Math Whiz Breaks Calculation Record

keyshawn632 writes "The Associated Press reports that Gert Mittring, 38, needed only 11.8 seconds to calculate the 13th root of a 100-digit number in his head at a math museum in Giessen, a small town, located in western Germany. It's worth noting though that his feat will not be recognized by The Guinness Book Of World Records because of the difficulty of standardizing such mathematical challenges."

38, ohhh

[photon_chac]

According to Neumann's thoery, a math guy reaches his peak at about 26, could this _Gert Mittring_ be a bit more 'number-crunching' at that age?

Re:What?

[ricotest]

The dude can memorize a 22 digit number in four seconds (according to the article) so I'm sure he can take a similar time to juggle the numbers around in his head. Perhaps his mental algorithm focuses on certain numbers at a time so that he can handle it.

Devi: another brilliant mathematical mind

[GreenPenInc]

When I was a kid, my dad lent me a book of Shakuntala Devi's book, "Figuring". She was famous some years ago (in the 50s, I believe) for her own computational ability, multiplying two 13-digit numbers in her head in 28 seconds.

The book itself was an interesting read, and at the time I just ate it up. It has a lot of tricks regarding number theory, mathematical riddles, calendar tricks, and calculation of pi, for example. It teaches how to figure the day of the week for any Gregorian date of any time in a few seconds, a trick which I still remember and use today!

As for the Pi, it contained a few poems and sayings whose letter counts signified the individual digits. I started trying to memorize pi, with my sights set firmly on the world record (as I am not without my own mathematical and mnemonic prowess). However, around grade 9, I decided to abandon my quest in order to get a life. I had memorized 1350 digits at that point.

One such quote held little significance for me at the time, but has since become hilarious. "How I want a drink, alcoholic of course, after the heavy chapters involving quantum mechanics!" Needless to say, my quantum prof found it quite funny. :)

I so call bullshit

[tomstdenis]

Unless there is some really trivial algorithm for finding 13th roots I totally call bullshit. If it takes him four seconds to memorize a 22 digit number how can he manipulate and find a 13th root for a 100 digit number in just over twice that amount of time?

There has to be a trick to it aside from "thinking really fast"

Tom

The future is here

[forgetmenot]

When I hear about people like this I can't help but think of "Dune" and it's Mentats.

I would like to know how much of this ability is genetically determined and how much is due to training and from what age did his "gifts" become apparent.

Either he needs to be stuck into some kinda breeding program (perhaps solving his virginity problem *hyuk hyuk*) or his training regimen needs to be studied and duplicated en masse. Imagine an advanced state-of-the-art military computer system that runs on 3-square meals a day and isn't susceptible to EMP bursts.

Re:Devi: another brilliant mathematical mind

[Skeezix]

I'm not sure which method you use, but I included one method in an article I wrote on memory improvement [pubcrawler.org] which some slashdot readers might find interesting.

With some practice, you really can get to the point where you can calculate days of the week for any date in just a few seconds. People don't realize it's not all that difficult so it's a nice parlor trick.

Also included in that article are methods for remembering long-digit numbers, the order of a deck of cards, etc.

dumb tricks...

[Vellmont]

It sounds impressive, but how usefull is doing something a machine can already do more quickly and efficiently? John Henry [wikipedia.org] learned this the hard way. As others have pointed out there's tricks and shortcuts that people use to doing these calculations, so most of it just amounts to mathematical parlor tricks.

The implicaton is this guy is a genius. Maybe he is, but calculating roots quickly doesn't make you a genius, it just means you know some math tricks. Isn't this just the mathematical equivalent of how many peanuts can you stuff up your nose?

Wizard...

[fourharpoon]

Apparently, most /.-ers don't value this feat highly. Well it makes sense, these geeks don't love computer for no reason :-)

But IMHO, it's a great gift, nonetheless. Heck, I can't even remember my girlfriend's birthday.


~ Sig is not parsed by modder

Increasing math ability

[_Hellfire_]

This guy appears to have "superhuman" math ability, and I would imagine that it's just the way this guy's brain is wired that allows him to do that.

I always wonder if there is a condition that works in the opposite way, a bit like dyslexia for reading/writing for maths, a sort of "mathlexia" if you will. Just as dyslexia doesn't mean you're stupid, it's just that your particular model of brain doesn't comprehend words straight away, a person with "mathlexia" can't add up 137 and 48 in their head to save their life, let alone do anything complicated like division or factorisation.

If there is such a thing as mathlexia, I'd say I've definately got it. The funny thing is, I love computers, I love programming (in C among other languages, though a mastering of assembly has persistently eluded my efforts), and I can understand even engineering diagrams and other geeky stuff. I kicked ass in English Literature at high school (even though I didn't particularly enjoy it and it's not where my passions lie); but I cannot do maths in my head if my life depended on it. Even with a calculator I get lost in the process of doing a complicated sum, but I would say I'm at least a half decent programmer. It's not that I have a problem with a logical process, it's the math part that throws me.

Is it just the way my brain is wired? Is there a big secret no-one's telling me that will make this all easy? Am I destined for a life of going "uh huh? righto..." when someone explains a (pure) math concept to me? Or is there some hope for a math dummy like me?

If anyone knows the answer(s) to any of this I would be eternally thankful.

Re:"memorizing 22 random digits in just four secon

[Class Act Dynamo]

Well, since there are only ten digits, I think ordering would have to be relavant. Otherwise, he could just count how many occurences of each digit there are, which would certainly not be quite as impressive.

If I recall...

[gravteck]

I don't remember if this was the same guy I saw on TV. But the guy I saw was performing large multiplications and finding large roots in front of an elementary school class. They later showed doctors or scientists doing brain imaging on him while he solved math problems. What they found was that he was using parts of his brain that most people utilize during visualization (not sure how they were able to separate it from him actually seeing something). He said he visualizes the number in his head and then he can perform various manipulations on them and he can "see" the math work itself out. Obviously some is probably genetic, but he also commented on practicing his methods for 5-7 years. He also appears to not be the only root master [recordholders.org].

To how many significant figures?

[jwise]

And how much about the problem did he know in advance? Did he know it would be a 13th root of a 100-digit number? Did he know that the number would be a perfect 13th power of an integer? I find it impossible to believe he calculated a 13th root of a 100-digit number in 11.8 seconds without knowing any of these things. Knowing all of them makes the problem a lot easier.

The 13th root of a 100-digit number will always have 7 digits. If you memorize the first few digits of the 13th powers of numbers between 49 and 58 and you are given a 100-digit number, then you immediately know the first 2 digits of the 13th root. Memorize the initial digits of 13th power of numbers between 491 and 588 and you immediately know the first 3 digits. By memorizing the terminal digits of 13th powers of numbers less than 100, you could similarly immediately get the last 3 digits. That leaves 1 digit to compute, which is a slightly less impressive-sounding feat for 11.8 seconds. It's not a trivial calculation, though, and not at all shabby for 11.8 seconds.

Jonathan

Roomie in College

[AlexTheBeast]

I roomed with a guy in college who would calculate a 10 digit by 10 digit multiplication in his head throughout the day on weekends. He would be grilling or watching TV and you would see him get him and write down 1 digit of his answer.

In grade school he had memorized 52 decks of shuffled cards in some insane short period of time. The teacher would ask him what the 12th card of the 17 deck was... and he would start listing them forward and backward from there.

We often went to the casinos with him. He would card count and we just would bet whatever he would bet. We would all make a $100 or so and leave. He was always afraid of getting caught.

Some government agency approached him for running sets of numbers from point a to point b. They liked the fact that he could just put all those digits in his head without a papertrail.

Last I heard of him, he was avoiding math as much as possible... he enrolled in some DO program in a medical school somewhere. Numbers came too easy for this guy... and he knew he would go crazy if he went into a math field.

So now he's a doc somewhere. Probably calculating 10 by 10 digit numbers in his head as he examines you...

Re:"memorizing 22 random digits in just four secon

[DanteLysin]

Reading this article reminds me of a fellow I worked with when I was younger. He could compute mathematical equations such as 56*83 or 123*281 in his head in just a few seconds. But if you asked him, what's 84-21, it would take him forever.

He was autistic and his brain was just "wired" differently from the "norm".

Re:That's easy.

[Rufus88]

because (N mod 10) always equals (N^13 mod 10)

I know I'm going to kick myself for asking this, but why is this necessarily true?

Re:High pi

[Nyh]

I read somewhere that you only need about 50 digits of pi to describe a circle the size of the observable universe to within the diameter of a proton, let alone a chocolate donut.

Well, let us see:
radius universe: about 15e9 lightyears
radius proton: 1.2e-15 m

circle with the size of the universe divided by diameter proton:
2*pi*15e9*365*24*3600*300000000/(2*1.2e-1 5)=3.7e41 .
So 42 digits of pi will do.

42? Where did I see this number before?

Nyh

Re:What?

[henrygb]

The square root involves the first five places after the decimal point of an irrational number, while the thirteenth root results in an integer - this allows some tricks.

Show me any 13th power of an integer and I can immediately tell you the final digit of the root. Similarly with 5th powers and 9th powers. But square roots of non squares don't give so many tricks.

Re:I think you missed

[ArsenneLupin]

the implied "generally".

It only needs to be tested for numbers (N^13) mod 10 = ((N mod 10) ^ 13) mod 10.

And yes, we do have:

0^13 = 0
1^13 = 1
2^13 = 8192
3^13 = 1594323
4^13 = 67108864
5^13 = 1220703125
6^13 = 13060694016
7^13 = 96889010407
8^13 = 549755813888
9^13 = 2541865828329

Ok, so all 10 modulos match (and there are only 10 possible modulos), so how much more general do you want it?

And yes, when asserting a property about supposed to be satisfied by the members of a finite set, a proof by exhaustive enumeration is a perfectly valid proof ;-)

Re:What?

[op00to]

Why don't people teach this in schools? Obviously this kind of stuff is pretty tricky, but there must be other interesting little tricks and relationships that when taught correctly, could have interested a whole lot more kids in math. Where'd you learn this?

Met him last week!

[rxmd]

Actually, I was quite astonished to see this on Slashdot, as I had lunch with the guy last Thursday where I work [caesar.de]. He's nice in persion, but one of the secretaries at work said he stinks and should wash more often. I'm afraid I didn't notice it quite as badly...

He has an interesting way of getting along financially. Basically, he's living off an exclusive contract with the German TV station RTL [www.rtl.de] where he's featured every now and then in shows. He also gives lectures on mathematical topics; RTL makes him charge a very steep EUR 2500 per lecture (about $3000). I think originally he studied psychology; he's still running the psychiatrist's practice in Cologne that he startet off with.

We were joking about him tackling the Millenium Problems now; I wonder if he's serious about that... but then, there's more to it than calculating in your head really fast.

Re:What?

[Com2Kid]

• oh wait, it's not designed at all. it's pretty much all crap. and intelligent design. even in private schools and grad schools, no one ever is taught to really think for themselves as a part of the curriculum.

You know I used to half way believe this, until I got some friends who came over from the Asian schooling system.

Americans are INCREDIBLE at taking story problems and real life scenarios and doing mathematical modeling on them.

Heck just yesterday I saw a Seasame Street game that was basically an introduction to Venn diagrams! That is at the pre-school level, once you actually get into the schooling system, well, here are just a FEW of the things American school children know about that seem common place to us, but are complete mysteries to others!

• The planets, names, sizes, that they even EXIST.
  
• Atoms, molecules, etc.
  
• Anything dealing with biology.
  

(Please note, this list may not be applicable in the southern states.)

The American Educational System needs an overhaul, for sure, our basic mathematical and linguistic education bites, but when it comes to finding creative, or just worldly, solutions to problems, well, we at least have that covered fairly well.