Atlan0001 beat me to responding to the tachyon in a bar scenario. But, I have a somewhat different "take" on the idea of traveling backwards in time:
Going backwards in time would look more like (1) tachyon flies through bar door into arms of bouncer, (2) who places tachyon at bar, (3) where tachyon spits whisky into 5 shot glasses, (4) which somehow jumps up into the mouth of a whisky bottle as the bartender holds it above each glass in sequence, (5) before the tachyon says "Whisky" and the bartended sees to respond with "What'll you have?" Or, should they be calling it "yksihW"?
My point is that we are trying to talk about what a tachyon would "see" as if it is looking at what we see being run backwards. That is not the correct way to look at an interaction with a tachyon from either its frame of reference nor ours.
Clearly, the laws of physics would not change in our frame of reference just because there is a tachyon nearby. So, whisky is not going to jump into a bottle above a glass just because a tachyon ordered it. So, the issue is what would an interaction with a tachyon actually produce in the way of evidence in our frame of reference?
Still using the joke as the scenario, if a tachyon could come into a bar and take down a bottle and pour itself a drink, what those watching (extremely fast) would see would depend on where they are in the bar, and how soon the light gets to them from each place that something observable happens due to the tachyon's presence. Those near the door where the tachyon enters would see the light of the tachyon going by before the light gets to them showing the whisky being poured, while those at the bar would see the whisky getting poured before they see the tachyon come through the bar to get to the bottle. It is the folks near the bar that would see the tachyon appear to leave the bar backwards after pouring the drink. Of course, this all depends on how much faster than the speed of light the tachyon is traveling. So, that tends to make it near impossible for "researchers" to agree on what they have seen if a tachyon makes some observable effects. Without at least some pre-existing knowledge of the tachyon's properties, observations would probably just look like "noise" in the data.