I did well on my reviews today and completely smashed the previous record review count, reaching 4284. And that’s even more impressive than it seems, because I also did three new lessons today, so my review count would have been only 4281 if I hadn’t done any lessons.
Incidentally, I don’t think it’s entirely coincidence that I’ve been blowing past one record after another recently. If you look at my graphs above, it may seem like I was basically just treading water for months, but I think one of the keys is that a lot of progress comes in ways that aren’t reflected in the review count graphs.
First off is the fact that almost a quarter of the easier items have advanced to guru or higher now. When I first started, most of the reviews I did were at Apprentice level, so whenever I tried to “sprint” by doing lots of reviews at once, my review count would just bounce back almost immediately as all the items I reviewed re-entered the review pile in hours or days. However, nowadays over half the reviews I do end up at guru+, meaning that they won’t add to the pile again for at least a week. This means that the bounce back is not as severe and takes longer to manifest, which has enabled me to dive much deeper than before.
The second major factor is increased accuracy. While there are some items I’ve missed repeatedly, I sometimes I do manage to remember an item that I missed previously when it comes back up for review. Combined with increased familiarity in general, and my accuracy has gone up a lot over time, which also helps keep my review count down for longer with less effort. Another side effect of this is that when my accuracy is higher, I’m a lot more motivated to do reviews. When I miss a bunch of reviews, I’ll just give up and quit early.
Back during the summer when I was just treading water, I told myself that every little bit helped and I was still slowly making invisible progress as described above. However, it’s one thing to think that and quite another to really see it in action.
On another note, at the end of August, I added a new stat to my nightly stat posts - reviews due in the next week. The idea behind this is the “bounce back” effect mentioned above. Early on, most of my reviews were Apprentice, which meant that the current review count was highly volatile. If I did a bunch of reviews, the review count would plummet, then immediately bounce back up.
The idea was that tracking the review count due in 7 days instead would provide a much less volatile picture of my progress, since it isn’t affected by Apprentice reviews at all. You can see this reflected in the graph above. The red line tends to track the blue line, but it goes up and down by smaller amounts each time, and more smoothly.
However, since over half my reviews lately have been Guru+, the +7 day review count is still relatively volatile, so I decided to add a new stat: Reviews due in the next month. This number can only be decreased by Enlightened and Burn reviews, so it will probably only go up and down very slowly.
Sun Nov 21 2021:
Time spent: 92
Reviews completed: 409
Reviews remaining: 4284
Reviews in next week: 4740 (+456)
Reviews in next month: 5094 (+354)
Accuracy: 75.79% (310/409)
Accuracy on resurrected items: 92.65% (189/204)
Accuracy on non-resurrected items: 59.02% (121/205)
Current item counts: