In science, as among movie stars, it is so often about Meeeee! Back in the early 90s, I was working in a minority-interest field of science. I was plugging away at the human genome, the aspergillus genome, the yeast genome, the candida genome, the drosophila genome . . . all before any of those genomes were delivered complete into the public domain. That's where I had two of the three big ideas I've had in science. We had a corner on a particular, peculiar way of looking at genetic sequences and were applying our technology to the genetic inventory of one species after another. The inside joke was that we should write a mail-merge program to write the same paper again and again only changing the name of the species, the tables of data, and a single sentence of the conclusions. The convention in bioscience is that the last author is the PI, the principal investigator, the Gaffer, the one who wrote the grant to land the money for everyone else's paycheck. The first author is often the youngest person in the lab; they are The Effective, the one who has done all the grunt work. As a youngster starting out , the number of first-author papers is what gets you interviewed for a permanent position. Anyone else involved in the project is shovelled into the middle ground. For us there were rarely more than 3 or 4 authors to be listed.
In biomedical science, later on in my career, it was rather different. I remember having a one-side-heated conversation with a post-graduate in the ophthalmic genetics lab one floor down. He was raging because he had been placed 5th / 8 in the billing for the lab's latest paper. He was convinced he should be 4th! I was amazed that someone would a) care b) be so precise in the algebra. The other thing that used to annoy me when I worked in a hospital setting was that the consultant surgeon and often his registrar would get their names on the scientific papers when all they had contributed was some very delicate butchery to provide the samples which the real scientists had analysed. I said at the time that you should only get your name on a paper if you could present it at a scientific conference if/when The Effective fell sick at the last moment. You can do this if you've written a chunk of the text, or you're the boss, of you've heard & seen the results thrashed out at numerous lab-meetings. The surgeons are far to busy to attend lab meetings in the research centre so they often haven't a clue how their abstracted tissue is analysed.
Sometimes, the paper results from the collision of two different trains of research. Then you have two young turks, of similar seniority, who have both put their all into the project for the last several months. The solution to that is "joint first author". I know at least one case where, the most ambitious and, let's face it, ruthless of the two has gotten his [almost always a he] name physically first among the joint-firsts. Papers are usually cited as Smith, J. et al. (20xx) so it does make a difference in external perception. So it's very much a case of "Eclipse first, the rest nowhere". In my own case, we were on a collision path when my old boss, who had moved to Nottingham, and my current boss in Dublin found that they were both analysing the same pair of organisms in essentially the same way. My oppo in Nottingham was much younger than me, just starting out and a woman-in-science. Of course, I let her take first-first place in the list of authors. This helps explain why I am a nobody at The Institute while the other actors here cited are Rulers of Empires at home or abroad.
get top billing on the promotional material. Even if it means the order of names bears zero relationship to the mugshots in the back-ground photo. Actors have agents to fight their corner on this, because the order means money. If two monster stars are neither of them backing down, you can stagger the names so that one is leftest and the other is toppest [R] and let their agents fight over how many pixels higher or lefter their client finishes up.