Day 100: What Actually Stuck vs What I Forgot

Day 3. Time for some actual data.

After 100 days of my "one problem per day, done deeply" approach, I tested myself. No hints. No looking things up. Pure recall.

The Test

I went through all 100 problems I had solved over the past 100 days.

For each problem, I scored myself:

The Results

Total score: 168 out of 200 (84%)

Breaking it down by topic:

| Topic | Score | % |

|-------|-------|---|

| Arrays & Strings | 36/40 | 90% |

| Two Pointers | 18/20 | 90% |

| Sliding Window | 16/20 | 80% |

| Linked Lists | 14/18 | 78% |

| Trees | 26/32 | 81% |

| Graphs | 20/28 | 71% |

| Dynamic Programming | 24/30 | 80% |

| Other (Heaps, Tries, etc) | 14/20 | 70% |

What I Noticed

**Patterns I understood deeply stuck the most.** Two pointers, arrays, sliding window - these I could apply to new variations easily.

**Patterns I had memorized faded.** Some DP problems I could only solve because I remembered the exact recurrence. When I tried variations, I struggled.

**Graphs were my weakness.** BFS vs DFS decisions, detecting cycles, shortest paths - I needed more practice here.

Compare to My Old Method

Remember when I tested myself on 50 problems after grinding 500 LeetCode problems?

That's a 5x improvement in retention with 80% fewer problems.

The Key Differences

1. **Struggling first** creates memory hooks

2. **Spaced repetition** reinforces at optimal intervals

3. **Pattern focus** transfers to new problems

4. **Teaching/explaining** forces deep understanding

What This Means for DSA 100 Days

I'm building all of this into the product:

The goal isn't to expose you to problems. It's to make you actually LEARN them.

Tomorrow: The pattern that was my breakthrough moment - Two Pointers.

— Marcus