Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Forget Me Not - How to win the U.S. memory championship. (slate.com)
36 points by senthil_rajasek on May 14, 2011 | hide | past | favorite | 23 comments


FYI, Ed's my co-founder at Memrise, and you can read his blog here:

http://edcooke.memrise.com


Wow! I really liked Moonwalking with Einstein and thought Ed sounded awesome.

He seemed like a great hustler in the book so I bet he is a fun co-founder to be around.


Hah! Yep, never a dull day. You can get a glimmer of things from here:

http://www.flickr.com/photos/memrise/3966430619/in/photostre...

and facebook.com/memrise has a few photos as well :)

If there's any interest, I can try and persuade him to do an 'I am a Grandmaster of Memory, AMA' on Reddit.


I'd like to see that!


The main reason that American times in memory events has improved over the past few years (besides more people spending more time and effort) is that we have started using the more advanced techniques of the Europeans. We have gone from one two-digit image per location to the 3 two-digit numbers per location PAO method of Dominic O'Brien to the 4 two-digits per location of Ben Pridmore.


Really interesting. Can you give some reference links about such methods. I am currently learning Chinese and trying to learn and remember 2200 Chinese Characters.


There's a great method for learning the characters in the context of Japanese, described in a book called "Remembering the Kanji" by James Heisig. It helped me work through about 1000 characters in a month; doing it full-time, you could probably to recognize 2,000 in a month, and read them in another month. He's published a similar book for learning the characters in Chinese; you should check it out.

Taking a quick look at the memrise website, I don't think their technique is going to be as helpful as Heisig's. For a variety of reasons (discussed in the book), memorizing the character for "man" by associating it with a picture of a man is not efficient for learning more than the most basic characters.


Thanks, "He's published a similar book for learning the characters in Chinese" - Do you know the name?

Edit: I think I found it: http://en.wikipedia.org/wiki/Remembering_the_Hanzi


Yes, that's the one. I haven't used it myself, but I expect it's good.


> memorizing the character for "man" by associating it with a picture of a man is not efficient for learning more than the most basic characters.

You're dead right - most Chinese characters are combinatorial, so it would be crazy to try and come up with an image for each. We combine a number of different techniques, including weaving the components into stories. See e.g. the first mem for 'meat' here:

http://www.memrise.com/item/14684/meat/

I don't know if this gets to the heart of your concern... I'm just trying to highlight that the simple animation from 'man' to an image of a man is just one of the approaches we employ.

Thanks for taking a look.


One of the great things about Heisig's approach is that the student makes their own images and stories for each character. Something that I invent is much more likely to stick with me than something I see online; and once you get used to it, the strategy only takes a few minutes, probably on par with the amount of time it would take to read through that page.

That page presents a lot of information, and most of it isn't ideal for memorizing the character. It's nice to see images and etymologies while you're looking at the character, and they might even allow you to recognize it when you see it later, but it's especially difficult to recall the exact form of the character later from information such as this.

I would suggest you take a look at one of Heisig's books if you're interested in chinese characters specifically. It's a very well thought out system, and he spends some time explaining its specific advantages in the beginning of volume 1 of Remembering the Kanji.


Apologies for shilling, but give us a try - this is exactly the problem we're trying to help people with:

http://blog.memrise.com/2011/03/why-chinese-is-getting-easie...


I apologize for any offense, but that post is riddled with things that give me the impression the author himself doesn't really understand Chinese that well. For starters, the vast majority of words are two or more syllables. Even when individual characters are a distinct word, they often have multiple meanings depending on context. There just isn't a simple 1-1 mapping of characters to "meanings". I'm a fan of Heisig's RTK system, which it looks like you borrowed from heavily, but its main utility is building a scaffold to assist with writing before getting the sufficient practice to do it naturally. However using the system for reading is a horrible waste of time since reading things in context is one of the few ways of learning the collocations and general diction patterns necessary to be a functional user of a language.

The bit about only needing 2000 characters is also clearly BS. Even an 11 year old elementary school student will read far more than that many (though the number they're required to write is lower especially in the first few grades.

In another post on the same site, "Why is Mandarin so incredibly easy to learn?", I read the following:

>To be an undergraduate at a Chinese university you need to know about 5,000 words

This is 100% absolutely false. The popular red Oxford E-C/C-E dictionary contains well over a hundred thousand entries, and a typical undergrad will know nearly all of them as well as many other words not in the dictionary. Learning a language is a major undertaking and this kind of blog post is really doing everyone a disservice.

Edit: Personally I've had experience learning over half a dozen languages to various levels, including Chinese. In my former work in Taiwan, I dealt customers almost exclusively in Mandarin for 3+ years. I can't claim to be a language expert, but I am familiar with what it takes to be functional in one. Learning a language is a major task.

For some idea of what's involved, check out this book written by one of the most successful polyglots of the modern era: http://news.ycombinator.com/item?id=2538379


How many people have learned Chinese using you method?


We're still young, so we've only had a few thousand people try us out for Chinese. Of those, a good number have learned hundreds of characters or more.

One caveat - until recently, we've only really tackled the problem of reading characters. We've just started to introduce approaches for learning the pinyin and pronunciation, but they're not as polished yet.

It's going to be a little while before we tackle writing - I'd try Skritter for that.


Awesome site! Any idea how long until you get official support for Korean? I saw some cool user generated stuff, but it's not useful without audio pronunciation.


I use memrise.com for French and it incorporates these techniques in language learning. Greg Detre and Ed Cooke have a great site there.

For general info about the memory methods I recommend Mnemotechnics.org


Moonwalking with Einstein (the book by the article's author) is a really interesting read. I recommend anyone who liked the article to check it out.

There are some interesting ideas related to Joshua's 'deliberate practice' that can be applied to software engineering as well.

http://www.amazon.com/Moonwalking-Einstein-Science-Rememberi...


Here's an excellent blog on some of these techniques and training by Nelson Dellis, the current US Memory champ: http://climbformemory.com/blog/

His older posts go into a lot of interesting detail.


Are competitions for these fair game for anyone? I would think there are people[0] who naturally have Rainman-like recall abilities which occurs subconsciously. The rest of us would have to employ an ever-expanding bag of tricks just to try and keep up.

[0] http://en.wikipedia.org/wiki/Kim_Peek


Joshua Foer, author of "Moonwalking with Einstein", actually met Kim Peek and makes a lot of interesting observations about him. Regarding these competitions, what's incredibly interesting is that all the memory champs have always been trained... people with naturally strong memories have never been able to compete with them. The same holds true for the world mental calculation championships too.


This previous Hacker News thread might be helpful for language learning. http://news.ycombinator.com/item?id=1603562


I wonder if contestants use amphetamines or methylphenidate before competitions.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: