Hacker News Reader

Comment by tutfbhuf | Hacker News Reader

Congratulations that's a great achievement. I think that almost everyone should be able to achieve a solid 2000 elo rating, with enough dedication over the years. Doing tactics puzzles is one of the most important factors for fast improvements.

> Some day, I want to become a titled player. A candidate master needs an ELO of 2200

2000 and 2200 are completely different skill levels. I have some friends who play chess for more than 10 years in a chess club, starting as a kid, but haven't reached titled skill level yet and probably never will.

My personal advice is: focus on the fun part of chess, there will always be a stronger player, but that doesn't matter.

View on HN

> I think that almost everyone should be able to achieve a solid 2000 elo rating, with enough dedication over the years

I went to check rating to prove how unlikely this is. Sure enough, that would put you just outside the top 100 players in Australia.

But then I noticed the guy who played one board above me at school when we played is rated above 2000, and he wasn't really much better than me.

So yeah maybe it is possible!

I think consistency play a good factor in it.

ignorant question time: isn't ELO relative not absolute? if everyone gets better doesn't that just make it a lot harder to get to 2k?

It's relative to the pool of individuals, and there are tons of caveats (as other posters mentioned).

Despite this, due to the size of the pools of individuals and relative seriousness of players who belong to official tournament organisations, the ratings have been fairly stable. Yes, there's been some inflation, but I'd say it's somewhat been in line with the increase in knowledge and training methods. While the system has all kinds of minor issues, it's still "quite good", especially when we compare it to other sports or activities with a more variable pool and quality.

In this sense, most people know what 1600, 1800, 2000, etc, mean to fairly precise degree.

Lots of new players that join up, drop their ELO points into the "bag" and then leave (or if online, sign up new accounts). ELO "points" aren't a finite resource.

That would quickly lead to hyperinflation of ratings. The rating lost by a new player is not won by an old player. If you join and is assigned a rating of 1900, but loose your 5 first games, you probably loose a 100 rating points each game. However, the player beating you at 1000 rating doesn't get 100 points, just a few points.

It's generally understood that online chess ratings are inflated over what the FIDE/table rating is.

The first "1900" was meant to be "1000" as well. Cannot longer edit.

there is inflation (deflation too) and all kinds of fraudulent schemes.

https://en.wikipedia.org/wiki/Elo_rating_system#Ratings_infl...

It is said that if a child (teenager) follows the step course, which is a Dutch youth training method, you should be able to start from then on at a 1600 ELO rating. I would assume following the course is self-selecting, you only make it if it is fun and you are good at it, so take that for what it is worth.

> I think that almost everyone should be able to achieve a solid 2000 elo rating, with enough dedication over the years

I think you are massively underestimating the natural range of human intelligence.

I don't think the average IQ 100 individual could ever reach an ELO of 2000 regardless of how much they practice, and half the population has IQs of less than 100.

https://news.ycombinator.com/item?id=23888814

> GM Hikaru Nakamura famously took the Mensa online IQ test and scored a very average 102

His elo peak rating is 2816. Just one example for an extraordinary strong player with average IQ (there are many other examples). I don't think that a high IQ is required to become a decent chess player (like 2000). It may help though.

While I think people underestimate how much hard work can overcome low cognitive talent, such that probably a lot of people with average iq can hit a chess rating of 2,000 if that was their primary goal in life, I also think Hikaru's IQ probably isn't actually only 102. Online tests aren't so rigorous. Just from hearing him talk he seems to have the usual nuanced understanding of the universe on a variety of domains that would not be typical of an average iq. he also took the chest while streaming and talking to chat

I think that the value of IQ is overestimated. IQ tests don't capture "nuanced understanding of the universe on a variety of domains". They measure something much more narrow, i.e., how good you are at the particular type of IQ tests.

> GM Hikaru Nakamura famously took the Mensa online IQ test and scored a very average 102

Nice clickbait. It you watched him take it on stream, you'd know that he was goofing around with his chat and ran out of time b/c he didn't know it was timed.

Fair enough, let me try again:

https://chess.stackexchange.com/a/24946

> The most comprehensive and recent publication I've found related to this question is The relationship between cognitive ability and chess skill: A comprehensive meta-analysis.

> The bottom line is that according to a meta-analysis of numerous studies, there is clearly a correlation between chess skill and cognitive ability, but cognitive ability explains only about 6% of the variance in chess skill (12% if you only consider numerical ability only). There must be other factors as well (e.g., training), some of which could be more important than cognitive ability.

From my personal observations I very much agree with 6%.

IQ test is bullshit and has been debunked a number of times.

Nowadays anyone bragging about their IQ should be considered a moron and left talking to themselves.

For example: https://www.youtube.com/watch?v=W3oUqKUx2o0 (first thing that came off my head)

IQ tests are universally accepted in psychology as the most robust psychometric ever developed. The idea that IQ tests are "bullshit" is nonsense pop psychology. Some people might try to over-interpret IQ tests as meaning more than what they do, but they are very good at measuring what they measure, and they predict a lot of things.

That said, I agree if you brag about your IQ you're a moron.

a test in isolation cannot be bullshit. when people talk about iq test many people imagine the person believes an iq test is a rigorous measure of all intellectual talent. or course it is not that. but it is also wrong to suggest it has no meaning at all. it does have some predictive power.

imperfect measures are not always useless, which is goof since all measures have some error.

> a test in isolation cannot be bullshit.

A test trying to measure something we can't even properly define is kinda flawed in it's whole premise.

As such the best an IQ test can do is measure how well somebody prepared for a test like that, but it will struggle to properly catch any knowledge or skill domains not covered by the test, of which there are a lot.

IQ tests correlate with things. Thus it captures more than just how good someone is at the IQ test.

You can say "There is this test, and people's scores on this test correlate with various things like, income, marital fidelity, religiosity, incarceration rates, job performance". We call this test the IQ test.

IQ tests don't even attempt to measure any knowledge or skill domains at all. I don't think you understand what an IQ test even measures.

If anything, top level chess players are likely to have missed some non-chess education when they were young due to being a chess prodigy, which would contribute to a lower IQ in adulthood.

I'm pretty sure IQ tests aren't supposed to work that way. At least theoretically, they're measuring intelligence, not education.

1) They are a measure of a very specific kind of intelligence: the kind of intelligence required to score high on IQ tests. I score pretty high on IQ tests but I can say without a shadow of a doubt that my intelligence in other areas is, at least as it appears to me, below average.

2) intelligence does have some elasticity. When you think of intelligence has a culmination of pathways/heuristics/low-level-techniques, it’s easy to see how many pathways are self-reinforcing and not easily taught or measured. You can be predisposed to a particular kind of intelligence, and not everything is attainable by everyone, but what we call intelligence is close to what we call talent: seems like an innate quality, but actually reflects long-term hard practice with countless hours dedicated.

Intelligence is a complex interplay of nature and nurture. There is no way to test only the "nature" component, it's akin to unmixing a pot of purple paint to extract the red and blue.

A lot of IQ test questions get easier the more mathematics you have studied.

that is the goal but at the extremes i imagine it fails. imagine you are a chess prodigy that was homeschooled and never had a sit down exam before. don’t know if hikaru fits the description.

IQ tests just measure your ability to complete IQ tests.

Garry Kasparov is smart but we're not talking Feynman levels of intelligence

The correlation between chess ability and IQ is positive but not that high. not as high as it is with physics and math abiltiy.

Chess is mainly about memory and recognizing patterns and implementing those patterns at that exact moment at the chessboard.

So I would definitely think there is correlation between chess and memory. There was a really good documentary about memory called How Does Your Memory Work. Highly recommended, you get to see a glimpse of the workings of Susan Polgar's brain. I might also confuse it with My Brilliant Brain from National Geographic :)

I remember seeing a live post-mortem analysis by Anish Giri. It was just astonishing how fast he can see combinations, yet also how deep those combinations are. And then on top of that throw in a sentence like "Oh, it is not new, it was played in 2006 as well, in that game between him and him". So really, fast thinking, slow thinking, and picking deep into your memory, all at once, seemingly without any effort. That is 2760 ELO for you :)

That's because physics and maths ability is included in the IQ test. If you included lots of chess or chess-like problems than Kasparov would magically rank a lot higher.

Funny that you bring up Feynman, he stated he only had an IQ of 125. High, but probably lower than Kasparov.

Don't you mean overestimating?

I think the parent comment does mean underestimating the _range_, but overestimating the IQ, as you point out.

Once again HN readers revealing how sheltered they are in their little bubbles.

If you pay attention to it you'll see how far removed from reality a lot of the readers here are. It's quite worrying really.

Can you elaborate instead of just dropping a sharp comment?

I'm curious about what you mean.

Sure. I see it almost every time I visit HN. I've encountered many people with a slightly warped sense of reality but none more so that on this website.

There are two broad patterns I often see:

The first is where they think everyone on the planet is somewhat like them. You'll see comments like "everyone should be able to program" or even "everyone should program". I've even seen people claim that ability to program should be considered a fundamental life skill. In general it's hugely overestimating not only the IQ but the interests and mindset of the general public. The comment I replied to here is an example of this type.

The other one is where their perception of a baseline standard of living is far higher than that of average people even within their own country. It's usually that they think some luxury item is absolutely essential despite people around the world living healthy and happy lives without it. They seem to be completely unaware that these other people exist.

My theory is that these people have been fed through schools in affluent areas, through university and straight into ridiculously high paying jobs and have simply never had (and probably never will have) any contact with people from other walks of life.

On the flip side you have people like me, I grew up working class poor in the north of England in the 1980's - worked minimum wage jobs in my late teens/early twenties and managed through hard work and luck to escape that life into a successful career as a software engineer.

While I'm sure a decent subset of users would fit your description, that's certainly not the case for every user.

I agree with what you're saying but it seems like you might have replied to the wrong comment.

The parent comment is making the same point as you about IQ: "I don't think the average IQ 100 individual could ever reach an ELO of 2000 regardless of how much they practice, and half the population has IQs of less than 100."

Reply on news.ycombinator.com