The Difference Between Scoring 605 And 705 On The GMAT®

What This Episode Covers

If you're scoring around 605 on the GMAT® and trying to break through to 705, the gap isn't about working harder or learning more content. It's about pattern recognition — how often you see something familiar on a new problem and can immediately connect it to an effective approach.

In this episode, Isaac breaks down pattern recognition into four specific components and explains how each one shifts between the 80th percentile (605) and the 99th percentile (705). The difference isn't mysterious. At 605, you're recognizing patterns maybe 30 to 50% of the time. At 705, it's closer to 70% or more. That's a 50% improvement in how often the right strategy comes to mind when you see a new question.

The episode also introduces the TGS two-part error log system — a practical tool for building pattern recognition faster. Most error logs fail because they become a dumping ground for every wrong answer, mixing execution mistakes with strategy gaps. By separating those two types of mistakes into different logs, you can apply the right fix to the right problem and accelerate your score improvement.

Key Points

What Pattern Recognition Actually Means

Pattern recognition gets talked about a lot in GMAT® prep, but it's rarely defined in a way that's actionable. Here's the breakdown:

There are four parts to it. First, how often you recognize something on a new problem — is it happening 30% of the time or 90% of the time? Second, it has to be on problems you've never seen before, not just ones you've already practiced. Third, you can associate that new problem with something you've seen in practice. And fourth — the part most people miss — you've associated that practice problem with an effective approach that actually works for you.

That last piece is critical. Recognizing a problem type is helpful, but it's not enough. If you recognize an overlapping sets question but can't quickly deploy a strategy that works, the recognition doesn't translate into a right answer. That gap between "I've seen this before" and "I can solve this efficiently" is where most score plateaus live.

The 605 vs. 705 Gap in Numbers

At the 605 level (roughly 80th percentile as of this recording), you're probably recognizing patterns on new problems about 30 to 50% of the time. Maybe one out of every three problems feels familiar.

At the 705 level (roughly 99th percentile), that number jumps to 70% or more. You're recognizing something you've practiced on the vast majority of new questions.

That's a 50% improvement in recognition frequency. And it translates directly into how the exam feels — instead of "everything felt unfamiliar" and "I couldn't get into a rhythm," you spend most of the exam connecting dots and deploying strategies you've already tested.

Why Most Error Logs Fail

Most error logs become a wrong-answer repository. They're a list of every question you got wrong, mixed together regardless of why you missed it. And because the list feels unpleasant to revisit — like taking out the trash — most people procrastinate on reviewing it.

The problem is that the list mixes two completely different types of mistakes. A question you missed because you did 2+2=5 is not the same as a question you missed because you didn't have a strategy. Those require different fixes. When they're in the same list, you end up applying the wrong solution to the wrong problem.

The TGS Two-Part Error Log System

The fix is to separate your error log into two lists.

List 1: Questions you knew how to do but got wrong. These are execution errors — you picked the right strategy but misread the problem, made a calculation mistake, or solved for the wrong variable. For this list, log every single time you make the mistake, even if it's the same mistake repeatedly. Then analyze which mistake comes up most often and focus all your study energy on fixing that one habit until you can reliably say you'll never make it again. Then move to the second most common mistake, and so on.

The 15th time you log the same mistake hits differently than the 5th time. Somewhere between the 5th and 10th time, your motivation to fix it shifts from "I'll try not to do that again" to "I need to change this today."

List 2: Questions you didn't know how to do. These are strategy gaps — you either didn't recognize the problem type, associated it with the wrong strategy, or your strategy wasn't fast enough. For this list, re-solve the questions on a regular basis. That repetition is what physically strengthens the neural pathways associated with those problem types. You're not just "reviewing" — you're building the recognition circuit that will fire the next time you see something similar on the exam.

How Many Re-Solves Per Study Session

If you're resistant to re-solving old questions, start with just one per session. If you're open to it but unsure of the value, try three. If you've tried the approach and seen results, five to ten per session is a good target.

The balance between new problems and old problems depends on your situation. If your last practice test showed a lot of questions you should have recognized but didn't, you need more re-solving. If you're seeing topics you've never encountered before, you need more new material. You can self-diagnose this by looking at your most recent practice test and categorizing each missed question.

Other Factors That Might Need Attention

Pattern recognition is the biggest lever, but it's not the only one. Three other things might need work:

If you're missing a lot of questions you know how to do, execution is your bottleneck. All the pattern recognition in the world won't help if you're consistently bumbling the math. You might need to go all in on List 1 for a month.

If you don't have enough effective approaches in your toolkit — you don't actually know the strategies for common question types — you need more content learning. That might mean a study program or a free study plan.

If time management is the issue, that may improve with better pattern recognition, but it could also stem from testing anxiety (seek professional help if needed) or simply not having enough reps under your belt yet.

Key Takeaways

Pattern recognition has four parts: how often, on new problems, associated with practice, and associated with an effective approach. All four need to be in place.
The 605-to-705 gap is mainly about recognition frequency. 30-50% recognition at 605 vs. 70%+ at 705 — a 50% improvement.
Most error logs fail because they mix execution errors with strategy gaps. Those require different fixes.
List 1 (execution errors): log every single occurrence, even repeats. The 15th time you log the same mistake is what creates the motivation to actually change the habit.
List 2 (strategy gaps): re-solve questions regularly. Quality repetition — not just exposure to new problems — is what builds recognition.
Start with 1-3 re-solves per session if you're new to the approach. Scale to 5-10 once you've seen results.
Other bottlenecks may need attention too: execution, content knowledge, time management, or testing anxiety.