nope , the term is based on the worst case scenario's which were if you retired in the years i listed . if you are testing any later time frames than you are not comparing apples to apples .
it is not just picking time frames out of a hat that you think are examples . any guesses at what gold did pre 1975 is just that , a guess and no one knows for sure what would have happened if it traded freely as it does today . sorry i don't accept tylers 1972 comparisons, we had that discussion when he was claiming the same thing for the permanent portfolio .
Nowhere in the original trinity study was the SWR ("SAFEMAX") defined as "the rate at which you could deplete your portfolio if you retired on years x, y, z, or w". It was defined as (my wording as I can't find the paper back) "the maximum rate one could use to safely withdraw from his portfolio without depleting it in a given timeframe, no matter what the starting date". The fact that the worse years (those which defined SAFEMAX) were the ones you listed is a result of the study, not a definition.
As for the apples to apples, and thanks to the above definition, nothing forbids you to calculate SWR on smaller timeframes if that's all you've got, to compare different AAs. Now, sure, you cannot compare 1926-present's SWR for a 60/40 with a 1972-present SWR for GB. You have to adapt the timeframes, obviously, and be aware the longer the timeframe, the safest the SAFEMAX.
don't forget a comparison has to take in to consideration the balance's left over too , not just the draw rate .
Once again, you're wrong. The SWR does not care about balance left. But I agree that has to be taken into consideration when looking at the whole picture. The Trinity study is just that, an academic study, and it's been stated many times that nobody does actually that : spend x% (inflation adjusted) of the original value of the portfolio, no matter what, and consider that dying with $0 left is as much a success as dying with millions.