying, sleeping enough) ,"cheats" by choosing easy decks, decks without writing, decks they know are fast to click through. Yes, you cannot cheat in the sense of telling people you've got more points than you've really got. But the way to win is definitely not the way to learn the most, to retain the most. I think I've read all the interviews with the winners (I think there have already been two or three rounds of the challenge) and I really don't think they learnt much or that the memorisation will do them any good in the long run.

I'd say the one sensible rule option could be something like this:
-any SRS would do, tracking would be done either with a bot (if anyone has the skills and will to create one), or on a thread (perhaps one post per person, perhaps a good use for our beautiful progress bars)
-any kind of "card"-1 word cards, sentence cards, conjugation cards, cloze deletion. the point is to help people with motivation to achieve their own goals, not to create a pointless clicking obsession the memrise challenges turn into
-10000 reviews OR 10000 new cards, no matter whether in one deck or more decks, made or downloaded
-the key making it a challenge: a deadline. The Super Challenge is a time frame of 20 months. So, how much time do we need for 10000 reviews or cards?

Cavesa wrote:-10000 reviews OR 10000 new cards

What ways are there to track the number of reviews?

I think it has to be reviews. After all I could generate 10,000 cards in seconds with a script and hey presto I won the challenge.

smallwhite wrote:

Cavesa wrote:-10000 reviews OR 10000 new cards

What ways are there to track the number of reviews?

In Anki, you always see the number of cards to learn and review left, they are even in three colors-new cards, review of newer cards, review of older cards, unless i remember it wrong. I think it was the same in the computer version and ipad one, no idea about the website or other platform.

In Memrise, you can see such a number as well. Or you know how many reviews per session you do and simply count sessions and multiply at the end.

No idea about Quizlet and other platforms though.

Reviews and new cards aren't equal, it's like comparing apples and oranges. It's breadth vs depth. The benefits aren't going to be the same. Also, to give you an idea how big the difference can be: over the last 1,252 days, I did 466,057 reviews, whilst the number of notes learnt I'd estimate to be only in the 30-35k.

Ezy Ryder wrote:Reviews and new cards aren't equal, it's like comparing apples and oranges. It's breadth vs depth. The benefits aren't going to be the same. Also, to give you an idea how big the difference can be: over the last 1,252 days, I did 466,057 reviews, whilst the number of notes learnt I'd estimate to be only in the 25-30k.

Of course, that's why I mentioned both options as possible. Each of them would probably call for a different deadline.

Each of these options is possible, it just depends on either the participants agree on or sfuqua decides.

What if we do it like this?

To win: At the end of the challenge you need to have at least 10000 cards that show up when you show click on "review" when you do browse in anki, and you have no cards show up when you click on due. If you use another SRS system you need to figure out how to find the equivalent information.

That way as long as you get through 10000 new cards, and you keep up the reviews, it gives you maximum flexibility about how you get there.

You can make the deck ahead of time, or, more interestingly, make a deck from hard sentences from your reading in the super challenge. There is really no need to make a deck ahead of time.

I'm definitely keen to join in on this challenge. Though I would be looking to do sentence cards across 2 separate language decks.

I use flashcards deluxe app but it sounds like it should be compatible for the kind of info needed to complete the challenge.