Silver Service Coach/Sleeper Analysis

[Anderson]

I'm performing a revenue analysis on the Silver Service (Meteor/Star) for the Washington, DC to Orlando, Florida market, comparing coach and sleeper revenue per car. I am using Washington and Orlando because there is a substantial amount of traffic traveling from Washington (and points north) to Orlando (and points south), often filling one or two coaches completely out of DC. Charges from New York to Orlando are slightly higher than Washington to Orlando.

I will only attempt a small number of scenarios, won't take into account discounts, etc. since I am not a one-man revenue management office. As such, it should be obvious that scenarios beyond what I generate are possible.

First, the base data, per Amsnag (with confimation via Amtrak.com)
(1) Buckets
Coach:
$269
$207
$159
$121

Roomette (room charge):
$537
$456
$376
$295
$215

Bedroom (room charge):
$1147
$998
$855
$705
$556

Note that in both cases, an additional charge of $121/person (rail fare) will apply.

(2) Capacity
An Amtrak long-distance (Amfleet II) coach has 59 seats. A Viewliner I sleeper has 12 roomettes and 3 bedrooms.

(3) Scenarios:
Coach:
-Coach is 100% occupied at low bucket: $121*59 seats=$7139 in revenue. Revenue per passenger: $121.00
-Coach is 100% occupied at high bucket: $269*59 seats=$15,871 in revenue. Revenue per passenger: $269.00
-Coach is occupied 50% at low bucket, 50% at high bucket: $11,505 in revenue. Revenue per passenger: $195.00

Sleepers:
-All rooms at single occupancy, low bucket:
3 bedrooms, $677 in revenue each: $2031 in revenue
12 roomettes, $336 in revenue each: $4032 in revenue
Total: $6063 in revenue
Revenue per passenger: $404.20

-Roomettes at sigle occupancy, bedrooms at double occupancy, low bucket:
3 bedrooms, $798 in revenue each: $2394 in revenue
12 roomettes, $336 in revenue each: $4032 in revenue
Total: $6426 in revenue
Revenue per passenger: $357.00

-Roomettes at 1.5 occupancy, bedrooms at double occupancy
3 bedrooms, $798 in revenue each: $2394 in revenue
12 roomettes, $396.50 in revenue each: $4758 in revenue
Total: $7152 in revenue
Revenue per passenger: $298.00

-Roomettes at single occupancy, bedrooms at double occupancy, high bucket:
3 bedrooms, $1389 in revenue each: $4167 in revenue
12 roomettes, $658 in revenue each: $7896 in revenue
Total: $12,063 in revenue
Revenue per passenger: $670.17

-Roomettes at 1.5 occupancy, bedrooms at double occupancy, high bucket:
3 bedrooms, $1389 in revenue each: $4167 in revenue
12 roomettes, $718.50 in revenue each: $8622 in revenue
Total: $12,789
Revenue per passenger: $532.88

A point should be made that the odds of Amtrak selling out an entire coach at high bucket are pretty low. Ditto selling out an entire sleeper. However, this does give an idea of the range of revenue possibilities. With that said, at least until recently the trend has been to sell a lot more coach space at low bucket than not outside of peak days, while sleepers tend to sell more space at higher buckets from what I can tell. Thus the coaches will tend towards lower-bucket scenarios while sleepers will tend towards higher-bucket scenarios...but coaches will also have more short-hop traffic and therefore more turnover. Revenue is therefore likely to be about equal between the two cars.

Finally, on the $269 high bucket coach fare: That caught me off guard, since it is a violation of the 2:1 fare limits that had been observed until recently. I'm not sure what's up there.

 

[Guest]

I don't know. Roomettes at a peek of 1.5 occupancy seems kind-of low. That means at best, you assuming that there is only one passenger in half the roomettes. While I don't go knocking on each roomette door to count noses, from my experience on the Silvers is that most roomettes have two passengers. So, to me, 1.5 occupancy should be the low side and 2.0 occupancy being the high side.

 

[Paulus]

Worth noting, I think, that looking at NARPs ridership data for Orlando, there are about 13,781 passengers on that city pair and, just from eyeballing, at least three quarters are coach.

 

[afigg]

A point should be made that the odds of Amtrak selling out an entire coach at high bucket are pretty low. Ditto selling out an entire sleeper. However, this does give an idea of the range of revenue possibilities. With that said, at least until recently the trend has been to sell a lot more coach space at low bucket than not outside of peak days, while sleepers tend to sell more space at higher buckets from what I can tell. Thus the coaches will tend towards lower-bucket scenarios while sleepers will tend towards higher-bucket scenarios...but coaches will also have more short-hop traffic and therefore more turnover. Revenue is therefore likely to be about equal between the two cars.

Finally, on the $269 high bucket coach fare: That caught me off guard, since it is a violation of the 2:1 fare limits that had been observed until recently. I'm not sure what's up there.

I would say the odds of Amtrak selling out an entire coach for NYP-ORL at high bucket are zero. Would be useful to get the typical percentage breakdown of how many seats are available/sold at each bucket price for Amtrak in general and then for the eastern LD trains. With 4 and 5 bucket prices, how evenly distributed are they? 20% / 25% each or something like 35% for the low bucket and smaller percentages for each higher bucket? How many seats and rooms are typically sold at the highest bucket price?
I am surprised at the spread of bucket prices for coach and roomettes. The top coach bucket is pretty dang expensive. But then again, the higher NE Regional bucket prices on the NEC have gotten pretty dang expensive.

 

[Anderson]

My read is that low bucket probably accounts for somewhere in the 30-45% range on coach seats, with the share depending on the day(s) in question. I get the feeling the sleepers have more of a "bump" in the middle, with most rooms going in buckets 2-4 and a small number in both 1 and 5.

 

[neroden]

Eyeballing it, and using assumptions for bucket distribution, load factor, and occupancy as noted by others, it does look like *for a trip pattern such as Orlando to DC*, a sleeper car generally generates more revenue than a coach car, though not a huge amount more. Certainly enough to pay for the 1 attendant per car (rather than 1 for every 2-3 cars in coach).

In business terms, the coach cars are really intended for the shorter trips, and that's where they make more money.

This doesn't surprise me at all. I've always said that sleepers are simply something you provide when the runtime on likely trips gets longer than a certain amount, because at that point they make business sense, whereas they don't make sense for trips with shorter runtimes. If you don't want to run sleepers, you run the train faster.

People taking coach over very long distances are generally price conscious. I think the high buckets in coach on ORL-WAS are not intended to be sold at all; those numbers are there to hold the seats open for shorter-distance traffic. I expect that high-bucket coach ticket purchases are mainly for shorter distances.

It would be interesting to see the numbers for Miami-NY. I wonder whether the price differential between sleepers and coaches would be larger on the longer trip. I would expect it to be, but Amtrak has done strange things before; for example, I find the ticket pricing on the California Zephyr bizarre, as it seems to actually subsidize longer trips at the expense of shorter trips, and I can't see any business sense in it. By contrast, Lake Shore Limited ticket pricing has always made sense to me.

 

[Anderson]

Eyeballing it, and using assumptions for bucket distribution, load factor, and occupancy as noted by others, it does look like *for a trip pattern such as Orlando to DC*, a sleeper car generally generates more revenue than a coach car, though not a huge amount more. Certainly enough to pay for the 1 attendant per car (rather than 1 for every 2-3 cars in coach).

In business terms, the coach cars are really intended for the shorter trips, and that's where they make more money.

This doesn't surprise me at all. I've always said that sleepers are simply something you provide when the runtime on likely trips gets longer than a certain amount, because at that point they make business sense, whereas they don't make sense for trips with shorter runtimes. If you don't want to run sleepers, you run the train faster.

People taking coach over very long distances are generally price conscious. I think the high buckets in coach on ORL-WAS are not intended to be sold at all; those numbers are there to hold the seats open for shorter-distance traffic. I expect that high-bucket coach ticket purchases are mainly for shorter distances.

It would be interesting to see the numbers for Miami-NY. I wonder whether the price differential between sleepers and coaches would be larger on the longer trip. I would expect it to be, but Amtrak has done strange things before; for example, I find the ticket pricing on the California Zephyr bizarre, as it seems to actually subsidize longer trips at the expense of shorter trips, and I can't see any business sense in it. By contrast, Lake Shore Limited ticket pricing has always made sense to me.

Yes and no. From what I can tell, Amtrak also uses them to grab revenue where they can at the last minute (i.e. when flying would be several times more expensive and/or unavailable). For example, I would expect that 10-20% of seats would go at high bucket over Thanksgiving weekend...simply because Amtrak can sell them at those prices that weekend while the sleepers will get sold out.

On the Silvers, things get odd past Orlando/Washington:

WAS-ORL: $121/269

WAS-MIA: $137/304

NYP-ORL: $134/297

NYP-MIA: $141/314

This is why I used WAS-ORL as a benchmark...through traffic from WAS-north to ORL-south will be filling up the same slots, and there is a lot of it. When I've been on the Meteor, getting loaded into one of the back coaches in DC, I regularly walk through well over a coach of traffic going to ORL, WPB, FTL, MIA, and so forth.

 

[Paulus]

Eyeballing it, and using assumptions for bucket distribution, load factor, and occupancy as noted by others, it does look like *for a trip pattern such as Orlando to DC*, a sleeper car generally generates more revenue than a coach car, though not a huge amount more. Certainly enough to pay for the 1 attendant per car (rather than 1 for every 2-3 cars in coach).

It's not just the attendant though; consider the food costs.

It would be interesting to see the numbers for Miami-NY. I wonder whether the price differential between sleepers and coaches would be larger on the longer trip. I would expect it to be, but Amtrak has done strange things before; for example, I find the ticket pricing on the California Zephyr bizarre, as it seems to actually subsidize longer trips at the expense of shorter trips, and I can't see any business sense in it. By contrast, Lake Shore Limited ticket pricing has always made sense to me.

Approximately 3,069 passengers per year; counting the pixels, 80% is coach traffic. For Orlando-DC, pixel counting is 72.72% as coach traffic.

 

[neroden]

So there is some sanity to the pricing: the prices increase when you take a longer trip on the same train, and the "room charge" for the sleeper increases as well. None of this "Green River to Denver is cheaper than Grand Junction to Denver" weirdness. Good.

I did a little scouting of NYP-MIA.

Bedroom:

$1147

$998

$855

$705

$556

So the bedroom charge is the same amount from NYP-MIA as from WAS-ORL -- which is sensible enough given the ridership patterns. If the charge was less, something weird would be going on.

Interestingly, *I had trouble finding all the roomette buckets* because there are so many sold-out-roomette days:

Roomette:

$567

$482

$397

$312

$227

So the roomette charge is more from NYP-MIA than it is from WAS-ORL. Again, perfectly sensible. The premium for sleeper service does go up as you take longer trips, though not by that much.

----

For a different example which is close to my heart, buckets from Amsnag on the Lake Shore Limited, Syracuse to Chicago:

Roomette:

$431

$371

$306

$246

$181

Bedroom:

$844

$726

$600

$481

$356

Coach:

$150

$120

$96

$77

Worth noting is that in this case, roomettes are sold out on many dates and bedrooms are sold out on many dates --- but coach is low bucket on almost every date, and the high buckets (above $96) are showing up *only* on this coming weekend.

I think you can assume that almost all coach tickets for SYR-CHI are low-bucket; but the sleepers are selling in the middle and high buckets routinely.

WAS-ORL is actually quite similar. But it has much stronger coach ridership. In this case there ARE some dates where the coach is showing higher buckets, but eyeballing it it's something like a quarter of the dates -- most of the seats are still being offered at low bucket.

Looking at this pattern, I'd guess that the "all seats low bucket" assumption is probably roughly correct for coach; it might be more accurate to say 3/4 low-bucket, with 1/4 higher-bucket on some dates and 1/4 vacant on other dates, but I think it comes out to much the same revenue.

The high-bucket prices and repeated sell-out conditions on the sleepers, combined with low-bucket prices and lack of sellout conditions in coach, indicates to me that additional sleeping cars on the Eastern trains *will* add to Amtrak's bottom line.

 

[Paulus]

The high-bucket prices and repeated sell-out conditions on the sleepers, combined with low-bucket prices and lack of sellout conditions in coach, indicates to me that additional sleeping cars on the Eastern trains *will* add to Amtrak's bottom line.

But would they do so more than another coach car? You've got, what, three sleepers on the Lake Shore Limited for instance, with ~50% of sleeper passengers going terminus to terminus (thus why Syracuse has issues, especially if it is at or near a turnover point for coach). Increasing supply by 1-2/3rds may result in significantly lower buckets for any gained ridership (and lower revenue if they aren't able to increase ridership significantly). If they're averaging only 50 sleeper passengers per train (which they are), despite 90 sleeper "seats," they may already be at the sweet spot for sleepers.

 

[neroden]

The high-bucket prices and repeated sell-out conditions on the sleepers, combined with low-bucket prices and lack of sellout conditions in coach, indicates to me that additional sleeping cars on the Eastern trains *will* add to Amtrak's bottom line.

But would they do so more than another coach car?

Yes, on the LSL at least, they most likely would. Because the existing coach cars aren't filling up.
If you add another coach car, you're either adding empty space, or you're reducing prices. It's certainly possible that the reduction in coach ticket price combined with increase in volume would make for better total revenue, but it's unlikely.

On the LSL in particular, most everyone is heading for South Bend or Chicago, with some heading for Cleveland and Toledo. This is because people going shorter distances within NY are taking the Empire Service. Amtrak's coach price from SYR-CHI is already less than the gas cost of driving (unless you have a particularly efficient car), less than the cost of flying, and substantially less than the cost of Greyhound. I don't think that lowering the coach price would be likely to get big increases in ridership.

You've got, what, three sleepers on the Lake Shore Limited for instance, with ~50% of sleeper passengers going terminus to terminus (thus why Syracuse has issues, especially if it is at or near a turnover point for coach). Increasing supply by 1-2/3rds may result in significantly lower buckets

Unlikely. Remember, the sleepers on the Eastern trains to which we are referring are currently selling out routinely.

The situation is quite different on the Superliner trains, where the roomettes are *not* selling out. Though the bedrooms still are selling out. (Some people have said that the bedroom/roomette balance was wrong in both the Superliners and the Viewliners. They may have been right.)

This is pretty simple to figure, actually. If the coaches are not selling out, but the sleepers are, then adding coaches means reducing ticket prices, but adding sleepers does not. If it's the other way around, then it's the other way around.

Obviously at some point the trains will have "enough" sleepers, they won't continue to sell out at high bucket prices, and it will not make sense to add more. It's quite likely that Amtrak's original planned Viewliner sleeper order of 100 would have saturated the sleeper market in the East. Or the current 50 plus the added 25 which Amtrak ordered from CAF may be the sweet spot; I would assume that Amtrak has made some attempts to predict the market.

Eyeballing it, and using assumptions for bucket distribution, load factor, and occupancy as noted by others, it does look like *for a trip pattern such as Orlando to DC*, a sleeper car generally generates more revenue than a coach car, though not a huge amount more. Certainly enough to pay for the 1 attendant per car (rather than 1 for every 2-3 cars in coach).

It's not just the attendant though; consider the food costs.

Meh, food costs are fairly small if you're already running the dining car. It'll cover that too. And I am very much of the view that you already have to be running the dining car for a runtime that long, even for an all-coach train. Half the LSL's dining car patronage is from coach passengers.

After looking at the typical prices, I'm even more sure that the added revenue from sleepers will cover the variable food costs. (On the Viewliner trains. I'm not at all sure about the Superliners.) Coach seats almost all sell for low bucket. Roomettes and bedrooms usually come close to selling out, more of them are assigned to the middle buckets, and so the average revenue is going to be more like the center bucket.

A couple switching from coach to a low-bucket roomette on the LSL from SYR-CHI more-than-doubles their ticket price. A single person buying a mid-bucket roomette multiplies their ticket price by just under five. For a couple switching from coach to the high-bucket bedroom, it multiples their ticket price by more than 6 times.

The premium seems to vary a lot depending on what train and what segment you're looking at, and I wouldn't claim generically that *all* the trains would benefit from added sleepers. The sleeper premiums for CHI-DEN are lower than those for SYR-CHI. The premiums for CHI-EMY are higher. (As I suspected, if you're crazy enough to ride CHI-EMY in coach, Amtrak doesn't charge you very much.)

I still consider sleepers to be simply "something you do when the runtime is longer than X". This is borne out by the situation in the Western trains.

For instance, the coach prices on the California Zephyr remains really minimal for the longest-distance trips, but the sleeper prices keep going up as the distance gets longer. Some sleeper service is probably a net financial positive on a route like the California Zephyr as long as the train is running from Chicago to California. The question is not a question of sleeper service -- that's just something you do when the runtime of the train gets long enough -- the question is whether running the train across two sets of mountains and the Utah desert makes sense at *all*.

A lot of people have speculated that Boston-Washington-(Newport News) will get its lone sleeping car back when the new Viewliners arrive. This makes sense: the route is just long enough. It's just short enough it doesn't need a dining car, though. It might give a better picture of the true situation with sleeping cars if it comes back.

 

[tricia]

The revenue-per-passenger totals for roomettes and BRs are affected not only by various proportions of tix sold at various buckets, but also by passengers booking accomodations with AGR redemptions. To what extent this matters, I don't know.

 

[Ryan]

So there is some sanity to the pricing: the prices increase when you take a longer trip on the same train, and the "room charge" for the sleeper increases as well. None of this "Green River to Denver is cheaper than Grand Junction to Denver" weirdness. Good.

You may be interested in reading some threads from a while ago. Some of us did some analysis like this once upon a time:

http://discuss.amtraktrains.com/index.php?/topic/35761-a-look-at-fares-on-the-crescent/

http://discuss.amtraktrains.com/index.php?/topic/36205-amtrak-fare-comparisons/

I wonder if some of the patterns have held up over the years or not. Maybe I'll dig up some numbers to compare...

 

[neroden]

You may be interested in reading some threads from a while ago. Some of us did some analysis like this once upon a time:

Thanks! I like looking at analysis like this. (While sitting at home sneezing and coughing.)

 

[neroden]

The revenue-per-passenger totals for roomettes and BRs are affected not only by various proportions of tix sold at various buckets, but also by passengers booking accomodations with AGR redemptions. To what extent this matters, I don't know.

It's probably totally impossible to compute without inside information. I also would consider it a cost of running AGR, and really nothing to do with train operations. Amtrak is not to the point airlines are (where almost nobody pays for first class on domestic airlines, and they use it strictly as a marketing tool).

 

[Paulus]

I am extremely doubtful of claims that the sleepers are filling up on the Lake Shore Limited as anything more than a freak occurrence or as a result of heavy seasonal demand (which is also going to fill up the coaches). The consist posted here says there are 6 coach cars and 3 sleepers. That's 354 coach seats and 90 sleeper "seats." For FY13, there were 359,192 coach passengers and 36,263 sleeper passengers. Or, with a daily train in each direction, that's 492 coach passengers per train (139% of capacity) and 50 sleeper passengers per train (55% of capacity). Obviously it isn't standing room only for the coach passengers, there's a deal of turnover present at various points. Note that this is with Amtrak shooting themselves in the foot by offering absolutely terrible times for Ohio travelers; coach ridership would absolutely soar with a daylight runtime.

It is impossible for the Lake Shore Limited to be routinely selling out if it is barely selling half of its sleeper "seats," especially if there is a strong seasonal or day of week bias. The appearance might be given though by a combination of selection bias (you're looking at traveling at the same time as everyone else) and a high percentage of single occupancy passengers. There may also be a greater readiness to raise fares on the sleepers than there is for the coach passengers. Anderson's noted some funkiness with how seats are allocated in Amtrak's system in the past and that might also be giving a misleading impression.

 

[neroden]

I am extremely doubtful of claims that the sleepers are filling up on the Lake Shore Limited as anything more than a freak occurrence or as a result of heavy seasonal demand (which is also going to fill up the coaches).

Just check Amsnag. :eyeroll: I'm not saying they're selling out daily, but roughly once or twice a week. You have biases against sleepers and want to believe that they're unpopular. You're wrong.

 

[neroden]

That's 354 coach seats and 90 sleeper "seats."

I don't know how you get to 90 sleeper "seats", but you're wrong.

12 roomettes and 3 bedrooms (including handicapped) per sleeper, after subtracting the sleeper attendant's room. Total of 36 roomettes, 9 bedrooms.

Subtract 10 (!!!) roomettes for coach, lounge, and diner staff -- which appears to be about right -- and you're at 26 roomettes, 9 bedrooms. Yes, the staff are eating up a LOT of space.

and 50 sleeper passengers per train (55% of capacity).

There's clearly quite a lot of sales of roomettes to individuals rather than couples. Assume that the bedrooms are all sold to 2 people each. If you consider that about half the roomettes are sold to individuals rather than couples -- the "1.5 passengers per roomette" scenario mentioned before -- you end up with a realistic capacity of 57.

So yes, the sleepers are selling out. Routinely. As long as Amtrak allows individuals to reserve entire roomettes (rather than the "share with someone random" procedure used in Europe), the LSL sleepers are running at roughly 87% of practical capacity. It may also be the case that roomettes are reserved by individuals even more than half the time.

(As a further point, you can't expect much turnover of these rooms -- the market is Chicago/S Bend to upstate NY/NYC/Massachusetts. Few will spring for roomettes on the "orphan legs" of the trip. Even the coach passengers preferentially take Empire Service if they're going from Buffalo to points east, and S Bend to Chicago is dominated by the South Shore Line. Would I love to see a second frequency which was a day train in Ohio? Certainly I would, and that train probably wouldn't need sleeper cars. But that train wouldn't serve the Upstate NY-Chicago or NYC-Chicago markets well at all, so it would have to be a *second* frequency. On another matter, if OTP was better, the coach ridership of the LSL east of Buffalo might improve, and the roomettes might even get resold.)

The "1.5 passengers per roomette" scenario, which seems to be roughly what's going on, still makes a sleeping car more profitable than a coach.

Now, if you want to start a discussion over whether the coach and cafe staff members should each get an entire roomette to themselves, be my guest -- I don't see why they shouldn't use both bunks at least. But that's the current employment contract, apparently.

If the staff on the LSL were 2 to a roomette, even accounting for "one off" due to gender-matching, it would free up 4 roomettes for paying customers. Which I'd estimate would be about 1.4 million dollars in revenue per year.

 

[Paulus]

\

I am extremely doubtful of claims that the sleepers are filling up on the Lake Shore Limited as anything more than a freak occurrence or as a result of heavy seasonal demand (which is also going to fill up the coaches).

Just check Amsnag. :eyeroll: I'm not saying they're selling out daily, but roughly once or twice a week. You have biases against sleepers and want to believe that they're unpopular. You're wrong.

Once or twice a week is entirely in line with what I stated.

That's 354 coach seats and 90 sleeper "seats."

I don't know how you get to 90 sleeper "seats", but you're wrong.

Rooms times berths, didn't think to account for the lack of a dorm. Idiotic levels of staffing on that train though; they should just all be dumped and treat it like a somewhat longer corridor train. The sleepers are almost a self-licking ice cream cone.

So yes, the sleepers are selling out. Routinely. And the "1.5 passengers per roomette" scenario still makes a sleeping car more profitable than a coach.

No, it had slightly larger gross revenue. Profit means subtracting costs, like the diner operations.

 

[neroden]

Again, you want to misapply fixed dining car costs to the sleepers for some irrational reason. I realize there is no reasoning with you on such matters, Paulus; you have a mindless vendetta against sleepers. At this point I'm just debunking you for the audience.

One thing we can agree on: the staffing situation on the "long distance trains" is off in substantial qualitative ways from what it should be. I wouldn't necessarily say that there are *too many* staff exactly, but they sure don't seem to be deployed for maximum effectiveness.

As the recent discussion about bathroom cleanliness showed, coach car attendants are needed on long runs -- perhaps even more than are currently employed -- but are they actually getting the necessary work done? Does it make any sense to have coach car attendants far away in a dorm car where the coach passengers can't find them? (I say no.) Why is there one coach car attendant trying to handle two or even three entire coach cars, but only one sleeper car attendant per sleeping car (with arguably a lot less work)?

Perhaps the dining car really does need 5 on-board staff -- but if so, why isn't the dining car open all day? Dining cars *used to be* open all day. (Yes, I know, they spend a bunch of the day on paperwork. That really has to be ended ASAP.) Why does each staff member get a *single-occupancy* roomette, which is the second-best accomodation available on the train, better than most paying customers can afford? Et cetera.

These are really questions of Amtrak's union work rules contracts. Unfortunately, in the last round of "negotiations", President George W. Bush's "Presidential Emergency Board" decided to give the unions everything and more than they'd asked for, in what I can only describe as another attempt to hamstring Amtrak.

I'd like to see the results for a sleeper on #66/67, because the only added staff member would be the sleeping car attendant. Therefore, it wouldn't have the weird accounting effects of the "dorm space" assigned to coach car and cafe car attendants or the loss-leader character of the dining car. It would give you a much cleaner picture of what's actually going on with a sleeper, as opposed to the rest of this union-rules situation.

 

[Anderson]

On the number of attendants in the sleepers vs. in coach: A good part of that is bad comms. In the Crescent PIP, Amtrak mentioned trying to sort out the comms between cars to allow a sleeper to be "attached" to the adjacent car. If that can be sorted out, Amtrak could go to 2 SCAs per three cars, which is a bit closer to the coach situation. That would also throw another room into inventory. Per the PIP, the SCAs under this setup wouldn't have to cover more than 21 rooms (at least, once the toilet situation is sorted out), which close to the number a normal Superliner SCA covers (and well below the number the Transdorm SCA can be stuck with).

As to the amount of sleeper space available: I'm going to agree that 90 passengers in three Viewliners is insanely high. You're assuming 100% occupancy at full capacity...that just does not happen. Nathanael pointed out, you've got a decent number of OBS to house, and you've got plenty of folks who travel single in a roomette. That would, by the way, include me about 90% of the time. 1.5 passengers per roomette is a realistic number, IMHO, though it might actually err on the high side a little. Likewise, I know bedrooms won't quite reach 2 passengers per bedroom, if only because the handicapped room can end up going to a single rider without too much trouble. However, 1.5/roomette and 2.0/bedroom is the best approximation I can come up with for passenger behavior.

Next, let's talk seasonal variation. You're almost never going to sell a train out every single day. If you do, that's a disaster because you're implicitly turning away a ton of traffic. However, if you can get close to selling out your train about a third of the time, you're doing pretty good. Mind you, that third isn't going to be spread out symmetrically: There are times (such as Thanksgiving weekend or peak season in the summer) that Amtrak could probably sell its long distance trains out twice over. On the other hand, there are also times (such as a good portion of January/February) that Amtrak could probably put part of their fleet in the shop and not turn folks away.

Finally, let's talk directional variation. Even the Auto Train doesn't get to 100% of its practical capacity...but the 65-75% they're regularly managing now is pretty good. Why? That includes sold-out trains that "pair" with partly empty trains because there's seasonal traffic in one direction or the other (thanks to snowbird migration patterns. There are also trains that are amazing one way and godawful the other: I can leave Washington on the Cap and hope to make a morning appointment in Chicago (and indeed I have done that). The reverse? Not so much. Ditto the LSL: I can leave New York only losing an hour or two of working time on each end of the trip. Leaving Chicago, I lose the entire next day crossing New York state. Going to Florida, I usually take the Meteor south and the Star north, since those are the "evening" departures from RVR and DLD, respectively.

 

[afigg]

On the number of attendants in the sleepers vs. in coach: A good part of that is bad comms. In the Crescent PIP, Amtrak mentioned trying to sort out the comms between cars to allow a sleeper to be "attached" to the adjacent car. If that can be sorted out, Amtrak could go to 2 SCAs per three cars, which is a bit closer to the coach situation. That would also throw another room into inventory. Per the PIP, the SCAs under this setup wouldn't have to cover more than 21 rooms (at least, once the toilet situation is sorted out), which close to the number a normal Superliner SCA covers (and well below the number the Transdorm SCA can be stuck with).

I wonder what is the status of the Crescent PIP proposal to have an SCA cover 1-1/2 Viewliner rooms? Splitting 3 Viewliners between 2 SCAa would clearly cut costs, much more so than many of the cost trimming measures now underway. Has the idea been dropped or is it, like so many other PIP recommendations, waiting on the new Viewliners and in this case, upgrades to the comms in the current Viewliners?
However, how would having 2 SCAs cover 3 Viewliners work on a train with 4 continuous Viewliner sleepers if the Meteor is expanded? One SCA would have only 1 car to cover resulting in an uneven workload. OTOH, on a 4 sleeper LSL with the BOS sleeper on the other end, the BOS sleeper SCA would logically only have one

Same goes for a 2 Viewliner sleeper configuration (after the bag-dorms are deployed). Lighter workload for the SCAs. Does the current agreement allow this idea to be implemented or would Amtrak would have to negotiate with the SCAs? Maybe pay the SCAs a little more when they have 1.5 Viewliners to cover?

With regards to directional variation, the AutoTrain may have the strongest directional variation of any train in the system during the peak periods. Easy to see this in Amsnag when the AT is frequently sold out or only the highest buckets are left in one direction over a month while the other direction has many days with available rooms and lower bucket prices.

What these discussions about the Meteor and LSL sleeper revenue numbers show is that the bag-dorms are badly needed. They will provide a quick boost in roomette capacity and revenue with no additional SCAs to pay for. Where the heck is the test set of new Viewliners?

 

[Paulus]

One thing I did neglect to point out: If they are filling up, the answer is quite clearly to raise prices until that is no longer a problem as has been done with the Acela (the slowest, lowest capacity, and priciest high speed train in the world).

Again, you want to misapply fixed dining car costs to the sleepers for some irrational reason. I realize there is no reasoning with you on such matters, Paulus; you have a mindless vendetta against sleepers. At this point I'm just debunking you for the audience.

No, it is for the entirely rational reason that the diners exist for the sake of the sleepers. You may have noticed a lack of diners on the trains without sleepers. The Lake Shore Limited has a high percentage of coach travelers using the diner (though that begs the question of just how many travelers at all use it), but they spend quite a paltry sum, certainly not enough to justify a full diner (as Amtrak recognized). Doesn't even make the LSL's F&B service perform much better for that matter.

As the recent discussion about bathroom cleanliness showed, coach car attendants are needed on long runs -- perhaps even more than are currently employed -- but are they actually getting the necessary work done? Does it make any sense to have coach car attendants far away in a dorm car where the coach passengers can't find them? (I say no.) Why is there one coach car attendant trying to handle two or even three entire coach cars, but only one sleeper car attendant per sleeping car (with arguably a lot less work)?

One person roving the entire train doing their job should ensure that the toilets remain clean. Quite frankly I've never understood why there are coach attendants in the first place if they aren't doing things like at seat meal service.

 

[jis]

I must say that I have never understood the need for Coach attendants either, specially when there are no pillows to hand out. They should be replaced by a single keep the toilets tidy person per 4 coaches. One person should be able to manage 8 toilets, perhaps even 10.

BTW, just the fact that one can raise fares on a highly traveled corridor with huge business clientale, where Acelas provide a useful service with the right O-D times etc.has very little relevance to what can be done in Sleepers. If anyone truly believes that these two are somehow relevant to each other just shows a profound lack of understanding of what causes people pay for a service.

Also except for the coneheaded railfans, none cares whether a train is fastest or the slowest fast train or has low or high capacity. As long as it serves a customers purpose they will ride and pay the fare that they deem to be the value of the service. So all of those other factors are mostly irrelevant too.

 

[Ryan]

You're like the worst railfan ever, Jishnu. All those facts and logic clouding up the simple truth that Amtrak would be so much better if Internet Experts were put in charge.