Monday, September 29, 2008
Your Numbers Up!
After taking an interest in Hadyn Greens post over at Public Address on the unpredictability of various professional sports I thought I'd check it out for the 2008 NRL season using roughly the same formula. I couldn't quite get it exactly right because I'm just not that smart but it's near enough.
According to this study done by Naim, Vazquez and Redner from Boston University of the EPL (or FA rather), MLB, NHL, NBA and NFL the number of upsets by percentage is
FA: 45.2%
MLB: 44.1%
NHL: 41.4%
NBA: 36.5%
NFL: 36.4%
There are some very clear trends (you can see them on the graphs on the above link) in each of the competitions since about 1980 which I guess is the dawn of crazy (read: mo' money) professionalism. NFL, NHL and MLB are becoming more unpredictable with more upsets while the Premier League and NBA are becoming more predictable with fewer upsets. The FA probably largely because of the obscene money flying around at the top end of it in the last two decades and increasingly since 2000 which has killed it as a 'contest' while in the NBA they have this 'interesting' thing where top players sign with sides with a good chance of winning them a championship ring thus concentrating the best players on a fewer number of teams. What would be nice in football is if the Saudi Arabian league takes off and they sign every single top player so the EPL chills the fuck out a bit.
So anyway, the results for this seasons NRL are in and they say... 39.2%
Which is a bit lower than I thought they would be but there's the entire possibility that I fucked it up. In fact I know I did because there were 4 teams at the end of the year that I had the incorrect points tally for (2 points too many for each of the 4 teams) which means somewhere going through the year calculating the weekly tables I ballsed it up which may have messed up a couple of results. Either way though the 39.2% figure will most likely be less than a percent out (margin of error +/- 0.6%) which would still put it after FA, MLB, NHL but ahead of NBA and NFL.
I discounted the one and only draw during the year between the Broncos and Panthers even though I know they count them. Apparently it makes jack all difference though.
There were 192 games played, with 58 results where the lower ranked team won and 90 where the higher ranked team won with 44 matches played where both teams were on the same number of points and so no favourite tags were handed out.
It's interesting to see a bit of a pattern. There tended to be 3-5 upsets one week and then only 1 or 2 the next. I'd suggest this is because the higher ranked teams refocus after seeing a lot of upsets in a round. Also in the last 3 weeks of the comp this year there were 19 results that went to the higher ranked team and only 2 upsets on paper. The old motivation factor coming into play there. After a few weeks there had been more upsets than favourites winning but that gradually crept down all year until those final weeks. If they shortened the season to say 20 games then the figure would probably be around 42-43%.
On this weekends game I think I'm edging towards the Sea Eagles. There's the old cliche about needing to lose one to win one. It doesn't always hold true. The Broncos have played 6 and won 6. The Storm have played 2 and won 2. But there does seem to be a kind of deathly serious focus about the Sea Eagles and a no prisoners attitude. It should be an excellent final and could go either way based on a refs call, a bit of luck or the proverbial moment of brilliance but I think the Sea Eagles have probably got a touch more desparation. Although that might be countered by the siege mentality running through the Storm who have been there, done that and would like nothing more than to ram rod it back up.... well everybody, actually.
According to this study done by Naim, Vazquez and Redner from Boston University of the EPL (or FA rather), MLB, NHL, NBA and NFL the number of upsets by percentage is
FA: 45.2%
MLB: 44.1%
NHL: 41.4%
NBA: 36.5%
NFL: 36.4%
There are some very clear trends (you can see them on the graphs on the above link) in each of the competitions since about 1980 which I guess is the dawn of crazy (read: mo' money) professionalism. NFL, NHL and MLB are becoming more unpredictable with more upsets while the Premier League and NBA are becoming more predictable with fewer upsets. The FA probably largely because of the obscene money flying around at the top end of it in the last two decades and increasingly since 2000 which has killed it as a 'contest' while in the NBA they have this 'interesting' thing where top players sign with sides with a good chance of winning them a championship ring thus concentrating the best players on a fewer number of teams. What would be nice in football is if the Saudi Arabian league takes off and they sign every single top player so the EPL chills the fuck out a bit.
So anyway, the results for this seasons NRL are in and they say... 39.2%
Which is a bit lower than I thought they would be but there's the entire possibility that I fucked it up. In fact I know I did because there were 4 teams at the end of the year that I had the incorrect points tally for (2 points too many for each of the 4 teams) which means somewhere going through the year calculating the weekly tables I ballsed it up which may have messed up a couple of results. Either way though the 39.2% figure will most likely be less than a percent out (margin of error +/- 0.6%) which would still put it after FA, MLB, NHL but ahead of NBA and NFL.
I discounted the one and only draw during the year between the Broncos and Panthers even though I know they count them. Apparently it makes jack all difference though.
There were 192 games played, with 58 results where the lower ranked team won and 90 where the higher ranked team won with 44 matches played where both teams were on the same number of points and so no favourite tags were handed out.
It's interesting to see a bit of a pattern. There tended to be 3-5 upsets one week and then only 1 or 2 the next. I'd suggest this is because the higher ranked teams refocus after seeing a lot of upsets in a round. Also in the last 3 weeks of the comp this year there were 19 results that went to the higher ranked team and only 2 upsets on paper. The old motivation factor coming into play there. After a few weeks there had been more upsets than favourites winning but that gradually crept down all year until those final weeks. If they shortened the season to say 20 games then the figure would probably be around 42-43%.
On this weekends game I think I'm edging towards the Sea Eagles. There's the old cliche about needing to lose one to win one. It doesn't always hold true. The Broncos have played 6 and won 6. The Storm have played 2 and won 2. But there does seem to be a kind of deathly serious focus about the Sea Eagles and a no prisoners attitude. It should be an excellent final and could go either way based on a refs call, a bit of luck or the proverbial moment of brilliance but I think the Sea Eagles have probably got a touch more desparation. Although that might be countered by the siege mentality running through the Storm who have been there, done that and would like nothing more than to ram rod it back up.... well everybody, actually.
Labels: nrl, nrl grand final
Comments:
Disclaimer: I am biased.
But I do seriously think that being underdogs will actually suit the Storm.
Think what you like about Wing Commander Bellamy, but he circled the wagons superbly last week.
But I do seriously think that being underdogs will actually suit the Storm.
Think what you like about Wing Commander Bellamy, but he circled the wagons superbly last week.
# posted by 
: 30 September 2008 at 20:33
: 30 September 2008 at 20:33
Post a Comment