Right, that’s basically what I was asking here. What is X? Technically, yes, the 4% rule does account for a drop right out of the starting gate, but if you are looking to minimize the risk of that, what amount of time makes sense? I’m sure there’s no scientific answer to this question but I’m still interested in how others approach it.
You're right, there's not really a scientific answer. Basically the lower your WR on your actual quit date, the better your odds.
1987 started as a booming year for the stock market. If you had hit your FIRE number in early to mid 1987, you would have experienced a SORR-relevant event only 6-10 months later. Had you not already retired, this event would have probably persuaded you to work another ~2 years until your portfolio recovered back to the FIRE number. But had you already retired, your portfolio would have eventually recovered and you'd be fine. These outcomes represent a tradeoff of 2 years' labor in exchange for a lower WR that probably didn't matter in the end, but at the time it would seem like a "can I retire or not" question.
2000 was different. It was a bad time to retire on a stock portfolio, because the S&P500 was destined to lose 40+% and not recover to its old peaks for seven years, and then the 2007 peak ended with another -40+% recession that took until 2013 to recover from. Basically, a Y2K retiree would have spent 13 years of their retirement deeply underwater, with a WR in the 5-10% range. Their 30-year and 40-year odds might not be so good after all that depletion. There was only a brief window of time in 2000 when this retiree could have worked 3 or 6 extra months to ensure the recent gains stuck, and then made a probably-correct decision to keep working. In other words, the odds of a "let the gains stick X months" strategy saving a Y2K retiree were slim to none. Only a conservative AA could have saved them from a lot more working years.
As these examples illustrate, the "make sure the gains stick" strategy might or might not persuade you to work OMY whether you need to or not. Also these extreme examples are somewhat cherry-picked. Most of the time, stocks will be higher 3-6 months in the future, and one will have wasted 3-6 months working unnecessarily thanks to the strategy. These much-more-likely small mistakes are the flip side of the occasional risk of making a big "retired at the wrong time" mistake. If you've made the right AA and WR decisions, you'll be fine the vast majority of the time.