A look at fares on the Crescent

Amtrak Unlimited Discussion Forum

Help Support Amtrak Unlimited Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Status
Not open for further replies.

Tracktwentynine

OBS Chief
Joined
Nov 23, 2010
Messages
588
Location
Washington, DC
A recent thread introduced the idea that traveling further on Amtrak could result in a cheaper roomette price than traveling a shorter distance. This piqued my interest, so I analyzed fares on the Crescent from four origin stations.

To do these analyses, I used AmSnag. It's a terrific tool for analyzing Amtrak fares if you're looking to make pretty charts or just save a few bucks on your next vacation. All of the ticket prices looked at are the low-bucket price only.

Coach fare (dark red line) is pretty straightforward. It's the low-bucket price from the origin to each station down the line.

Roomette fare (green line) is the roomette upcharge added to the coach fare. So essentially, it's the price if one person wanted to be in a roomette singly.

Bedroom fare (blue line) is the bedroom upcharge added to the coach fare. Essentially, it's the price if one person wanted to be in a bedroom alone.

Note: Only major stations are shown on the chart, but the analysis considered the fare from the origin to each other station on the Crescent.

First, let's look at low-bucket fares on the Crescent starting from New York (NYP):

5327875175_e6bb10c2a0.jpg


(larger image)

A few notes:

Generally, fares from New York act as we expect. Note that the lines start at ALX (Alexandria, Va.). This is because passengers cannot alight between New York and Washington, so fare data is not available. Beyond Alexandria, coach fares increase in a shallow step function as far as Anniston, Alabama, after which the fare ceases to increase.

However, what's most interesting is the roomette fare. Note that after ATL (Atlanta, Ga.) the roomette fee drops. That means that someone traveling from New York to Atlanta could buy a roomette ticket from New York to Anniston (and just get off in Atlanta) and save on the order of $50.

Because the roomette fare drops south of Atlanta (ATL), let's look there next:

5327875195_82f62ab3f1.jpg


(larger image)

The above chart should be read in two ways. Southbound trips start at Atlanta (where the lines have a gap) and go to the right. Northbound trips start at Atlanta and go to the left.

Fares for the portion of the Crescent south of Atlanta are much cheaper (in all three fare categories) than they are north of Atlanta. This is due primarily to the demand of the route, which is much stronger north of the Big Peach.

The most interesting piece of data, though, is for roomette and sleeper trips taken from Atlanta to any point south of Washington (including DC). For all of those destinations, the low-bucket bedroom is cheaper than the low-bucket roomette.

So what does Charlotte (CLT) look like?

5328485908_045df727df.jpg


(larger image)

Like Atlanta, the chart should be read starting from the center and moving outward.

Note that roomettes are more expensive in the low-bucket than bedrooms in the low-bucket for trips starting from Charlotte and ending between Atlanta and Washington (inclusive). There is a significant drop in the fare when one compares Charlotte -> Atlanta and Charlotte -> Anniston. Buying a trip from Charlotte to Anniston in a roomette and then just getting off in Atlanta would save a rider $106 over a ticket from Charlotte to Atlanta.

A similar phenomenon happens for destinations north of Washington, although with not as severe a drop.

One final chart, New Orleans (NOL):

5327875227_4b041d54f8.jpg


(larger image)

This chart should be read right to left. Trips start on the right at New Orleans, with longer trips moving further left across the graph.

Note that the lines representing coach, roomette, and bedroom are all roughly parallel from New Orleans to Atlanta. After Atlanta, all three see a steep increase, after which, all of the lines begin to diverge. This is because roomettes increase at a greater rate than coach, and bedrooms increase at a greater rate than roomettes, especially north of Washington.

Anyway, I hope all this is clear. I'll be happy to clarify anything that's not.
 
Great Analysis! I'm glad my confusion in the previous thread piqued someone else's interest enough to put together this impressive data set. That really helps explain how to save $$ on the Crescent, and when to book Bedrooms vs. Roomettes. It also solidifies for me the high demand for the NYP/WAS and ATL route ... maybe if someone at Amtrak saw this it would sway their decision towards another train...? :cool:
 
Just like airfares, who can make heads-or-tails out of the logic behind any of this.

Interesting analysis however.....
 
Last edited by a moderator:
Thanks for the detailed analysis. Here is an actual fare for comparison.

CLP (VA) to NOL on the Crescent this coming June

$917 R/T for two in a bedroom. ($458.50 per person) Also booked 11 months out.
 
Are you in the Business Intelligence field? That's the type of stuff we like to see.

Now if you can just create a simple dashboard and . . .
mosking.gif
Sorry - I started adding work requirements.

Makes me think about doing something similar from MSP.

Tom
 
AWESOME!!! Thank you for all the hard work. I'm going to give this a further look when I've got more time.
hi.gif
 
What makes no sense to me is the dip in roomette prices just north of WAS.
 
OP should be working for Amtrak, so does the creator of AmSnag. AmSnag should be part of the booking process on amtrak.com

It is a shame that Amtrak can't even figure this stuff out and random people have to supplement their inadequacies. It is like bringing duct tape with you on a trip, it shouldn't be needed.
 
I think that as a collaborative project we should undertake to do this for more routes. I just whipped out the Silver Meteor real quick (took maybe 20 minutes to do):

97_NYP_MIA.PNG


I didn't have time to pretty it up like Matt did (I actually have to go do some real work today), but the graph levels off at Jacksonville - once you get to FL, you may as well go all the way (the small increase at the end is a $3 increase that comes into the coach fare at Fort Lauderdale).

I also wonder if the same behavior is evident across all buckets. I would theorize that at the high bucket prices the shape of these graphs would look quite different.
 
OP should be working for Amtrak, so does the creator of AmSnag. AmSnag should be part of the booking process on amtrak.com

It is a shame that Amtrak can't even figure this stuff out and random people have to supplement their inadequacies. It is like bringing duct tape with you on a trip, it shouldn't be needed.
Why? What's in it for Amtrak for them to try and help the public figure out the cheapest way possible to travel? I'd say that both of these efforts work in direct opposition to Amtrak's revenue management folks (in fact, I'd be willing to bet that all these graphs exist, they just don't publicize them for obvious reasons).

I actually wouldn't be surprised if Amtrak tried to shut down AmSnag (not sure if that's possible) to keep this kind of Business Intel more tightly under wraps.
 
I actually wouldn't be surprised if Amtrak tried to shut down AmSnag (not sure if that's possible) to keep this kind of Business Intel more tightly under wraps.
Southwest Airlines was able to shut down a web site a few years ago, (or maybe they changed THEIR OWN web site, not sure) that would automatically check you in, EXACTLY 24 hours before your flight, thereby helping you get one of the coveted "A" boarding passes.

I used it while it worked, it was a gem.
 
OP should be working for Amtrak, so does the creator of AmSnag. AmSnag should be part of the booking process on amtrak.com

It is a shame that Amtrak can't even figure this stuff out and random people have to supplement their inadequacies. It is like bringing duct tape with you on a trip, it shouldn't be needed.
Why? What's in it for Amtrak for them to try and help the public figure out the cheapest way possible to travel? I'd say that both of these efforts work in direct opposition to Amtrak's revenue management folks (in fact, I'd be willing to bet that all these graphs exist, they just don't publicize them for obvious reasons).

I actually wouldn't be surprised if Amtrak tried to shut down AmSnag (not sure if that's possible) to keep this kind of Business Intel more tightly under wraps.
I have to wonder how advanced Amtrak is when it comes to such analysis.

I also have to wonder if Amtrak's strategy is based on fewer passengers at a higher fare, or more passengers at a lesser fare.

When someone publishes this information in a volume for "dummies," please let me know - I'll be the first in line.
 
OP should be working for Amtrak, so does the creator of AmSnag. AmSnag should be part of the booking process on amtrak.com

It is a shame that Amtrak can't even figure this stuff out and random people have to supplement their inadequacies. It is like bringing duct tape with you on a trip, it shouldn't be needed.
Why? What's in it for Amtrak for them to try and help the public figure out the cheapest way possible to travel? I'd say that both of these efforts work in direct opposition to Amtrak's revenue management folks (in fact, I'd be willing to bet that all these graphs exist, they just don't publicize them for obvious reasons).

I actually wouldn't be surprised if Amtrak tried to shut down AmSnag (not sure if that's possible) to keep this kind of Business Intel more tightly under wraps.
I have to wonder how advanced Amtrak is when it comes to such analysis.

I also have to wonder if Amtrak's strategy is based on fewer passengers at a higher fare, or more passengers at a lesser fare.

When someone publishes this information in a volume for "dummies," please let me know - I'll be the first in line.
Amtrak's revenue management system and staff are very sophisticated as far as analysis is concerned and all you have to do is try to book a sleeper from ATL-WAS (even though OBS take up some of the space) and you will see that there very rarely is a low priced sleeper available. Also, looking at the charts tells you that south of ATL, where there is generally space, the prices are lower to attract the business. Their system is not random selection of pricing and/or inventory control. Granted, there generally is an indication of the need for more inventory when the trains are full, -but that discussion has been had on this board for years, i.e. - when do you add more space and how can you guarantee that by adding an additional car you will sell enough to justify the additional costs involved??
 
I am currently between Crescent trips. I rode 19 down from DC (WAS) to Charlotte, NC (CLT) 01/4-5/11 and am heading back north on 20, early this Sunday, 01/09/11. I booked my entire excursion last July 2010. Coming down, bedrooms were a definite must. Leaving Sunday, it's just me, so it didn't really matter. Funny thing was, at the time of making the reservation, no bedrooms had been sold yet, so lowest bucket, and bedrooms were 34 dollars cheaper than roomettes, EACH way. No brainer, I'm in a bedroom, by myself for 34 bucks less. I assume some of the roomettes are taken out of revenue service right off the bat, for the crew, but I could be way off. In any case, I prefer the bedroom! :)

P.S. I am not inferring that the overall cost was cheaper than say, farther south, it was not, but was definitely affordable to me at low bucket.
 
Last edited by a moderator:
OK, I spent a little bit more on my lunch break. I went through and got (hopefully) the high bucket prices.

It seems as though in the high bucket, there's a small step downwards in price once you get to Yemassee - either there aren't any high bucket trips past there or Amtrak is trying to increase the number of through riders (and push people going shorter distances to the Palmetto, presumably).

Hopefully the graph makes sense - the dotted lines are the low bucket fares and the solid lines are high bucket. You can also see that it's possible to get a low bucket bedroom for cheaper than a high bucket roomette on this route. I wonder if that's the case across the Amtrak system.

97_NYP_MIA_HL.PNG
 
A question, if I might: There's a goofy pricing dynamic on the Silvers that I've found, namely that aside from the late-night 67, the fare is usually far cheaper on the SS or SM than on the Regionals from NYP-RVR. As this is becoming a somewhat frequent run for me (visiting NYC with friends is likely to become increasingly frequent as time goes by), the fact that the Silver buckets get stuck low strikes me as a bit of a surprise considering the through nature of that train. Is it just the lack of competing WAS-NYP passengers that's doing this?

P.S. Very nice charts. Ryan, is 728 Savannah? And 977 Jacksonville or Orlando?

P.P.S. The Roomette/Bedroom thing makes sense...Roomettes are more in demand, but they're the natural pick for both a lone traveler and for a couple not wanting to fork over for a bedroom...that was actually one thing I noticed at Christmas: Roomettes were sold out all over the place, but you could often get a bedroom or some variant thereof.
 
Last edited by a moderator:
They're Charleston and JAX. I've done a little of cleanup to try and emulate Matt's style (I hope that you don't mind):
I certainly don't mind.

One note on style (and I can't tell if you did this in the most recent graph), I didn't make a line graph, I made an XY scatter. What that means is that each point (station) on the x-axis is spaced according to its distance from the origin station. If you do a simple line graph, each of the stations are evenly spaced no matter their mileage.
 
Thanks - I was doing a line graph, XY scatter would be a much better way to go.

Edit: Rather than repost yet another graph, I replaced the one above with one based on an XY Plot.
 
Last edited by a moderator:
Well Ryan, I'm no statistician BUT...I'm paying $41 for the H room from NOL to Atlanta Saturday and I've paid less than $50 for the same room from Orlando to Miami. It all comes down to supply and demand. Sometimes we leave NOL with just one or two other rooms occupied but when we pull into Atlanta it looks like Noah's Ark is about to fill up!
 
These charts are neat. Useful way to see where the peak market pricing is for LD trains. Of course, if people going to Atlanta start booking roomettes or bedrooms to stops beyond Atlanta but getting off in Atlanta becomes widespread, Amtrak will have to adjust the pricing to compensate once Amtrak figures out that people are doing this.

The FY2010 numbers for the Crescent and Atlanta are interesting.

Crescent: 298,688 total passengers

Atlanta boarding & alightings: 112,364

Since Crescent is the only train serving Atlanta, that means a full 38% of the Crescent passengers are either traveling to or from Atlanta. Which has got to be close to the largest market share for a city that is not the endpoint of one of Amtrak's LD trains. Shows that Atlanta is an under-served market that could support more Amtrak trains if Georgia and Atlanta political leadership were so inclined.

Another thing to point out about the Crescent, is that it is one of the 5 LD trains that is to be the subject of a Performance Improvement Plan for FY11. The FY10 PIP reports for the Cardinal, CL, CZ, Texas Eagle and Sunset Limited were released last October. The FY11 PIP reports are due for the Silver Star & Meteor, Palmetto, LSL, and the Crescent. So there must be a staff group inside Amtrak that has either started or will get started soon to review options for improving the Crescent. But I guess they are not set up to take public comments or suggestions.
 
So there must be a staff group inside Amtrak that has either started or will get started soon to review options for improving the Crescent. But I guess they are not set up to take public comments or suggestions.
Do you honestly think they would solicit pax comments ??? A long time Excedrin headache has been the NYP Crescent diner crew mixed with the NOL crew running the rest of the train. The diner LSA has complete charge and sometimes shows how big his head can get~ sometimes so big that he has to get off in Washington to keep from knocking the centenary down on the NEC. I'd be willing to bet the diner stays NYP based just because it can. :angry2:
 
The Crescent comes up for review in 2011. And I think I mixed up the Palmetto and the Crescent in an earlier discussion.
 
Status
Not open for further replies.
Back
Top