So.

[Boidhre]

I meant normally distributed. I suppose it was erroneous to assume the distribution should be normal, or even symmetric. I am not sure what to make of the KGS data though, as I havent thought about how their ranking algorithm will effect the shape of the curve.

I've seen chess ratings reported as normally distributed.

With KGS, remember that's accounts not players. It's not uncommon for someone to have several accounts and I imagine that it's more common to find an sdk doing it than a ddk for instance.

[daal]

Having said the above, I would also be very interested what were Ed's motives.

I think it's pretty obvious. In another thread a grown man with a wife and small daughter (not huge amounts of free time) who has albeit reached 10k without much effort, had declared his intention, as many have done, to reach shodan within a year. Instead of cheering him on, as would probably be appropriate when responding to an ambitious kid, Ed offered up a dose of cold sauce.

My take on Ed's first post there is that he felt that the aspiring shodan was not aware of the magnitude of his planned task. Ed's post was called demotivational by the aspiring shodan, and perhaps it was. The sentiment to ignore Ed's opinion was subsequently echoed (by you, among others).

Ed then started this thread, calling into mind that our collective experience shows that the road to shodan in a year, let alone pro, is rocky and steep, and regularly underestimated. He showed that his pessimism is well founded, and rather than being based on simple negativity, it was grounded in his knowledge of and, I presume respect, for all of the hard work put in by so many of us who have nonetheless not achieved our stated goals.

[Bantari]

Having said the above, I would also be very interested what were Ed's motives.

I think it's pretty obvious. In another thread a grown man with a wife and small daughter (not huge amounts of free time) who has albeit reached 10k without much effort, had declared his intention, as many have done, to reach shodan within a year. Instead of cheering him on, as would probably be appropriate when responding to an ambitious kid, Ed offered up a dose of cold sauce.

My take on Ed's first post there is that he felt that the aspiring shodan was not aware of the magnitude of his planned task. Ed's post was called demotivational by the aspiring shodan, and perhaps it was. The sentiment to ignore Ed's opinion was subsequently echoed (by you, among others).

To be honest, while I might have agreed with this sentiment and still do, my main point of posting in that thread was a side-line discussion about pro teachers vs. ama teachers. And I have to admit after some more PM discussions with Ed I am leaning towards his opinion on that at the moment.

Ed then started this thread, calling into mind that our collective experience shows that the road to shodan in a year, let alone pro, is rocky and steep, and regularly underestimated. He showed that his pessimism is well founded, and rather than being based on simple negativity, it was grounded in his knowledge of and, I presume respect, for all of the hard work put in by so many of us who have nonetheless not achieved our stated goals.

Guesswork?
I personally know people who set themselves ambitious goals and then went on to achieve them. I suspect each of us does, including Ed. As some people from the very threads Ed brought up seem to have mentioned, their time is not yet up and they seem to be right on track towards these goals. So I will reserve my opinion on that until I hear some more.

[Bonobo]

[..]

My take on Ed's first post there is that he felt that the aspiring shodan was not aware of the magnitude of his planned task. Ed's post was called demotivational by the aspiring shodan, and perhaps it was. The sentiment to ignore Ed's opinion was subsequently echoed (by you, among others).

Ed then started this thread, calling into mind that our collective experience shows that the road to shodan in a year, let alone pro, is rocky and steep, and regularly underestimated. He showed that his pessimism is well founded, and rather than being based on simple negativity, it was grounded in his knowledge of and, I presume respect, for all of the hard work put in by so many of us who have nonetheless not achieved our stated goals.

Thank you, this is exactly how I, too, thought it is.

Ed wouldn't have had to do that much work to put anybody down if he had wanted this—and I’ve never read anything here where he’d put somebody down deliberately, rather the opposite, supporting others and constructively criticizing people’s play.

[Solomon]

I think some people would step back and take a better look at the rationality of their goals if they made a bet of some sort, e.g., if I don't reach 5d in 3 months then I will go to work dressed up as Sai, purple hair included.

[jts]

I think some people would step back and take a better look at the rationality of their goals if they made a bet of some sort, e.g., if I don't reach 5d in 3 months then I will go to work dressed up as Sai, purple hair included.

There's a bit of a commitment problem inherent in this idea, though. If you can't force yourself to achieve the goal you set, how are you going to force yourself to go through with the penalty you set?

It might work better as a cash bet among two or more people with improvement goals; then you can set aside the wager in advance and distribute it to the winners of the bets at pre-arranged checkpoints. But that wouldn't prevent all of the wagerers from admitting defeat and unwinding the bet.

[Solomon]

It might work better as a cash bet among two or more people with improvement goals; then you can set aside the wager in advance and distribute it to the winners of the bets at pre-arranged checkpoints. But that wouldn't prevent all of the wagerers from admitting defeat and unwinding the bet.

Yes, I agree. I took part in a cash bet to lose weight with a group on dietbet.com and found it much easier to commit to my goal of losing 4% of my weight putting down the money than if I were to just force myself to commit mentally.

[Splatted]

I think some people would step back and take a better look at the rationality of their goals if they made a bet of some sort, e.g., if I don't reach 5d in 3 months then I will go to work dressed up as Sai, purple hair included.

Sounnds like a win win.

[deja]

I want to hear what Ed thinks.

Perhaps, but the topic itself (ambiguity and all) has generated a lively discussion. I sort of want to see where it goes without Ed clearing things up for us.

I think where this thread goes is pretty obvious because it seems to already have gone there.

Normal distributions and standard deviations? I couldn't have predicted that... In any event, I've found the discussion interesting and haven't seen any finger pointing or ridicule, except perhaps toward Ed for his presumed motivations.

What interests me is the broader phenomenon of people turning a hobby, usually a sport of some kind, into a professional aspiration. The expectation is almost always the same - if I just work hard enough at it, I can achieve it - and the outcome is almost always the same, it doesn't happen. Hard work simply is not enough for such grand aspirations, but the myth of hard work guaranteeing success persists despite all the evidence to the contrary. I guess that's why myths persist.

Yes, I often see the glass half empty... because it is.

[SoDesuNe]

What interests me is the broader phenomenon of people turning a hobby, usually a sport of some kind, into a professional aspiration. The expectation is almost always the same - if I just work hard enough at it, I can achieve it - and the outcome is almost always the same, it doesn't happen. Hard work simply is not enough for such grand aspirations, but the myth of hard work guaranteeing success persists despite all the evidence to the contrary. I guess that's why myths persist.

I have yet to see someone who really works hard at the goals he posted here (for more than a week). I for sure don't. I simply work at them when I have the time to do so.

In my opinion hard work leads more often than not to success, although it obviously can't guarantee it. No work leads to nothing.

And the glass is half-full! =)

[deja]

What interests me is the broader phenomenon of people turning a hobby, usually a sport of some kind, into a professional aspiration. The expectation is almost always the same - if I just work hard enough at it, I can achieve it - and the outcome is almost always the same, it doesn't happen. Hard work simply is not enough for such grand aspirations, but the myth of hard work guaranteeing success persists despite all the evidence to the contrary. I guess that's why myths persist.

I have yet to see someone who really works hard at the goals he posted here (for more than a week). I for sure don't. I simply work at them when I have the time to do so.

In my opinion hard work leads more often than not to success, although it obviously can't guarantee it. No work leads to nothing.

And this is why I put "such grand aspirations" in bold, which refers to "the broader phenomenon of people turning a hobby, usually a sport of some kind, into a professional aspiration." I never claimed that hard work rarely leads to success.

And the glass is half-full! =)

No it's not!

[Bonobo]

I’ll take the smaller glass then, thanks

[emeraldemon]

Normal distributions and standard deviations? I couldn't have predicted that...

Off-topic statistics rant time!

Every probability distribution has a standard deviation, but standard deviation isn't a particularly intuitive concept to think about. It roughly represents the average distance a particular random sample will be away from the mean value. Here's a test of your intuition: imagine rolling a simple die, with the numbers 1-6 each equally likely. What's the expected value of a single die roll?

3.5

Now what's the standard deviation?

about 1.7078

Instead, I'd recommend a more robust and intuitive measure of variance: the interquartile range (IQR). In the same way you can find a median by sorting the numbers in a list and picking the middle one (or averaging if there are two), you can pick the first and third quartile by sorting the numbers and dividing them into four chunks. The IQR is the gap between the 1st and 3rd quartile. What would you guess is the IQR for our die roll?

2.5

Was that closer to what you would expect? It's a slightly less-well-known statistic, but in most situations I think it's a better descriptor of a distribution, telling you roughly where the "middle half" falls. The median is usually also better than the mean, but that's another rant...

[Solomon]

Instead, I'd recommend a more robust and intuitive measure of variance: the interquartile range (IQR). In the same way you can find a median by sorting the numbers in a list and picking the middle one (or averaging if there are two), you can pick the first and third quartile by sorting the numbers and dividing them into four chunks. The IQR is the gap between the 1st and 3rd quartile. What would you guess is the IQR for our die roll?

2.5

Was that closer to what you would expect? It's a slightly less-well-known statistic, but in most situations I think it's a better descriptor of a distribution, telling you roughly where the "middle half" falls. The median is usually also better than the mean, but that's another rant...

I don't think it's intuitive at all, as there is no standard definition on how to compute a quartile, some methods better than others depending on what you're trying to do. For more info: http://bit.ly/YMjmFW

[drmwc]

Normal distributions and standard deviations? I couldn't have predicted that...

Off-topic statistics rant time!

Every probability distribution has a standard deviation, but standard deviation isn't a particularly intuitive concept to think about. It roughly represents the average distance a particular random sample will be away from the mean value. Here's a test of your intuition: imagine rolling a simple die, with the numbers 1-6 each equally likely. What's the expected value of a single die roll?

Minor nittery: The Cauchy distribution doesn't have a mean or standard deviation (or any finite moments). The integrals are undefined.

I don't know if anyone has done research on the size of the set of distributions on R which do not have finite moments. I suspect it's a reasonably big set.