Right now you can’t reliably learn a new word, until you go through your whole review queue. What I mean is that after you add new words they might be in the middle of your queue. And I know that some people will say it is more important to review words, rather then add new ones. But there are times when I want to prioritize learning new words:
a) If I have a review backlog of 600 items, I would prefer to split it to 200 review, 10 new words. Reviewing is far more boring then learning new things. It would help me get motivated to use skritter after a break
b) Sometimes I need to learn specific words for an exam , It’s easier to learn a few words every day for a week than to review everything for 5 days, and then have only 2 days to learn everything else.
c) It allows me to predict better how many new words I will learn in a week. There is a bunch of words I just can’t remember and reviewing them takes a lot of tim, which I sometimes don’t have, and the benefit of being stuck seems negligible. I would prefer to have an option to review all but those words, then go to learn new ones, and then back to reviewing.
d) Sometimes I just don’t feel like reviewing. Given a choice between efficiently reviewing, inefficiently learning a few words, and not using skritter, if only first and third are an option I often pick doing nothing.
Similiar system is employed in memrise and many other learning apps, and I must say I prefer it greatly.
Hey, thanks for writing out your thoughts. You make some good, valid points, and we here on the team definitely have similar ideas. Tweaking the SRS is a little tricky because it has to work for a lot of use cases. We have brand new users with 5 items, people who study religiously every day with only 100 items in their queue who are constantly adding new items, and people who’ve been gone for months with thousands of reviews. So defining how to re-prioritize item order and tweak scheduling is definitely an art that we’re constantly thinking about, but are pretty conservative about actually changing because it can have lots of weird effects that take time to test.
Though SRS tweaks to keep you motivated are certainly on our agenda, our more short-term solution for users like you is to improve our auto-adding criteria and to offer more study modes. A single section review mode will help you to bypass your queue and zero in on critical new vocab you need to learn quickly for tests.
We’re in a technical transition between databases and APIs, which is a huge undertaking, but opens a lot of doors. We’re beginning to see the fruits of the labor in some small, subtle ways (speed boosts, small features and little tweaks that weren’t possible before), but will keep increasing in momentum as more things are transitioned over. The end result is that we will have more flexible control over the data, be able to more rapidly develop features, and be able to build features not possible before.
I think we’ve all been under a mountain of reviews of some point, and while SRS is great, it’s not the only path to learning and mastering characters, and not always the tool you need or want all the time. As a team we want to reflect that more in the functionality of the application and offer alternative routes to users in addition to that classic study mode. Right now, we’re still on the path to being able to deliver that functionality, but making situations such as those you mentioned something you’re eager to tackle with Skritter is absolutely what we’re headed towards.