Looks like the Great Firewall or something like it is preventing you from completely loading www.skritter.com because it is hosted on Google App Engine, which is periodically blocked. Try instead our mirror:


This might also be caused by an internet filter, such as SafeEyes. If you have such a filter installed, try adding appspot.com to the list of allowed domains.

Skritter is addictive

Foo Choo Choon   February 23rd, 2010 7:20p.m.

Other users affected as well?

Evidence (30 hours a week, not during vacation): http://img691.imageshack.us/img691/1575/skrittersh.jpg

Luckily you're based in the US. - In the EU, regulators might get interested in the addictive potential of skrittering.

Hobbes828   February 23rd, 2010 8:03p.m.

is that real?

400 characters learned in a week and 30 hours with 95% retention rate?

that is insane....

the good news is that at that rate you will finish all the characters you need in a couple more weeks. the bad news is that your review queue must be quite long every day, but I guess you will really notice it go down as 1000s of characters get pushed back to longer intervals, haha....

murrayjames   February 23rd, 2010 8:17p.m.

Dude, that's crazy. Congrats on the rapid progress though. I do an hour a day, tops.

nick   February 23rd, 2010 8:48p.m.

Wow! You're tops for this week by a good margin, but you haven't hit the record (43 hours in one week) yet.

戴莉絲婷   February 23rd, 2010 10:20p.m.

Wow - those stats make mine look so sad! lol Congrats on the hard work. Now go hang out with some real life people. ;)

shinyspoons   February 23rd, 2010 11:40p.m.

props for the commitment, that some serious skittering. But even if you only spend an hour a day, you can still get pretty good results. Last week i put in 9 hours and learned 400 characters, but my retention rate was lower at %86.

Hobbes828   February 24th, 2010 2:15a.m.


Well that is still much better than I can usually manage.. for example I did 6/7 days for 53 minutes a day if I look somewhat recently on my stats, but I only added about 100 characters.

Besides person to person I guess it varies on where you are in the studying process, etc... I used to add a lot when I first started because 1/2 or more of the characters I recognized and learned to write quickly, plus new words always had new characters...

Now I feel like my lists/queue is about 2 words for every new character, and I'm at 1750 characters or so so most of the really common characters (or ones that I already knew) are gone :)

Well that is all the excuses I have for myself...

百发没中   February 24th, 2010 2:28a.m.

impressive. there really seems to be a clear correlation of how much time you spend on skritter and progress made. you study in one day what I manage in about a week...and in that day you made as much progress as I did in a week:)

keep up the great work^^

阿軒   February 24th, 2010 2:49a.m.

So you guys learn new characters by bulk? You do not associate them to some kind of lesson or context?

I mean i could learn many characters, but I don't know if I would learn their use before long, and by then the time I do I would have forgotten the character imo.

skritterjohan   February 24th, 2010 4:36a.m.

Yeah those stats are crazy. Please let us know how studying that much for a week works out in time. I wonder does your retention and # chars learned drop off dramatically if you go back to studying 1-2hours per day or does it actually level out.

If it does level out it might be that learning bursts of 20-30 hours per week and then just spending a month catching up and then bursting for a week again is really the way to go. If you have a hard time with the characters you learned in that week later on it might not be a good idea after all.

I wonder whether all the learning stats here on skritter could be processed by someone knowledgeable about languages and statistics to formulate a few power rules for efficient skrittering.

Also how did you manage to do 30 hours outside of work? And did you know many of the characters you studied or were they totally new?

shinyspoons   February 24th, 2010 5:06a.m.

I'm on a diet of nothing but single characters, no words or lessons. For me this is the best way to do it.

Spaced repetition means you wont forget them. Sometimes there are ones you cant recall, but usually they only need to be looked up in a dictionary to be remembered.

By just focusing on the characters, you can develop a much better feel for them - so your can learn more, faster. I dont think I could have learnt that many characters in a week if I had not been learning in this way.

There are a few problems. Having no dialogues or vocab means there can be a lack of context. To get over this I watch a lot of Chinese TV so even though im not 'learning' I get to see the characters used. Also i have wasted some of my time learning characters that arent really important, but I think most of that was down to bad planning on my part. Also it can get a little boring - sometimes you just yurn for a sentence.

But if you can put up with it its defiantly worth it. Before I started on this binge, I was struggling to make the transfer from elementary to intermediate on chinesepod. Even with lots of time and had work, getting my head around intermediate lessons was proving pretty difficult. Now, the seem so easy and even the advanced lessons arent that difficult to read - although Im sure Im not getting the full meaning.

Foo Choo Choon   February 24th, 2010 6:32a.m.

Some added info:

[1] Just about a month ago, before the Skritter age, I could barely write 50 characters, although I recognised many more.

[2] I threw in some character lists "Chinese Primer (characters)" for fun (most characters known, took a few hours), but the bulk of it is from the HSK lists.
Besides, there's a lot of words I simply like, so I just add them to the queue.
As for me, I don't think it makes much sense to learn single characters if I don't know the corresponding words.

[3] The fact that my learning curve has been rather linear does surprise me. Many of the characters were completely new to me. There should be a diminishing marginal return in the long run. This effect is probably reduced by the fact that knowing a character makes it easier to study similar characters.

[4] Four hours a day probably won't be sustainable in the long, but two hours a day should be manageable.

[5] Most of the time is spent in the evening (e.g. 19h - 01h) , additionally I usually spend about half an hour for morning and noon revision.
Yes, I don't have other hobbies. :)

[6] I recommend listening to Chinese radio to increase learning efficiency.

Shisoik   February 24th, 2010 9:31a.m.

Which Chinese radio are you listening to?

Foo Choo Choon   February 24th, 2010 10:09a.m.

上海网络广播: http://radio.bbtv.cn/radio/live/
- especially 故事广播 (parts of the programme suitable for elementary learners) http://radio.bbtv.cn/radio/live/2009-07-07/649.shtml
- sometimes 东广新闻台

北京广播: http://listen.rbc.cn/
- 教学广播 (depends on the schedule, sometimes quite interesting)

jww1066   February 24th, 2010 10:22a.m.

@穆 That's amazing! I hope you don't develop carpal tunnel.

Here are your reps per item learned per hour and items learned per hour for that one week:

Reps/learned Learned/hr
Words 6.27 58.23
Characters 18.44 48.77
Tones 6.84 51.27

What's interesting about this for me is that your learned/hr numbers are much higher than mine, but your reps/learned are not significantly different, and in the case of characters are actually worse. Here are mine for comparison (since I started over in January):

Reps/learned Learned/hr
Words 8.21 23.59
Characters 14.18 33.00
Tones 9.87 28.47

Notice for example that the ratio of your "reps to words learned" to mine is 6:8 or 0.75, but the ratio of your "words learned/hr" to mine is more than 2:1 instead of the 1.5:1 that would be expected.

For characters there is an even weirder situation: my reps/learned is lower than yours, but you learn more per hour! This can be explained by a different number of reps per hour, i.e. you must have gone through the characters very quickly. Calculating your reps/hr and seconds/rep:

Reps/hr Seconds/rep
Words 364.9 9.9
Characters 899.4 4.0
Tones 350.8 10.3

You are only spending an average of 4 seconds on each character! You spend more time on average on tones and words, and in both cases you have a lower reps/learned rate.

Here are my reps/hr and seconds/rep for comparison:
Reps/hr Seconds/rep
Words 193.8 18.6
Characters 467.9 7.7
Tones 280.9 12.8

I have an Excel spreadsheet I used for these calculations if anyone is interested.


Foo Choo Choon   February 24th, 2010 2:01p.m.

Thanks, James. I'm quite obsessed with statistics indeed. To put it simply, there appears to be a trade-off between speed and memory.
Seconds/rep is slightly confusing because it apparently doesn't show the time you need for a repetition but the interval.

However, just as a "warning", the fact that I primarily study for fun leads to some distorting factors. For instance, I rate some characters as unknown simply because I like them.

Does the fact that I fixed my Target Retention Rate at 97% have any effect?

jww1066   February 24th, 2010 2:18p.m.

I read a paper a while back where they did an experiment with different repetition schedules. Basically they took a fixed amount of study time (40 hours, let's say) and divided it up in different ways; some people studied for 2 weeks at 20 hrs/week, while others studied the same material for 10 weeks at 4 hrs/week, etc. Their conclusion was that spacing out studying was much better for long-term retention.

Now, this is not directly applicable, because we all vary in our motivation levels; if someone can study 10 hours a week on a regular basis, that's clearly better than studying 1 hour a week. However, it does suggest that at some point studying in these huge bursts will suffer from diminishing returns.

I agree that the seconds/rep is confusing; I am not referring to scheduling, but to the number of seconds on average that you spent on each item. So 4 seconds/character means that on average you reviewed 15 characters per minute.

Another issue is that, if you start studying something, then get distracted and go to another page, Skritter doesn't necessarily know that and will continue counting until 30 seconds (or something like that). So there's a fairly large possibility of overestimates when it comes to the total time studied, particularly if (like me) you Skritter while doing something else that often calls you away from Skrittering.

Generally a different target retention rate would make a big difference, but in our case I'm also using 97%.


Foo Choo Choon   February 24th, 2010 3:08p.m.

My estimates, probability calculations (referring to 3 hours a day):

Keeping it up for ...
a week: 70%
two weeks: 50%
a month: 30%
two months: 20%
six months: 15%
a year: 10%

I know my learning characteristics: If I can't keep it up, I'm lost.

But I'm not risk-averse, I've got nothing to lose: I go for the riskiest option.

阿軒   February 24th, 2010 8:47p.m.

I support the method of learning over longer periods of time. Depending on how often i encounter a character or need to use / write it, I will learn that character very easily (which is common sense right?). However I could maybe learn 50 new characters in a day if I wanted to; but I think the next day I would forget how to write (recognizing would be easier) most of them. I need a context to properly learn a character.

Is there a list with all the radicals only?

nick   February 24th, 2010 8:54p.m.
taylor04   February 24th, 2010 9:12p.m.

Calm down and step away from the computer:) Someone start a SA group for this man (Skritters Anonymous)

nick   February 24th, 2010 9:32p.m.

A consequence of spaced repetition principles is that, mathematically speaking, the best treatment for excessive amounts of Skritter is more Skritter.

klutz14159   March 8th, 2010 5:12a.m.

@Nick: ROFLMAO.... oh my god that hurts... so true...

Anyhow, like all good addictions, going cold turkey is always the hardest way to quit.

This forum is now read only. Please go to Skritter Discourse Forum instead to start a new conversation!