Amtrak will lose this passenger/revenue with a stub.
They are going from 30 bedrooms and 84 roomettes per week between SAS and NOL to 1,036 coach seats per week.
NARP figures say the average sleeper fare on the Sunset is $260. Assuming sold out trains, the tri-weekly (ie current) Sunset has 114 first class tickets available, that yields $29,640 in potential sleeping car revenue.
The average coach fare on the Sunset Limited is $74. Assuming sold out trains, the proposed daily service from NOL-SAS has 1,036 coach tickets available, for a yield of $76,664 in potential coach revenue. That doesn't take into account the business class seating, which will obviously be priced above coach.
And I think I know roughly what Amtrak's methodology was regarding the 100k increase in ridership when the Sunset/Eagle combination happens.
There are approx 34,009,022 people within a 25 mile radius of the Texas Eagle.
There were 245,900 passengers on the Texas Eagle in '08.
That gives us a constant of .007% of the service area population that rides the Texas Eagle.
Now, if we combine the service areas of the Sunset Limited and Texas Eagle, we get a 25 mile radius figure of 59,381,367 people.
Using the same constant of .007% of the service population riding the train, we get 415,670 riders on the combined Sunset/Eagle.
However, we need to deduct the people between SAS and NOL. The easiest way to do this is to simply deduct the entire ridership number of the Sunset Limited, which is 70,168. 415,670-70,168 = 345,502.
Recall that the Eagle's ridership in 2008 was 245,900.
That comes out pretty darn close to a net gain of 100,000 passengers.