Took IAD-PEK this past Saturday (Jan 3rd).

Flight took a eastbound path versus the usual westbound path over canada/eastern Siberia.

Flight Path Map showed progress for about 11+ plus hours, then my TV monitor changed and all the sudden my seatmate's stayed the same.

See photos below.

About an hour before the the screen changed, I looked out the window and could see Lake Baykal below the plane. Both on the physical identification of the lake, the flight path map shown on my monitor appeared to be incorrect.

What was going on? Anyone experienced this before?

I landed at JFK once at about 12:30am and just before when I woke up I had the flight map on and it showed us somewhere off the coast of West Africa.

I've seen something similar before EWR-HKG, though not the difference between the two monitors. At a guess, you have two different quirks happening simultaneously:

1) In trans-polar routes, the map often freaks out if you're headed straight over the pole. I've seen the East/West line on the map flip back-and-forth again, as it can't decide whether you've gone East or West to get to your destination. (When, in fact, you've actually done neither... due North, then due South!)

2) Sometimes the line shows the projected route (the smooth line), while other times it shows the actual route (the "bumpy" one). I've never figured out the determining factor in whether you get to see estimated or actual route info, nor ever seen an option to change it.

It looks like your flight path was effectively trans-Polar, and one of your maps showed the estimated path going Westbound, while the other showed the actual path, headed Eastbound. Weird!

I've seen this on the p.s. planes too. Note the coordinates of the following location:

https://www.google.com/maps/place/0�00'00.0"N+0�00'00.0"E

My best guess is that after the aircraft parks and various systems start to get shut down, some part of the navigation system reports 0�N, 0�E to the entertainment system's map as the current location, and the entertainment system therefore finds itself off of the coast of West Africa. One of our resident pilots may be able to confirm, though.

Probably has to do with jet stream and other winds aloft. When they are strong and in the right places, east coast trips to Asia can often shift on a slightly eastwards polar track to avoid as much of the winds as possible and thereby mitigate the fuel and flight time drag caused by plowing into headwinds.

This is normally not the case from the west coast as the distance is much shorter and going east would make no sense.

i noticed this on the app Saturday as well for EWR-PEK.

On a recent EZE-IAH flight, the flight screen indicated that we had traveled 18,000 miles. I thought that somebody forgot to hit the reset button on the odometer.

Airshow is pretty buggy in general, and since it's certified they're not achieving a lot of development on it.

I noticed this problem on my flight from ORD-HKG. We took to polar route going "east" the same route your map initially showed. Then I also noticed that it suddenly changed. I know it was wrong because I was recording the GPS route on my phone so I know how we flew. (Plus I later looked at it on flightaware, luckily it caught enough of the route)

I'm pretty sure it is a software bug. I wouldn't put too much trust into its accuracy.
My phone may not be that precise, but I trust it more than the screen.

Here is what my ORD-HKG flight last week showed. If only UA would credit me for the routing shown in the second photo!

A flight map is only as good as the programmers who implemented it. A second-year CS student could do a better job.

On a certified IFE system? Doubt it.

I would say if the framework was documented with inline code, and the schema was clearly understood then yes a 2nd year cs STUDENT could do a better job.

This is a pretty legitimate issue stemming from the representation of a sphere on a square map. When you get up on those polar flights, the map doesn't know how to draw the route and may bounce back and forth to draw the shortest line based on the aircraft's current location.

The only way to deal with it would be some kind of exception in the programming, but even that's hard to do because no two flights will ever fly the exact same path.

Having worked on Air Force certified systems (in the distant past), I don't.

These kinds of map projections have been around for a long, long time. It's a very well-understood problem with well-understood solutions.