720 to 560 on test day. How did this happen?? My personal theory.

Sorry I posted this subject on the incorrect subsection of the forum, but I meant for it to be in this one as I wanted to help other test takers learn from my mistake.

I'm the last person to make excuses ever, but I am going to own up to this situation and just get better. Full disclosure: I have no idea if my observation is accurate, but if this post can help at least ONE person on this forum, I'll gladly spend the time to write it. I started studying about 4.5 months ago starting with a high 200's diagnostic score, which I would argue is not an accurate indicator because I did not even remember the definition of an integer at the time. I read through the entire MGMAT series which I found to be helpful and after reading through the entire series, I scored a 560 on the MGMAT cat diagnostic. The score was something around the range of low-mid 40's on quant with a low-mid 20's on verbal. I found this score to be reasonable as I had relearned all the fundamentals of standardized testing, but I also understood I needed to work on advance and higher level questions. I studied literally every single day for the past 2.5 months for at least 30 hours a week, disciplining myself after work and spending hours on end throughout the weekends. I also started using GMAT forum to take 40+ questions every day mixing up quant/verbal with all sorts of random subtopics. I found that hammering out questions will NOT improve your score, there are certain concepts that require you to go against the grain and just fundamentally relearn concepts. With this mentality in mind, I watched several MGMAT videos on their website to reprogram my brain to think differently, which I found to be incredibly helpful. Verbal was definitely my weakness so I hammered SC and CR with an iron fist.

Fast forward 2.5 months later, I decided to use up my 2 free GMAT Prep CATs as my test date was approaching in 2 weeks. To my pleasant surprise, I scored a 750 on the GMAT prep diagnostic with the scores of a 50Q/42V - to be fair this score may not be entirely accurate as I saw about 3 questions I have seen before on previous GMAT forums although I wasn't sure about the answer to the question. Fast forward again to 4 days prior to my actual test, I decided to use my free Veritas CAT as I did not want to waste my last GMAT prep; although I found their quant to be pretty accurate, the verbal completely threw me off. I scored a 51Q/33V giving me a 690; I was incredibly disappointed in this score as I was tilting my focus to verbal and thought I had made incredible strides. I freaked out about the accuracy of this score, took a quick power nap, and then proceeded to take my second GMAT Prep exam (yes, on the same day). Even though I felt pretty drained from the earlier CAT in the day, I managed to score a 720 (50Q/39V) which I found to be more representative as I did not recognize any of the questions as one that I had seen before. This score was more accepting as I felt it measured my accuracy. To be fair, I never practiced with IR/AWA on these practice exams.

I took the official GMAT this past week and on test day, everything appeared to be incredibly normal. I got about 7 hours of sleep, arrived at the test center early, ate a banana during the break. There was some jackhammer and construction noise in the building across the room, but I didn't find that to be too bothersome. In fact, I thought I wrote a stellar AWA argument and did well on IR - scored a 7. During the quant section, I saw a lot of questions that I did not find familiar, but upon carefully reading the question, I was able to figure out the correct answers or so I thought. After coming back from my break, I was waiting for the proctor to check me back in, but there was some guy arguing with her leading to a 2-3 minute delay of me signing back onto the exam. Needless to say I wasn't happy, but I managed to get the timing back on track.

So what happened? and what's my theory? I have no idea what happened to quant as I thought the questions were pretty fair, but I did notice after I selected my first answer, if I used my mouse to click another part of the computer screen (out of personal bad habit), my answers would change, but I thought I had caught that hiccup early on in the exam. Maybe because of this, I got the first a few on the first 10 questions in quant incorrect? I saw a few difficult 700+quant problems, which I solved in under a minute as they have appeared on similar GMAT preps before with slight twists of course. Does the software assume that I'm a 600 level tester and IF I do manage to solve 700+ questions in a short amount of time, they assume I just got lucky? My first 1-7 verbal questions in verbal were pretty challenging, they threw some of the harder 700+ SC and CR questions at me rather quickly in addition to an incredibly lengthy science passage, which I also assumed to be a 700+ type of passage. This science passage was probably harder than any other passage I have seen before. Please keep this in mind, I usually do pretty good on RC - whereas I struggle in SC/CR the most. Anyways, toward the middle of the exam, I started noticing that they gave me some super easy SC questions and I even managed to finish the exam early, which has never happened before in my life. I push the complete button and my score showed 560 with a breakdown of 42Q/25V. My mouth dropped wide open and I ended up canceling my score at the last minute; looking back, maybe I should have kept the score so that the improvement would look more impressive as it shows AdComs that I studied my butt off?

Can any GMAT experts or test takers PLEASE explain to me what happened? Maybe by me randomly clicking the computer screen, some of my answers got changed without me noticing? Some would probably argue that since I did not practice with AWA/IR, the practice scores would not be accurate since I would be drained, but I can reassure everyone, I was not drained toward the end of the exam at all. I'm someone that can take two back to back CATs and still function. Although I have always refused to believe in the "first 10 questions matter the most algorithm theory", I'm starting to believe that it's entirely true. Or maybe it's the fact that even if you get some 600 level questions wrong in the beginning, they just drop you to the 500 level leading to a very slim chance for recovery? If any experts can please explain this LARGE discrepancy, I would really appreciate this. To anyone reading this, I highly recommend they spend extra time on the first 10 questions and be extra careful without sacrificing enough time to adversely affect your later sections of the exam.