Some considerations for thinking about IQ:
(1) From a god's eye view or else from the point of view of the detached archobjectivist, humans are more similar than different when it comes to most things, including intelligence. For example, even William James Sidis, with his alleged IQ of 300, could not think eleven-dimensionally, for he too was reducible to an earthbound mind that only differed in the quantity of information obtainable and processing speed rather than anything qualitatively distinct.
(2) Since there was nothing qualitatively distinct, one can effectively propose that Sidis is to average intellect as Camaro is to Neon or as body builder is to average build. Each case entails more of the same stuff rather than a qualitatively distinct departure. With cars and bodies this fact is immediately observable, yet with intelligence the tendency is for people to thoughtlessly become mystics and ascribe to it some magical quality rather than applying the same standards we apply to everything else.
(3) At any rate, IQ tests are designed to test a person's ability to solve abstract puzzles quickly. Solving abstract puzzles requires an ability to reason. Correct reasoning is the business of logic and philosophy. The master of reason is therefore above all a logician and philosopher. It's quite likely then that a logican-philosopher will have a high IQ, since solving IQ puzzles is one form of reasoning; but having a high IQ does not necessitate the ability to think philosophically or to master reason, for their are other systems of reasoning not addressed by IQ tests, but which a person might have a knack for that places the one in question in a superior position to other humans.
(4) From this follows the perfectly logical possibility for one to have an IQ of 190 but be unable to construct arguments from scratch with the kind of sysematicness, precision, and rigor that a logician-philosopher who has mastered reason is able. For instance, it came as a suprise to one of the members of my university's chess team when, at the Pan American Games in Texas that we competed in, we got into a conversation where I put forth a carefully crafted argument for why traffic congestion misallocates scarce societal resources and how certain strategies could conduce to a more effective optimization of road space during peak hours. He called me a genius at the dinner table; now, despite being indifferent to this claim (due to the implications of an inflated or deflated sense of intellectual worth that may follow one way or the other and prevent me from treating myself as objectively as prior to the comment), what I found most interesting was that he was surprised at my argument, as if only someone who could see one move deeper on the board could be worthy of higher intelligence. The pattern was repeated with some of the other members, including a student who's doing a PhD in mathematics who I beat in the qualifier but who still ended up on the team, who I had an in-depth conversation with on the plane ride home and was likely suprised at the level of depth I computed all things philosophical, purely from raw intellect and intensive thought. The point implicated is that, IQ tests as they are lack the resources to measure a person's logico-philosophical capacity, but that is precisely what makes a person more or less intelligent for, as Descartes has noted, we are above all a thinking thing. Therefore, a person who can think better with less effort is more effective; therefore, more intelligent. For the brilliant person, there may be a lag between the one and the other; for the genius, it is accomplished effortlessly. Since some things still require effort to compute, I cannot effectively be called genius by my own definition; therefore, this is the other reason why I let the prior gentlemen's claim pass in silence, since he necessarily has a different working definition than I, but I prefer mine to his.
(5) It seems logical that if a person sees only four moves deep, he should design a puzzle that is only three moves deep or less; for if one finds oneself seeing X moves deep, one can freeze the position in mind and search for a deeper move. Since, however, there are always deeper moves, it is dubious to establish a puzzle that requires the same level of depth as yourself to solve; for there is no small chance that the test-taker will see that one move deeper, which may invalidate the proposed answer. On that note, if a person sees 300 moves deep, to test their full intelligence the test should be designed by one (or a computer) who sees at least 301 moves deep. Now, statistics demonstrates that gains in precision follow from closeness to the mean; the further away, the less precise. For example, a test that is designed to measure with precision those who can see 3 moves deep shall be ineffective in measuring a person's ability to see 300 moves deep. Further, IQ tests are typically timed, which means that value is placed on solving puzzles in a timely fashion. As a result, depth is traded off for time. So, if one can see 2 moves deep in 1 minute, this is usually preferred to one who sees 16 moves deep in 32 minutes, even if one who sees faster will not with all the time in the world be able to focus hard enough to see anything beyond 10 moves. It is not difficult to understand why this is the case. Whether capitalist or communist, no sensible agent in a modern state wants to administer IQ tests that will place a higher value on those who see deep enough to reject the system or commands from those who tow the line. The mindless person, therefore, who sees three moves deep but quickly will on average be more lucrative militarily than the neo-Socratic thinker whose depth in reasoning will prompt him to disagree with commands on logical grounds. In short, despite the obvious observation that academia should be detached from politics and the market in order for scholars to pursue research as objectively as possible, it is often the case that, due to funding and various other devices, research is pursued in a way that reflects these economic and political interests. In short, between the lack of ability of test-designers to etablish tests that consider deep patterns with precision, and economic and political interests that usually entail a preference for assigning higher mental values to those with speed over depth, it follows that much of this demonstrates why one can perform well on an IQ test without being a master of reason--of raw intellect, the animation of intelligence.
(6) If chess is any guide, where visuo-spatial pattern recognition is necessary for success though not sufficient, I have beaten 2300s and lost to 1700s. This is perfectly comparable to IQ tests. If a person is hungry and irritable--perhaps by the ticking of a clock or watch, it is quite likely that a person will not be able to attain peak concentration. If a person is not able to attain peak concentration, then the outcome of the test will not be an accurate indication of one's ability. Now, since IQ is supposed to measure potential it is concerned with peak and not average. What are the odds that you will take the test on a day where your intellect is keenest and sharpest as ever? Certainly, I have noticed that some days my ability to reason dispassionately and lucidly is more effortless than others. Therefore, adding to the imprecision the further away from the mean one is are a host of considerations that may lead one to test at less than peak capacity. Let another example suffice. Some musicians do not like sight reading in front of others, likely because mistakes will be made and thus it will inevitably be done imperfectly. Others do not like when someone is watching them read something for their reaction; for the want of excitation demanded by a feeler to establish connection may stifle one's ability to focus in a sensible and logically coherent way. It follows that if a person installed a camera in the room of such a person sight reading alone or if some device was used to make transparent what the thinker was thinking as he thought about the passage, the outcome would be quite different. For these reasons plus the aformentioned ones, the off-the-shelf approach to IQ testing may not be effectively tailored to a person who requires these conditions to test optimally. On that count, it is also interesting to discover under what conditions a person is most intellectually ripe so that one can replicate these conditions for activities that require peak concentration. For example, during the LSAT I was bothered by little tiny watches ticking at different frequencies that were extruciatingly loud and lacked consonce and were out of synch with my internal rhythm and clockwork. Knowing this, it would seem that ear plugs is a good bet, except that those too distract me in a way tantamount to putting a sea shell to one's ear: there is still noise though in a closed rather than open sense. Therefore, the desirable conditions would be the freedom of empty ears without the ticking of watches and everything else, but this was not possible in the test. In short, testing environment and conditions will almost certainly alter performance. As a self-contained system, I know the precise specifications that are required to attain optimal performance, but test conditions will inevitably run counter to some people's specifications, which means once again that IQ testing will be inaccurate.
(7) In the end, IQ tests are redundant; for do you need to ask a seven foot tall basketball player what size his shoe is to determine if he has big feet relative to the rest of humans? In short, it is quite likely that those who are the masters of reason know who else is and from this one can put humans in quantitative perspective with 0 being the mathematical entity denoting a completely feckless person who cannot reason and 100 being a perfectly logico-philosophical person. Those who necessarily need the number are inefficient and such inefficiency is antithetical to intelligence as previously defined; therefore, they do not deserve to know the number since knowing this prevents them from engaging in the sort of sober thinking that will give them an approximation of that number, which is all the number is anyways given all the aformentioned imprecisions in testing.
(8) Still, there will inevitably be some of you who mindlessly romanticize IQ and will labor like little beavers to improve your scores. This is not a task I recommend, for it is treating the symptoms rather than cause of a deeper issue rooted in psychological incoherence, however should this endeavor be pursued then it should at least be pursued effectively. Pursuing this effectively will include the following: (a) Increasing the total volume of intellectual activity and output. (b) Philosophy, logic, math, chess, music, and economics all help discipline thought, which forces the mind to concentrate and think sharply, which can conduce to a keener ability to recognize patterns and solve problems. However, I suspect that most of the top percentiles already do these things by default for the appreciation of the thing itself rather than a means to an end. So, in principle it seems undesirable to have a flood of people into these areas for the purpose of increasing a number so they can wear it fashionably as they do a pair of earings. Such people will invariably bastardize philosophy and cheapen it for the seasoned ones who really put in the time, dedication, and deep thinking. On the other hand, if a little more logic and philosophy means a less thoughtless populace then in that sense it may be considered desirable. At any rate, my default position is to sit on a fence when there is a fence to sit on; for, to firmly be on either side is to be on the wrong side for taking any side stifles one's ability to search deep into the costs and benefits without staking an interest that will undermine the point of view of objectivity. Assign only the direct weight of what was said and nothing more, and do with it only what you will and nothing less.