Patience, monitoring, knowing what the buckets are (thanks
@niemi24s!), and doing inventory checks up to 8 rooms when things are at high(er) buckets to gauge how they're selling and the likelihood (or not) of an inventory reallocation, has worked consistently for me to get low(er) buckets that work for my travel plans. I don't see it as outwitting, but having something of an understanding of the objectives of yield management and playing the odds using whatever information I can glean. I treat it as a game. I also don't hold out for absolute lowest bucket, but will happily grab a lower one. It also means I cannot book just whenever I feel like it but often have to wait and watch.
With that said, I think that
if there are a
lot of last minute drops and/or rooms going out empty, it would be an indication that the yield management practice has gotten too aggressive and that should drive a revision of their practice to reallocate inventory into lower and middle buckets somewhat earlier than doing last minute drops to lowest bucket. However, some anecdotal evidence here really isn't enough of an indicator that those occurrences have become statistically significant enough to drive such a re-evaluation.
Ironically, I do find these last minute drops to be somewhat heartening. Not that I think they help people wanting lower prices much, they are far to late to plan around, but it shows Amtrak yield management is paying attention and adjusting even if late in the game. It would be far worst if it looked like they just left them at high bucket and ignored it, letting rooms go out empty with no attempt to adjust. This shows they aware of and are reacting to facts, which makes me think that they eventually will make a course correction if this happens enough. I strongly doubt any course correction would include returning to the old practice of routinely allocating some inventory to low bucket 11 months out, though. That ship has irrevocably sailed.