~ spaced repetition cannot model learning gains happening outside the system

there was also some other source making this point; as in stop going crazy over algo's, you can't model real life, just use any kind of #spaced-repetition , -that- is the main benefit

some ◊ ACT-R algos try, it seems (❐ Using optimally selected drill practice to train basic facts mentions it and wants to improve in future versions)