Card scores seem off

[koreth]

I'm trying to figure out why some of my imported cards from PD1 have much higher scores than others. Here are two examples that seem wrong.

1. 拿 - created 1/29/06, difficulty 100, record 24-0, score 31590
2. 个 - created 10/20/05, difficulty 100, record 27-0, score 20900

The difficulty being 100, I can understand (PD1 had no notion of difficulty, just a history of right/wrong answers, so unless the converter assumed a particular aggressiveness level and answer quality, it'd have no way of computing a correct difficulty.) But why is the score on 拿 50% higher than the score on 个, which was reviewed more times with no incorrect answers?

The card with the highest score is an oddball too: 物理学家, created 9/30/06, difficulty 108, record 29-2, score 49421. I've hit that one since upgrading to PD2, and haven't hit the other two yet, so maybe the high score here is just a result of PD2's score adjustments getting applied here and not on the others.

 

[mikelove]

The scores are actually based on card ranks; if you were using frequency-adjusted testing in 1.0 they'll be calculated to correspond to the frequencies with which cards appeared, if you were using repetition-spaced the score for each card will be 100 points times the number of days in that card's repetition interval.

So it seems like 拿 was in a higher rank than 个 (or you had ranks customized in a funny way where cards with a higher rank would appear more often than cards with a lower one) - if you check those cards in 1.0 does the rank on 拿 seem to be higher? Do you remember changing the way cards advanced in rank at some point? Perhaps 拿 was reviewed after you'd adopted a more aggressive rank-advancing system.