Algorithms and Staying Gold
Anybody have particular strategies for skills staying gold for longer? I finished my French tree gold, while going through the tree slow and mostly only letting myself take on new skills when all the others are still gilded. I've regilded it a couple times since if memory serves, but still a week can go by and it already shows 35 more skills I need to practice. Not sure how the algorithms work but I think trees in other languages seem to stay gold longer? but maybe that is just my experience. I tend to like the timed practice, but maybe I should do more untimed? If I kept on a skill after it was gold would it stay longer? Do algorithms not change at all with % fluency?
I like to keep everything gold. I agree with Actiregularis that finishing without mistakes seems to keep them gold longer. I also discovered that continually reviewing the final twenty skills seems to keep the entire tree gold -- probably because all skills and most vocabulary appear in the advanced skills.
After 60%, it takes a lot of work to add percentage points, doesn't it. it took me ten days to move from 67% to 68%.
Anybody have particular strategies for skills staying gold for longer?
This is my favourite method
Duolingo's web version (www.duolingo.com) strengthens much better than the App
Not sure how the algorithms work ......
Not in my account in the Android App.
You can easely test it in your own account both in the Web version and in "Android on Bluestacks":
- by strengthen a skill and count the words, which are strengthened in the "Words tab"
In one strengthening in the web version, my "Words tab" showes up to 4 times more strengthened words than in the Android App.
(Strengthening method: https://www.duolingo.com/comment/6176795/Best-Way-to-Make-Your-Tree-Turn-Gold-and-Stay-Gold)
Maybe, it depends of the A/B testing in our accounts.
I find if I do several timed practices in a row with few errors (and few hints), then a skill will stay gilded for quite some time indeed. That said, this is something that is subject to a whole lot of A/B testing it seems, which I think makes people's experiences differ more than we might expect.
Here is an interesting post a linguist posted a couple of years ago https://www.duolingo.com/comment/6176795/Best-Way-to-Make-Your-Tree-Turn-Gold-and-Stay-Gold Bonne chance!