Too many repetitions means 750 days on Insane to go from level 4-5
Even with the new system, which I quite like, I still want to get all my German skills to gold. You haven't really completed the course if you don't do that.
So, I've already got four of the first five lessons at level 4 (and on at level 5). The fifth lesson, on accusative case, has 30 repetitions to get from level 4 to level 5.
The number of points you get per repetition seems to have gone up, so let's say that, with bonuses, I'm getting 15 points per repetition.
So just by doing one level of that fifth lesson, I am going to get up to 450 points. I have my daily target points set at 50, which is quite literally labelled 'Insane' by Duolingo. So it's going to take 9 days just to get that fifth lesson from level 4 to level 5.
Let's especially bear in mind here that, so far, this lesson has been the same riffs on "The man drinks water" and "The girl, the boy, the woman, the man" for four levels.
However, moving on, so far the 30 repetitions is the largest number to reach level 5. Let's take the average over those four lessons out of the first five. That's 80 repetitions, or 20 per lesson. The German tree currently has 125 lessons/skills. That's 2500 repetitions, or 37500 XP if I do each repetition well and get the 15XP each time. At the 'insane' 50XP/day, that is 750 days.
750 days just to get from level 4 to level 5. In one language.
So don't do it that way. I go through the tree doing all skills that have the lowest level. Currently, I'm doing the fourth pass on my 2-crown skills. Any skills that have a higher level, I just skip, so as I go I get fewer and fewer skills to actually do for each pass. Once I'm through and all skills have three crowns, I'll just start from the top again and do all that haven't got four crowns.
This way, I get varied repetitions of the entire tree instead of just repeating the same skill lots of times and then probably forgetting it again in a month.
That's one of the reasons why I'm going to stop working on my German tree as soon as I get to level 18 and move my learning elsewhere. It's just become too boring. Would do the same with Irish but there aren't many resources for learning this language... Fortunately (sic!), the Irish tree is rather short.
The best thing I can say about the new system is it convinced me to give up on Duo on move on to spend my time on more valuable things.
The old system offered a lot of repetition that seemed intelligent and reflected your strengths and weaknesses. I felt I was still spending time usefully by practicing things that I hadn't seen for a long time, or lessons where I made frequent errors.
The new system seems to offer only mindless repetition that reflects nothing of what you've accomplished, your correct answers or your mistakes.
I was excited by the initial promise of "more advanced content" being made available but it just isn't there. There is only invariable, unresponsive grinding.
I owe Duolingo great thanks for initially sparking my interest in German, helping me to maintain motivation and good habits, and making the whole process seem like something achievable. I'm disappointed in the recent changes, and I feel that they diminish the chances of new, incoming learners feeling the same sense of progress, encouragement, and excitement.
Having been here for 5 years, everything changes. We have Crowns today. If it even still exists in the future it will likely be different from what we are interacting with now.
Duolingo never stops crunching data on things like sentence success and fail rates, user drop-off, user retention, what changes inspire users to study more each day and for more days and complete more of the course or multiple courses.
If Crowns has a statistically significant negative impact compared to the last system, Crowns won't continue as a feature. If an alternative has a statistically significant positive impact, it will replace Crowns.
0.00000002% of users liked your comment, I guess it wasn't a very good comment.
I'm obviously only kidding about this ;-), but my point is that statistics can lead you to the wrong conclusions and blind you to what is actually important.
What is the point in having 3 billion users on Duolingo who only learn five words in another language? Would it not be better to have only 1,000 users learning a language to fluency? It certainly feels like Duolingo tends to the former. And the more you rely on statistics, the less you can see the more specific use-cases. *(Although the question at the beginning of this paragraph is also up for debate. Maybe 'Duolingo' really believes that's what is best.)
Here is an article that sort of touches on the subject: https://www.theguardian.com/commentisfree/2018/may/27/world-distraction-demands-new-focus. I thought of Duolingo immediately when I thought of this. Duolingo, the small start-up which set off on such a noble goal.
I am not against statistics; they are very useful. But Duolingo has, in the past, used statistics and claimed they've "done science".
The more I use Duolingo, the less I feel like the developers are actually using it themselves seriously* (dogfooding). I don't know if you or other people on here feel the same way, but I often come across bugs that simply have no reason to be there. Things which often almost break functionality and can take years to fix (your example of the markdown quotes is a nice example, but one of the more trivial ones). These things aren't hard bugs to fix. But they do cause unnecessary frustration.
But then there's the general usage. The features, the products they launch, the things they choose to invest their time in (and, more importantly, the things they, as a result, choose not to invest their time in): Duolingo for Schools, Crowns, the Fluency Meter, iOS. Then the things that were removed: Words, Activity Streams, Skill Strength, being able to see friends' progress. I'm too tired to be able to put into words why this doesn't make any sense to me whatsoever. I just feel the lists are self-explanatory to anyone who's actually used Duolingo for an extended period of time seriously*.
Sorry this was a little disjointed. There are a number of thoughts that have been bouncing around/snowballing for the last couple of years and this is one of them. Like you, I've invested a lot of time in Duolingo and I know how beneficial and important it can be (although my experience sounds a lot less dramatic than yours). Because of that I am willing to be patient and look forward to improvements and changes. But because of that I do care when things get worse or when time is spent on the 'wrong things'. And I do care when things grow in the wrong direction.
*By seriously I mean actively trying to learn a language for fluency. Seeing that 50xp is described as 'insane' seems ridiculous to me. I feel like I've not done enough work if I've only done 50xp in a day. And I know my memory of the words will start to fail if I keep that up. But then 50xp is pretty relative. Is that 50xp of Timed Practice (which is really nothing), or 50xp on the app (which is also nothing, but that's a rant for another day), or 50xp doing the same 5 sentences for half an hour (which make take you a while, but won't teach you much).
Edit: this might be a little to technical, but it also generally touches on topics of 'feature farming' and finding out real user requirements which are related: https://www.youtube.com/watch?v=2JNXx8VdbAE.
"Duolingo never stops crunching data on things like sentence success and fail rates, user drop-off, user retention..."
That's great, but it suggests that the developers of the system are using trial and error to expand the system, rather than having a clear vision for how the language courses should evolve.
Don't get me wrong, trial and error works ... eventually (that's how biological evolution works). However, it's not a very efficient system. The problem is, when a bad feature is implemented, the users become the guinea pigs until the developers figure out that the feature is flawed - and judging by the speed with which they respond to user feedback, that could take years.
As things are right now, if we decide to stick with the program, we have to accept that the upper skill levels are broken. Our only other option is to quit and come back later in the hope that things have improved.
This is a terrible way to build and retain a community of language-learners.
I hope you read my reply as informational and not argumentative. As I have said in another discussion elsewhere, with new features, I am almost always on the "wait and see" side, rather than the "This feature should stick around for ever/this feature is bad and needs to go away".
For transparency, I like things about Crowns. I've set up this year's learning goal for myself based on Crowns. However, there are things I find less than optimal about Crowns. If Crowns gets replaced, I am neither here or there about it at this time. I think people should use what works for them, give helpful feedback to staff (both positive and negative when applicable), and after move on and focus on what works and find outside resources to fill-in areas where Duolingo is lacking.
I often reply but with comments that attempt to speak to a larger audience by filling in information that might be missing for some readers, even if it is not missing for the person I'm reply to. I have a hard time guessing who already knows what. Some people I've replied to mistake this as me trying to be condescending towards them. I do not feel condescension towards you. So, I hope you won't read me as doing that.
Tone over the internet is very difficult to navigate. So, I wanted to put all of that out there in as plain of words as possible to make up for possible sources of tone-impression interference.
That's great, but it suggests that the developers of the system are using trial and error to expand the system, rather than having a clear vision for how the language courses should evolve.
Duolingo is closely associated with Carnegie Melon University. One thing Duolingo seems to have been doing from the start is forge new discoveries for online language education.
Towards that end, Duolingo tests data on things as small (and smaller than) as the number of Duo tears shown at the end of lessons (an old thing). They found that there was a significant difference in user-retention between the number of tears Duo cried from 1-3. That's not something I would have suspected. But, because Duolingo tested it, it became a discovery that lead to increased user-retention. It was retired when the team discovered something even better.
Like that, Duolingo puts every detail, every bit of data under a microscope to test. Whether or not it is inefficient is not for me to say because that's not my area of proficiency.
The graduate students involved publish research papers on what they've learned along the way. Your assessment that Duolingo employs trial and error here is accurate. On the matter of that indicating a lack of vision, I don't see it as incompatible with vision.
Some folks here don't know that there is a distinction between course contributors and feature development teams. So, for them I want to illuminate that difference: course contributors provide vocabulary and sentences (course contributors) and at least some if not all set the order of the skills. People who make interface and feature-creation decisions (ie Crowns, skill decay, decide what to do with fail rate percentages, etc.) are staff. Staff set the fail rate percentage and decide whether to retire the sentences that fail or not. While course contributors wrote all of the material people are currently interacting with, they do not decide if learners encounter all of the material they've written.
Having laid out the distinction as I understand it, I am not doing so to say I disapprove of staff or contributors. ^_^
I hope this comment has not been too messy. I jumped here and there while writing it.
I'm reviewing the tree every year and going to the next crown up each new time I review it. The main goal is learning the language. Repetition is good. Racing to "win" maybe less good for a lot of folks. Duolingo has picked it's priority and is testing it with this first iteration of Crowns. You gotta pick yours. Do you race all of the way to level 5 for a gold tree, or, do you go at the best pace for getting the language into long-term memory. (For a small minority of folks they can do both at the same time. For most of us, we probably gotta pick one.)
My question is this. How much new content was added to this new system? When you reach level four, for example, say across the entire tree, are there more words, does the content change? I haven't been able to tell so far. Or do the harder levels purely just mean repeated content (i.e. sentences), that the learner has already been exposed to within duolingo?
I don't have all of the info as I was in an early Alpha test group for Crowns. Things were adjusted during that and after.
One thing I do know is that,yes, harder content was added. At least part of it includessentences that had been retired because of the "fail rate". That is, the sentences were hard enough that people failed to answer correctly in much higher number than other sentences. Those retired sentences were brought out of retirement for this feature.
I cannot speak to any other changes though, as they happened after I was a tester and also I am newly returned from a hiatus I took from moderating.
Repetition is good only as long as it helps you get a word or phrase or a rule of grammar into long term memory. After that, repetition is completely useless. The problem is, Duolingo's higher levels include a whole lot of such wasteful repetition, and the result is that we waste time repeating things in the hope that the even higher levels of that lesson will present us with new content. So far, in my experience, they don't. But there's no way to tell that's the case for all the skill blocks.
There is no way to tell that the higher lessons of a skill will not be pure repetition until you complete the entire skill to level 5. People are beginning to realize that Duolingo's upper levels are pure repetition, but they haven't fully realized it yet, because no one who has completed an entire course to gold has reported on this issue.
Duolingo starts off good - you learn a lot. But after the first couple of levels, the system lures people into a procedure that seems to be a complete waste of time.
And sure, if I were fully aware that the upper skill levels were completely useless (as I strongly suspect they all are) I could just give up after the second or third level. But I'm not 100% sure. So I continue, hoping I'm wrong. But if the 4th and 5th levels are completely useless (as they absolutely are for at least some skills), surely that needs to change.
Repetition is important for learning. It's the way brain defines that the information is "important" and keeps it in the long term memory. But this repetition must be "spaced in time", with intervals between the repetitions. This is the most effective way to learn a new language. So, there is no point in trying to level up in a single topic very fast. Go forward and regularly going back is much better.
Repetition is important, but not mindless repetition. If something is already in one's long term memory, it doesn't make sense to contine repeating it ad absurdum. Having to "study" lessons that one has already internalized for 750 days is not a sensible use of one's time, and it is silly that Duolingo asks us to do so.
I agree. It's surely certain that, by level 4, all the vocab, phrases and grammar in the course is already in one's long term memory. Personally, I think they need to reduce the number of repetition levels and instead give us some more useful vocabulary and phrases to learn. Of course, the problem is, that takes a lot more work.
It's surely certain that, by level 4, all the vocab, phrases and grammar in the course is already in one's long term memory.
I disagree with this. That's only going to happen if you've done it over a long period of time. Blitzing a skill through to level 4 in an hour or two won't help much. Especially not it's a small skill with only 15 lessons at level 4. Which is why the strength meter was so important.
Although perhaps that's what you meant ;-)
duolingo is not asking you to do so. you get the opportunity to rehearse as much as you want. after the first crown you can continue with the following lessons. Personally, I also work on multiple levels. for me it does not all have to be at level 5. And duolingo does not require you to do so.
"duolingo is not asking you to do so."
But it is! The problem is, if we want to finish the course, we have to do so. Are you seriously suggesting that the makers of Duolingo put in all those grinding levels expecting us to just ignore them? Surely not. When you build any kind of a system, it's supposed to work. This system works fine at first, but it starts to break down the more you use it.
This is not how an effective language course should be. Duolingo is supposed to offer an effective way of learning a language by turning it into a game. But the game becomes an interminable grind for no gain in language skill at the upper levels. Even if you try doing multiple levels, eventually, you're going to find that you're left with a bunch of lessons, all of which require you to grind out lessons with the same content over and over and over again.
This is not the user's fault for being inflexible. It is poor design.
"Having to "study" lessons that one has already internalized for 750 days is not a sensible use of one's time, and it is silly that Duolingo asks us to do so."
They give you the option to, but they don't ask you to. You don't need to go up to level 5 on everything. A better use of your time is probably go through everything once up to level 2- use the practice between lessons... and then back through everything to level 3 once everything is level 2.
Regularly use the "practice" in between leveling. There really is no reason to go above level 3 on anything, unless you feel the need to for a particular subject.
You don't need to go up to level 5 on everything. A better use of your time is probably go through everything once up to level 2- use the practice between lessons... and then back through everything to level 3 once everything is level 2.
Well, I like the timed practice as it forces me to think and react quickly, like in real life situations. However, now the timed practice is enabled only after you have earned all five crowns for the skill. So, I do need to go up to level 5.
(Couldn't reply directly as there are too many nested comments already.)
@40ShadesOfGreen, In the web you can.
That's why I had specified the app only.
This is another matter with the Duolingo, that annoys me a lot... The differences between web and mobile...
True, I especially miss the Tips and Notes in the app and the Clubs on the web.
"They give you the option to, but they don't ask you to. You don't need to go up to level 5 on everything... There really is no reason to go above level 3 on anything..."
But that's not an argument supporting the system - it's basically an admission that the higher levels are useless, which is precisely what we're saying.
Telling us that we should accept Duolingo as it is or create work-arounds within the lower levels to overcome Duolingo's flaws is not an answer. We aren't the problem here - Duolingo is. If Duolingo is to be an effective tool for learning languages, IT needs to change. Telling us that we need to accept things as they are is basically to surrender and admit that Duolingo isn't going to be able to do the one job it's supposedly built to do - i.e. use gamification to teach languages as effectively as possible. Currently, Duolingo's upper levels are like an old-style MMO grindfest. It's poor game design, which means it's also poor learning software design.
Something needs to change, and it's not the users.
it's basically an admission that the higher levels are useless, which is precisely what we're saying.
this is the sentence that I and many others responded to:
"and it is silly that Duolingo asks us to do so."
And it is not useles for some of us. you can choose. see my comment above
As I was re-reading my previous response (the one I made earlier today, not the one you responded to here), I realized that Duolingo DOES ask us to, and edited my previous answer to reflect this.
Let's not forget, you're on level 12 in your highest skill language. I only started to notice the problem at this point. If you notice, most of the folks criticizing Duolingo are in the 20s, while most of the folks defending Duolingo are around our level or below. I suggest to you that perhaps you and I haven't experienced the full effect of the problem yet.
The thing is, you can do work-arounds to overcome flaws in any system (and that's what the system's defenders are telling us to do), but in this particular system, that will only work for so long. I suspect that when we get to level 20, the system's inherent problems will overwhelm any work-arounds we put in place.
For some people the higher levels ARE useless... for others they may be useful.
I can think of one situation where they proved useful. Me- and Spanish. I started Spanish when I started Esperanto, but felt I needed to just concentrate on Esperanto. When I came back to Spanish, my basics skill was a three- but I had forgotten a lot.
For me, having a level 4 was useful to get my feet wet and ready to start learning Spanish properly this time.
They give you the option to, but they don't ask you to.
The system is a teaching aid; it's supposed to nudge you in the right direction. It's supposed to have some idea what you should be learning. Saying that you should ignore what it's suggesting you need to do is, as IanBrettCooper suggested, an admission that the system is flawed.
It was fun the old way- getting a full golden tree. There was something rewarding about being all golden and keeping a tree golden.
I think some people are missing that; it's not actually necessary to do- and I suspect it's more the people who have been here longer who are probably more upset by the change.
Newer people will probably feel less compulsion to be all gold because they won't have experienced the old system.
Good discussion! I find this fascinating and super odd that many people have such a strong opinion, when after looking at some individual levels many people who don’t like Duo are in two different camps: They either couldn’t have completed through Five Crowns as their level is too low (since it’s such a grind and takes so much time, but then how do you know the advanced content sucks?) or they are level 25 of a bunch of languages (ummm, it seems to be working for you - why are you hating so much?)
It makes me want to weigh in here with some opinions (and maybe a few facts) 1. Duolingo is going for gamification to help learn languages 2. Learning generally requires repetition. 3. Some people learn more quickly than others 4. Duo is not perfect 5. But, Duo is free 6. Crowns are here 7. These forums are a good learning tool (and entertainment) 8. I find a tremendous amount of value in the native speaking moderators and discussions on “why” a sentence is what it is - not all programs provide this. 9. Duo is free (oh, that’s number 5) - sorry
I remember playing the “Dragon’s Lair” video game as a kid. I would go through the same six or so levels and lose every time. There goes a few more quarters. It took me dozens of times to get the timing just right on the correct move that would allow me to advance to the next scene. That meant going over the first five scenes over and over and over. Eventually I made it to the next level and again, it took me quarter after quarter to try to get through the next scene. Eventually, I gave up. It became boring to get to the same place without feeling any additional sense of accomplishment. My goal there was to conquer the game.
Just about all of life is based on gamification: -Facebook = How many likes do you have on your posts? -Foursquare = Check in here! - tell the world that you know this place the best and you’ll get a message telling you how cool you are. -Many jobs = complete these steps and you get a raise -Almost 3M apps and games available on your phone
One reason there are so many game apps on your phone is that people start to feel the need for that immediately gratification. If they haven’t “accomplished” something they move onto the next game which will immediately give them a badge of honor for completing level one. Wow, that felt great, I’m going to keep going to the next level! My sense of accomplishment is learning something, not measuring how many skills are level five. (I do admit, I kinda look at my XP and levels against others and the questions they ask)
So, you have to ask yourself: Are you here to beat the game, or are you here to learn?
If your goal is to beat the game, break out the quarters because it’s going to take a while.
If your goal is to learn something, then learn will playing the game. Have fun with this, people. Enjoy the experience. It’s better than sitting in a class doing “ich bin, du bist, er/sie/es ist” and did I mention that it’s free? It is free, you know.
If you’ve learned all you need to learn, then great - go on your trip to your country of choice or move or speak with your friends. I have not learned all I need to. I am at level 2 through the whole German tree and fives on the first couple sections and 3/4 in the middle. Sure “Der Mann isst einen Apfel” is drilled into my skull due to what felt like countless repetitions, but that’s what works for me. That said, however, “Danach ist er zur Universität zurückgegangen” still takes a bit of thought.
You all are probably smarter than I. I need more work to feel comfortable. Sure, I can read the sentences and translate like mad. My own little game now is to close my eyes: Hear the words, don’t see them and try to translate for what makes sense and what doesn’t, then look.
Lest you think I’m being a mole for the owl or an apologist there are a number of things that could be improved: -Additional content - especially when you get to level four and five - I would like to be challenged. Though I have no idea if the level four/five content at the advanced levels will feel as repetitious as it did at the first levels. I do know the first levels well, so bring it on! -Advanced AI that could calculate based on accuracy and speed the number of reps an individual may need in a particular subject. -Lingots - Are they really worth anything? If they are, feel free to give me some. -Speech clarity and speech recognition could be improved (In addition to taking DE from EN I’m taking the EN course from DE and there are times when I don’t recognize my own language nor does it recognize me) -Duo’s odd sentences happen so often that they are not representative of the real world. So many times I have to look up the discussion with the question “what did they mean by that?” But the stories are an okay compliment for “real world” speaking. -Duo needs some correction on additional correct translations and needs to take away some incorrect ones that have slipped through.
So, have you completed through level five on all skills, do you have super high streaks and high levels on a bunch of languages? I’m in the middle - I am neither, therefore, I’m giving this a chance. I’m still learning a little each day and that’s my goal. I’m also surrounding myself with German Netflix, Music, news, etc.
Overall, I am happy with my free, self-paced, self-scheduled study here.
Duolingo is not "free". It has costs. For example...
Internet use isn't free, unless you're accessing the internet at your local library. You need some kind of computer; you need access to the web. These things aren't free.
Duolingo costs us in time spent. Time is money, and the time we spend on things is important. Duolingo costs us a lot in these terms, because when it's teaching us stuff we already know, it wastes a lot of our time - time that would be better spent actually learning.
It is a huge amount of work and I find that If you go straight through it, it will feel even more repetitious and boring. I feel that going back and forth between hard and easy lessons is much more satisfying. When I feel like I'm getting stuck or am tired, I take a few minutes and blast through a couple easy lessons to pick my spirits up. And when I'm getting bored, I go straight back to the "harder" lessons. Easier levels are fives and I'm through at least two on the whole tree. My sense of order is completely messed up from the half circles and different levels all over the place but I'm not sure I feel a sense of order with most of German. :-)
It does get kinda ridiculous. If they must keep that number of reps, I'd be much happier if they had the same number of skills, and the same total repetitions, but break them up into more levels, so we have maybe ten levels. That way, it would feel as if we were getting somewhere. With the time spent on a single segment of a skill getting longer as you get towards the bottom of a tree, it's getting to the point where I'm dreading the higher levels.
If it seems pointless, it probably is. It's better to find another way to learn the language or another way to use Duolingo.
I'm currently learning Korean and I take advantage of the crown system by first using the word bank for three levels and moving on to a new skill every time I get the first crown in a previous skill. I have three or four skills on different crown levels between 0 and 2. When I get a skill up to crown level three, I leave it and come back later and start typing my answers. This way I can focus on understanding the language first and later as I review I have the added challenge of typing, so the repetition doesn't seem too boring . I'll keep doing it this way until it starts feeling too easy and then I'll think of another way (or until Duolingo changes and I have to find a new strategy).
Usagiboy has often said that the only constant thing about Duolingo is change, so as a user I choose to adjust my learning strategies to fit what Duolingo offers me rather than feel like Duolingo is forcing me to do something.
It might be fun if there was more vocabulary in the lessons, so if you work your way through all the levels you're rewarded with learning more words, and more words unlocked in your practice. If someone wants to repeat the vocab they already have, they can still choose to do that by practising the levels they already have.