Tuesday, 5 April 2011

Are NSW players over-represented in the Australian cricket team?


In every sports competition in the world, fans of one team will claim another team gets more favourable treatment than their own, whether it’s selection in representative teams, concessions regarding player salaries or favourable refereeing decisions. Cricket in Australia is no different. The dawn of summer is almost inevitably accompanied by bleating, generally from Victorian fans, about how players from New South Wales are more likely to be selected in the national team than players from other states. This opinion found a voice in David Hookes, who claimed:

"When they give out the baggy blue cap in New South Wales, they give you a baggy green one in a brown paper bag as well to save making two presentations."

It’s about time we put this idea to the test. Since the 1977/78 season, all 6 Australian states have played in the Sheffield Shield, Australia’s domestic first-class cricket competition. Since 1982/83, the season has culminated in a grand final (previous to this the winner was determined league style by whoever won the most throughout the season). I will use data from 1982/83 till now for consistency and so that all states are represented. The following table shows the players who have debuted for Australia since the start of the 1982/83 season, the state they were playing for at the time of their debut, and the number of Tests they played throughout their career.

PlayerYearsTestsStatePlayerYearsTestsState
Carl Rackemann 1982–9112QldMatthew Elliott 1996–200421Vic
Kepler Wessels 1982–8524QldMichael Kasprowicz 1996–200638Qld
Tom Hogan 1983–847WAJason Gillespie 1996–200671SA
Roger Woolley 1983–842TasAndy Bichel 1997–200319Qld
Wayne B. Phillips 1983–8627SAShaun Young 19971Tas
John Maguire 1983–843QldSimon Cook 19972NSW
Greg Matthews 1983–9333NSWStuart MacGill 1998–200844NSW
Steve Smith 19843NSWGavin Robertson 19984NSW
Dean Jones 1984–9252VicPaul Wilson 19981SA
David Boon 1984–96107TasAdam Dale 1998–992Qld
Bob Holland 1984–8611NSWDarren Lehmann 1998–200427SA
Murray Bennett 1984–853NSWColin Miller 1998–200118Tas
Craig McDermott 1984–9671QldMatthew Nicholson 19981WA
Simon O'Donnell 19856VicAdam Gilchrist 1999–200896WA
Dave Gilbert 1985–869NSWScott Muller 19992Qld
Robbie Kerr 19852QldBrett Lee 1999–201076NSW
Merv Hughes 1985–9453VicSimon Katich 2001–56WA
Geoff Marsh 1985–9250WAMartin Love 2002–035Qld
Bruce Reid 1985–9227WABrad Williams 2003–044WA
Steve Waugh 1985–2004168NSWNathan Bracken 2003–055NSW
Simon Davis 19861VicAndrew Symonds 2004–0926Qld
Tim Zoehrer 1986–8710WAMichael Clarke 2004–67NSW
Chris Matthews 1986–883WANathan Hauritz 2004–17Qld
Greg Dyer 1986–886NSWShane Watson 2005–25Qld
Peter Taylor 1987–9113NSWShaun Tait 2005–083SA
Mike Veletta 1987–908WAMichael Hussey 2005–57WA
Tim May 1987–9524SABrad Hodge 2005-086Vic
Tony Dodemaide 1987–9210VicPhil Jaques 2005–0811NSW
Ian Healy 1988–99119QldStuart Clark 2006–0924NSW
Trevor Hohns 19897QldDan Cullen 20061SA
Mark Taylor 1989–99104NSWMitchell Johnson 2007–40Qld
Greg Campbell 1989–904TasChris Rogers 20081WA
Tom Moody 1989–928WABrad Haddin 2008–30NSW
Mark Waugh 1991–2002128NSWBeau Casson 20081NSW
Shane Warne 1992–2007145VicCameron White 2008–094Vic
Wayne N. Phillips 19921VicPeter Siddle 2008–20Vic
Paul Reiffel 1992–9835VicJason Krejza 20082Tas
Damien Martyn 1992–200667WADoug Bollinger 2009–12NSW
Justin Langer 1993–2007105WAAndrew McDonald 20094Vic
Jo Angel 1993–954WABen Hilfenhaus 2009–15Tas
Michael Slater 1993–200174NSWPhillip Hughes 2009–8NSW
Brendon Julian 1993–957WAMarcus North 2009–21WA
Glenn McGrath 1993–2007124NSWBryce McGain 20091Vic
Matthew Hayden 1994–2009103QldGraham Manou 20091SA
Michael Bevan 1994–9818NSWClint McKay 20091Vic
Damien Fleming 1994–200120VicRyan Harris 2010–4SA
Phil Emery 19941NSWTim Paine 20104Tas
Greg Blewett 1995–200046SASteven Smith 2010–3NSW
Peter McIntyre 1995–962SAPeter George 2010–1SA
Stuart Law 19951QldXavier Doherty 20102Tas
Ricky Ponting 1995–151TasMichael Beer 2011–1WA
Brad Hogg 1996–20087WAUsman Khawaja 2011–1NSW

Of the 104 players who debuted after 1982/83, 28 were playing for NSW when they were first picked for Australia (27% of the new players).

PlayersPercentage
NSW2827%
WA2019%
Qld1817%
Vic1615%
Tas1010%
SA1212%

Clearly NSW players have played more Tests. But is this unreasonable? There are two measures we can look at here. Sheffield Shield results and state populations. The following shows the results of the Shield since 1982/83.

Season Winner Second Third Fourth Fifth SixthSeason Winner Second Third Fourth Fifth Sixth
1982–83 NSWWASATasQldVic1997–98 WATasQldNSWVicSA
1983–84 WAQldTasNSWSAVic1998–99 WAQldVicSATasNSW
1984–85 NSWQldSAWAVicTas1999-00 QldVicWASATasNSW
1985–86 NSWQldVicWASATas2000–01 QldVicNSWTasWASA
1986–87 WAVicQldSANSWTas2001–02 QldTasWASAVicNSW
1987–88 WAQldNSWVicSATas2002–03 NSWQldVicSAWATas
1988–89 WASAQldNSWTasVic2003–04 VicQldTasWANSWSA
1989–90 NSWQldSATasWAVic2004–05 NSWQldWAVicSATas
1990–91 VicNSWQldWASATas2005–06 QldVicSATasWANSW
1991–92 WANSWVicQldSATas2006–07 TasNSWVicQldWASA
1992–93 NSWQldWASATasVic2007–08 NSWVicWATasSAQld
1993–94 NSWTasWAVicSAQld2008–09 VicQldSATasWANSW
1994–95 QldSAVicWANSWTas2009–10 VicQldNSWWATasSA
1995–96 SAWAQldTasNSWVic2001–11 TasNSWQldWAVicSA
1996–97 QldWANSWTasVicSA

NSW has won 9 Shields, ahead of WA’s 7. If you look at the percentage of Shield wins per state and compare this to the number of players picked for Australia from that state, you will notice that these results are remarkably similar. Is it any surprise that the most successful team over this time has more players picked for representative honours? If you look at all the Tests played since 82/83, 34% of the Test positions up for grabs were occupied by NSW players, and the order of states is exactly the same as the order for number of Shield wins. This is strong evidence that rather than a selection bias, players are being picked either because they are the best players or because they have been a part of successful teams.

PlayersTestsShield wins
NSW2827%98334%931%
WA2019%54018%724%
Qld1817%51618%621%
Vic1615%38013%414%
Tas1010%30610%27%
SA1212%2087%13%

The second way to look at this is by state population. I don’t particularly like this method because sportsmen, especially in this professional age, move teams for many reasons, including for better opportunities and more pay, and don’t necessarily play for the state of their birth. The most populous state is not necessarily going to have the best team. But because this is often the first measure people look to when analysing team results (for example, Olympic results), and because arguably a larger population means a bigger economy and therefore more money flowing through the team, we shall include some analysis. Since 1982, NSW (+ACT) has averaged 36% of the total population of the states (excluding Northern Territory). Again, this matches quite closely the number of players picked for Australia – indeed, you might argue that more players from NSW should have been picked. It would be by this measure that Victorians may claim some bias.

PopulationShield Wins
NSW + ACT36%31%
Vic25%14%
Qld18%21%
WA10%24%
Tas3%7%
SA8%3%

Finally, let’s look at the one-test wonders – the players who only played one test.

One Test wonders
NSW3
WA3
Qld1
Vic4
Tas1
SA4

Victoria and SA share this honour with 4 one-test wonders – and this is possibly why David Hookes, of SA stock and coach of Victoria after his retirement, was cranky. In the case of SA, this is a third of their players who have made their debut since 1982/83 – indeed, 7 of their 12 Test players did not play more than 4 Tests. Tasmanian players also fair poorly in this regard, with 6 of their 10 players playing no more than 4 Tests – Tasmanian results are severely skewed by Ponting’s 151 and Boon’s 107 Tests.

No matter what state a Test player is from, he has roughly the same chance of playing 15 or more Tests. Thus the data does not support the idea that selectors are more likely to stick with NSW players through a patch of poor form.

Percentage of state Test players (15+ Tests)
NSW43%
WA40%
Qld56%
Vic44%
Tas40%
SA42%

I would be interested to hear your thoughts regarding this – I know it can stir some passions. But the data would suggest that there is no unfair bias towards NSW.

Further thoughts:
  • I have made no attempt to look at players moving state during their careers. Simon Katich, for instance, had a career revival when playing for NSW, even though he was originally picked for Australia when playing for WA. Nathan Hauritz similarly had his fortunes revived when playing for NSW, even though he debuted when playing for Queensland. There are also many examples of this working the other way – Jason Krejza moved to Tasmania, and Adam Gilchrist to WA, both from NSW and then made their debut. Even Don Bradman first played for NSW before an ongoing career with SA.
  • You could do the same analysis if you have time and patience on your hands for one-day and Twenty20 cricket.
  • It was quite difficult in some cases to track which state a player was playing for when they made their Australian debut, especially if they moved state soon after. I have put the data here (some of the array formulae don't work in Google docs, and I have stripped out the macros, but you can redo them). If I have made a mistake, let me know!

4 comments:

  1. What rubbish, what about adjusting all the NSW players who went interstate to then get in the test team as well ? Gilchrist et al ? the distortion becomes stronger again. Talk about only trust statistics you have manipulated yourself ! The NSW CA is a joke and everyone knows it.

    ReplyDelete
    Replies
    1. you think the selectors are so biased they pick on place of birth?

      Delete
  2. Suggest also analysis be done of of squads chosen for overseas tours, Australia A one day sides. Also look at replacements flown in at short notice to fill in for injured players during overseas tours (Phil Emery being chosen to deputise for Healy in Pakistan in the 90s being a glaring example). The Australian selectors choose many teams, not just the starting 11 in test matches: broaden the analysis to include the full range of teams/squads, target it to the seasons between the retirements of Lillee/Marsh/Chappell and the retirements of McGrath/Warne/Gilchrist and the bias I am sure will be statistically proven. I suspect that bias has reduced in recent years - it more or less had to, since it become so embarrassingly obvious. The bias is inarguable - the explanation for it (given that the selectors themselves have typically been from states other than NSW) is actually the more interesting question. And one I don't have an answer for.

    ReplyDelete