The lesser of two evils? Rating loss vs Revenue loss

[Pinstar]

Say you have a small-medium town. You want to plug it into your passenger network. It is too far to properly use road vehicles so you hook up a single train. That single train has way more than enough capacity to handle the town's passenger/mail output.

The dilemma I'm facing is how to do the orders. If I let the train take whatever is on the station and depart immediately, the passenger/mail gets transported as fast as possible leading to a higher payout. However, my station rating takes a hit as there are longer periods where passengers/mail are accumulating but the train is still out making its trip.

On the flip side, I could tell it to "Full load any cargo" where it will sit in the station and slowly fill with passengers/ mail and only depart once one of those is full. The upside is much higher station ratings, but at the cost of a lower payout since most of the train's cargo is "in transit" for a very long time since it's sitting on the train for awhile.

Which of these two, in your opinion, is the lesser of two evils?

[jfs]

The long term value is in growing towns larger, and more frequent service is better value for that.
The actual best value may be to set up a bus transfer service picking up passengers inside the town and dropping off with a transfer order at the train station. That gives you lots of active stations, while also letting your train make larger profits.

[Eddi]

i haven't ever had a passenger line that wasn't immediately overcrowded...

multiple smaller trains usually win out over one big one, because of the rating.

[odisseus]

There is a middle option. You can use a timetable to make the train wait at the station for a fixed time (say, 10 or 20 days). Note that the timetable doesn't have to be complete in order to work. In my experience, this option works better than either waiting for full load or not waiting at all.

[skc]

There is a middle option. You can use a timetable to make the train wait at the station for a fixed time (say, 10 or 20 days). Note that the timetable doesn't have to be complete in order to work. In my experience, this option works better than either waiting for full load or not waiting at all.

I was about to suggest exactly the same thing