## BEST GUESSING STRATEGY (ALL SECTIONS)

Pragmatism

Posts: 68
Joined: Thu Jan 11, 2018 10:42 am
Points: 68

So, upon taking numerous test and realizing that the limited time heavily conflicts with my accuracy to completed questions, I have decided to take the guessing strategy a bit further. A strategy that reveals, on average, a 30% chance of getting the guesses right rather than the statistical average of 27% per section at best. I have taken 12 test tests and have got an average of 13 questions right solely on guessing this way (mentioned below):

* Cautionary note: it is very imperative to know, if you have answered numerous questions with great accuracy and you intuitively feel that you have exhausted the particular selected choice mentioned below for the question you are on, then it is best to default to the most recent statistical trend (MRST), which will also be mentioned below.

MRST: B (the historical trend leans towards D)
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
C D C B B C B B D B A B E D B B E
RC
Screen Shot 2019-03-16 at 3.37.44 PM.png (136.86 KiB) Viewed 147 times

Logical Reasoning:
MRST: E (the historical trend leans towards D)
15 16 17 18 19 20 21 22 23 24 25 26
B B D E D D E E E E D E
LR
Screen Shot 2019-03-16 at 3.53.18 PM.png (122.97 KiB) Viewed 147 times

Analytical Reasoning:
MRST: A (the historical trend leans towards B)
17 18 19 20 21 22 23
B D C B A A C
LG
Screen Shot 2019-03-16 at 3.39.01 PM.png (74.36 KiB) Viewed 147 times
Robert Carroll
PowerScore Staff

Posts: 500
Joined: Fri Dec 06, 2013 7:18 am
Points: 436

Pragmatism,

I want to stress first what I think are the features of your recommendations that jibe with our recommendations: choose a single answer choice letter and guess that same letter for all the questions where you're purely guessing. I think you're suggesting doing that in the "MRST" case. So, if I have almost no time to do a certain number of questions, I should pick a letter (A-E) and answer every question left USING THAT SAME LETTER.

So, imagine I have 30 seconds left. I've unfortunately not been able to answer questions 23-27 in RC yet. I should pick a letter and put that same letter down for each question. You'd pick answer choice (B), if I'm interpreting what you say correctly. So:

23 B
24 B
25 B
26 B
27 B

This is fine. It's important not to switch letters - you're increasing the variance of your guessing results without increasing the accuracy.

You can adjust this strategy somewhat. If I have enough time to check a couple answers, I might say "I'll guess B for all of these, and then, if I have any time at all left, make sure that B isn't an obviously wrong answer." This might be easier to do in Logic Games, where you could quickly compare your setup or a mini-diagram for a particular question and see whether that provides a quick proof that your guess was wrong. But keep in mind that we're already considering a "pure guess" situation, where you have so little time to complete the remaining questions that you're just trying to make sure nothing is blank.

If you're suggesting that you choose a different guess answer for each question number based on supposed trends, that is advice I absolutely do not agree with. There are two main problems here. First, that strategy requires memorizing a different letter for each question number. That's overcomplicating the guessing process. Second, the trends that you identify are really just statistical noise. You can extrapolate almost anything you want from historical data - what matters is how that extrapolation matches the next test. And there's no good reason to think there will be any consistent match.

At some point, if guessing strategy isn't simple and is involving a complicated and time-consuming process, I wonder why that extra time isn't devoted to figuring out another answer! And then guessing on one fewer question. This is a surer course of action to achieving a higher score than a byzantine guessing algorithm.

Robert Carroll
Pragmatism

Posts: 68
Joined: Thu Jan 11, 2018 10:42 am
Points: 68

Hey Robert,

So, couple things:

1. This is a time consuming process, but a very rewarding process as well. What I have seen, more so recently, the guessing strategy based on the the MRST isn't as promising as one would expect. For example, if according to the MRST the best answer choice is D, but you have selected close to 5-8 D's already, the likelihood of the guesses being another D is extremely limited. Furthermore, there are some questions later on that heavily favor one type of answer choice over another. The example to this is in Logic Reasoning Section with 26 questions which lends nearly a 41% chance that the answer choice E is correct over another.

2. This strategy helps to address the issue of quantity over quality. In other words, helping a test taker to focus on greater accuracy and less on compromising accuracy for number of questions. I am basing the assertion purely on statistics which has proven that what were the correct answers in earlier tests is no longer helpful in the newer ones. Furthermore, the propensity of getting certain questions right significantly outweigh those that have a marginal chance of being correct.

3. Regarding the timeframe of having 30 seconds left, even 10 seconds for that matter, the process to select these answer choices is done before hand by placing a small dot on the letter of the answer sheet, so that you don't have to think about what answer is associated with that particular question. This has proven significantly helpful in easing me with respects to focusing on the question at hand than worrying about questions that could impede my progress.

4. Finally, before I touch upon the this point, I have to admit I chuckled at the byzantine empire comment . So, getting down to brass tacks, I agree that investing ample time in guessing and less on getting more questions right is counterintuitive,
but this method is as simple as picking answer choice B straight through. Realistically, it will take you maybe 10 seconds more to do the byzantine approach than to simply guess straight through in the end. The only difference between the two is that the byzantine approach requires taking a few seconds to gauge the test taker's ending point per section and place a small dot from questions thereon to ease the guessing part, whereas the simple strategy requires one to guess nearing the end of timing.