It also has Lara's 153 at No.53 in the all time greatest innings list, Tendulkar's 248 against Bangladesh at No.28 (!), and VVS's 281 is nowhere in the top 100.
So, yeah, a tiny bit flawed I'd say.
While I would like Lara's 153 to be up there, it's hard to rate it high based on a formula that's the same every single innings.. It was a great innings, but Warne was not in good form back then (he was in his mid career). Even then, it's rated an ATG innings and it's actually the lowest score to be in the top 100.
Tendulkar's 248 is automatically discredited because it's vs Bangladesh, but if you look at the scorecard - he had very little support, and batted very well with the tail (Zaheer for the last wicket).. aside from the fact the bowling was average, that innings ticks a lot of the right boxes.
VVS's 281
just missed out on the top 100.. the innings gets a bonus because it's made on a follow on and results in a win, but he got great support from Dravid in that innings.. also even though it seems a great bowling attack, only McGrath was in great form in 2001.
You might note that high scores are automatically given a high rating. While low scores in low team totals are rated higher, if actual runs don't have a significant enough base reward you start seeing names that shouldn't be up there in the ATG career lists (which are compiled by an average of all a player's innings ratings with a longevity bonus).
Good observations, but when you're going by a formula that applies to all innings, it doesn't matter what innings are remembered most, if the scorecard data doesn't allow it to be rated higher it's hard to do much about it. Any suggestions on anything else I should consider is appreciated:
cricrate | Test Batting Ratings