Qantas and Virgin Australia are each claiming to be Australia's most on-time airline, based on official Government statistics released last week.

"Qantas has been crowned the most on-time major domestic airline for the third straight year in the latest official Australian aviation statistics released today," crowed the Red Roo in a press release issued last Friday morning.

Later that day, Virgin Australia shot back with its own statement insisting it "outperformed the Qantas branded airlines and Qantas group of airlines in on-time performance for Financial Year 2012, as per the BITRE statistics published today."

Who to believe? We decided to cut through the PR spin, crunch the official numbers and slice and dice the stats to reveal who's right, who's wrong -- and which airline really takes the Gold Medal in the on-time stakes.

As it happens, although Qantas and Virgin Australia are using the same numbers they're counting those numbers in different ways -- ways which allow both airlines to be correct in their respective on-time claims.

But those numbers come with several riders, and as we dug into the statistics we found several reasons why even the seemingly simple 'on-time' measurement can't be taken at face value.

Dicing the BITRE data

The wellspring for the raw data is The Bureau of Infrastructure, Transport and Regional Economics (BITRE) -- a statistical organisation that monitors, among other things, aviation statistics, and generates monthly reports of on-time airline performance.

Eight "reporting airlines" -- covering the majority of Australia's flying -- participate in BITRE reports: Jetstar, Qantas, QantasLink, Regional Express (Rex), Skywest Airlines, Tiger Airways, Virgin Australia, and Virgin Australia  ATR/F100 Operations.

Here's the relative size of each reporting airline in terms of "sectors flown" during June 2012:

"Sectors flown" means one-way flights that actually took off and landed, so it excludes cancelled flights.

In terms of pure numbers of sectors flown in June 2012, that's:

• Qantas: 9,541
• QantasLink: 8,540
• Jetstar: 5,569
• Virgin Australia: 10,722
• Virgin Australia (ATR +F100): 1,148
• Rex: 5,221
• Skywest: 831
• Tiger: 935

Since Rex, Skywest and Tiger aren't party to the Qantas-Virgin stoush, and are a relatively small proportion of flights, we'll put them aside in the stats.

Three fruity flavours of 'Qantas'

One of the complications around BITRE's statistics on Qantas is that it's possible to count "Qantas" in three ways:

• there's the mainline Qantas domestic fleet which carries the Qantas name
• there's the combination of Qantas and its regional arm QantasLink – together counted as 'Qantas-Branded' airlines
• and finally, you can add the Qantas-owned Jetstar to get the corporate body known as the 'Qantas Group'.

Similarly, the BITRE data splits out Virgin Australia's version of QantasLink.

Those are flights operated by Alliance Airlines (which usesFokker F100s) and SkyWest Airlines (running ATR72 turboprops) as "Virgin Australia - ATR + F100 Operations", listing them separately from what you might call Virgin Australia's mainline Boeing 737 and Airbus A330 services.

Naturally, both Qantas and Virgin use this multiplicity of brands to spin the statistics in their favour.

Virgin Australia measured itself against "the Qantas branded airlines and Qantas group of airlines" – which respectively includes QantasLink and Jetstar – rather than just the Qantas mainline.

However, Qantas excludes QantasLink from its on-time tally.

Given that BITRE reports QantasLink as flying 48% of flights marked "Qantas" in June (not including Jetstar flights, of course), with QantasLink flights using QF flight codes and there being no obvious difference between Qantas and QantasLink when you book, it seems a fudge. 

On-time arrivals matter more than departures

One issue we have with the BITRE on-time statistics being trumpeted by Qantas and Virgin Australia is that these are for departures, not arrivals.

Stop and think: as a traveller, which of these matters the most to you -- that a plane leaves on time, or that it arrives on time?

We'd suggest it's the latter.

As annoying as it is to be sitting on the plane or at the gate for an extra 20 minutes before the scheduled take-off time, it's a far greater inconvenience to business travellers for a flight to arrive 20 minutes late. 

And if a flight takes off late but makes up the time en route for an on-schedule arrival, that's just fine by us.

As a result, an on-time rating based on departures is far less relevant than one based on arrivals.

(And, as an FTI: there's 15 minutes of 'wiggle room' built into the BITRE reporting. According to the agency, 'on time' means that your plane departs or arrives within fifteen minutes of the scheduled time.)

Cancellations should count

Here's another oddity which may surprise you: the on-time statistics numbers don't include cancelled flights. So if an airline cancels a flight, this doesn't count against its on-time record.

That may be correct from an airline operations point of view, but we'd suggest that cancelled flights are usually more inconvenient for passengers than a late depature or even a late arrival.

It's also worth noting that Qantas and Virgin were citing numbers based across the entire 2011-2012 financial year.

Financial years may have deeply significant meaning to bean-counters but we'd suggest a record over the past 12 months have less significance to a traveller than, say, the last three months or even six months.

The numbers that matter

So we dived head-first into the BITRE data (you can download it and dissect it for yourself at the BITRE website) to nut out the numbers for on-time arrivals.

We threw out the 2011-2012 financial year timespan to focus on just the first six months of this year, with the on-time arrivals for June 2012 as our most recent snapshot.

And we used the number of scheduled flights instead of the number of flights that actually flew in our on-time measurements so that cancellations are factored in.

For the mathematically-inclined, or if you want to try this exercise using the BITRE data, this means an airline's most accurate on-time percentage is calculated by dividing its scheduled departures by its on-time arrivals.

Where we need to combine the results of specific airlines into a group, it's just a matter of adding together the group's scheduled departures, and dividing that by the sum of the group's scheduled on-time arrivals.

Note that we haven't simply averaged the percentages, which might seem temptingly simple but which would be inaccurate and misleading.

So which is really the most on-time airline?

Using our revised formula, for the month of June 2012, Virgin Australia turns out to be Australia's most on-time airline.

Mainline Virgin beat mainline Qantas in the "getting you there on time" stakes by a nose: 83.1 percent for Virgin Australia to 82.7 percent for Qantas.

Here's the full month's on-time rankings:

1. Virgin Australia: 83.1 percent
2. Qantas: 82.7 percent
3. Virgin Australia - ATR + F100: 80 percent
4. Jetstar: 77 percent
5. QantasLink: 65.1 percent

QantasLink is hands-down Australia's worst on-time airline, with just 65.1 percent of planes arriving on time according to our counting.

And that's not counting the 3.7 percent of all QantasLink flights that BITRE says were cancelled completely, compared with between 0.5 percent and 1.1 percent for the rest.

In the 'group' stakes, Virgin is even further ahead if you include QantasLink and Jetstar on the the Red Roo's side of the ledger.

And the Virgin Australia Group (mainline, ATR and F100) scored 82.8 percent on-time, well ahead of Qantas-Branded (74.4 percent) and Qantas Group (74.9 percent) flights.

On-time trends: January-June 2012

In addition to the most recent month's results, we applied the same formula from January to June to map the overall trend and give some context to that one-month snapshot.

The pictures speak for themselves: a generally downward trend for Qantas and QantasLink is matched by a largely upward trend for Virgin Australia (and the Virgin Australia - ATR + F100 operation).

On a group level, Virgin Australia Group has been ahead of Qantas Group since March -- by a very slight (less than one percent) margin until May, but since then by 3 percent in May and now by 8 percent in June.

Splitting out the Jetstar and Qantas-Branded numbers is a very similar picture: since QantasLink flies so many of the Qantas-Branded and Qantas Group flights, it's pulling down the rest.

So that's our take on the on-time game, using what we believe is a superior set of the core BITRE criteria than those applied by either Qantas or Virgin Australia in their most recent round of chest-beating.

We're interested in hearing your thoughts on this methodology and how to best represent an airline's on-time performance: hammer your keys in the comment field below!

Follow Australian Business Traveller on Twitter for the latest updates for business travellers: we're @AusBT.