SHUPirate1
Topic Author
Posts: 3428
Joined: Sat Sep 20, 2003 2:53 pm

How Was "City Pair Combinations" Statistically Calculated?

Thu Dec 22, 2011 6:50 am

OK...so I'm dating myself a bit here, but going back about, oh, 15 years, American Airlines listed something called "city pair combinations" in their hub profiles, for example, http://web.archive.org/web/199610280...r.com/aa_home/servinfo/hubprof.htm

My question is, if anybody has any idea, how was this "city pair combinations" statistic calculated? Presumably it was based on connecting times and deviation from straightline routing, but does anybody have any idea how that was calculated?

Thanks!
Burma's constitutional referendum options: A. Yes, B. Go to Insein Prison!
 
LAXintl
Posts: 20183
Joined: Wed May 24, 2000 12:12 pm

How Was "City Pair Combinations" Statistically Calculated?

Thu Dec 22, 2011 8:58 am

Actually sounds pretty straight forward. Its simply math distribution model number of how many spokes(nodes) the hub has multiplied to give to total number of potential connection possibilitiies. For example a 20 spoke hub has 380 theoretic connection potentials.

Airlines will often utilize such marketing sound bites in things like DOT route applications regardless of the unlikeliness that many such markets would really never connect in real life (eg. BOS-LGA via ORD hub), or that some connections might have terrible timings especially for those nodes with low frequency service.

[Edited 2011-12-22 01:04:04]
From the desert to the sea, to all of Southern California
 
SHUPirate1
Topic Author
Posts: 3428
Joined: Sat Sep 20, 2003 2:53 pm

How Was "City Pair Combinations" Statistically Calculated?

Thu Dec 22, 2011 9:37 am

That's my question though...are they doing it that way? I couldn't quite get the numbers to work out that way for any of those four examples I gave (based on other data listed, and in the case of San Juan, not even close), and I have a feeling it's because they didn't, for instance, figure in LHR-ORD-ORY, just to give an example. (I believe that was before AA moved kicking and screaming over to CDG, and also, I'm not even sure they flew ORD-Paris at the time, but just one example of a route combination they probably didn't count)
Burma's constitutional referendum options: A. Yes, B. Go to Insein Prison!
 
SHUPirate1
Topic Author
Posts: 3428
Joined: Sat Sep 20, 2003 2:53 pm

How Was "City Pair Combinations" Statistically Calculated?

Thu Dec 22, 2011 6:05 pm

Anybody? Help? Bueller?

I'm stumped.
Burma's constitutional referendum options: A. Yes, B. Go to Insein Prison!
 
mogandoCI
Posts: 1247
Joined: Thu Jun 04, 2009 2:39 pm

How Was "City Pair Combinations" Statistically Calculated?

Thu Dec 22, 2011 6:11 pm

Quoting LAXintl (Reply 1):
Actually sounds pretty straight forward. Its simply math distribution model number of how many spokes(nodes) the hub has multiplied to give to total number of potential connection possibilitiies. For example a 20 spoke hub has 380 theoretic connection potentials.

Do they count EVERY combination or just logical combinations? SFO-ORD-LAX doesn't sound like a "city pair combo" to me....
 
Flighty
Posts: 7688
Joined: Thu Apr 05, 2007 3:07 am

How Was "City Pair Combinations" Statistically Calculated?

Thu Dec 22, 2011 6:23 pm

Not every city pair is realistically served. Most likely, it is around 25% of the total (because there are a number of small stations going with infrequent schedule to 1 hub, for an example.) So, with 200 spokes, you might claim 10,000 city pairs instead of 40,000. Including limitations on triple connect, layovers greater than 4 hours etc.

Who is online

Users browsing this forum: AJMIA, Baidu [Spider], Bing [Bot], dennypayne, fkfnz, flydia, Google Adsense [Bot], grbauc, hayzel777, ikolkyo, Lindegaard, log0008, rutankrd, seahawk, sise1989, TK787, TWA772LR, ty97, womenbeshopping, Yahoo [Bot] and 290 guests