Some people no doubt already know this, but in American sport one way pundits try to predict future performance is by looking at close game data. The theory goes that any very close game is more or less a coin toss - evidence suggests that there isn’t a specific skill to winning close games. So what they do is look for teams that have won or lost a disproportionate number of the close games they’ve played. Teams that win a lot of their close games are likely (all things being equal) to be worse in future - basically, because their luck will run out - and vice versa.
So I’ve done the same thing with the last three seasons of NRL games, and actually it’s very interesting. Before I reveal the results, some notes on the data:
- I considered a close game to be any one settled by six points or fewer. There are no qualitative assessments: if team A was winning by 18 points and team B jagged a couple of tries at the death to make the final margin six, it went down as a close game. If team A was desperately defending a six point lead and grabbed an intercept try on the siren to make it a 10 point win, it didn’t go down as a close game. You have to be fair to the data.
- The data over three years suggest that it is true that teams do not control their fate in close games. There is a lot of variance in the numbers but no team consistently out- or underperformed zero. The Storm, for instance (who I bet most people would expect to be consistently good at winning the close ones) actually ended up winning three more close games than they lost over three seasons - and year by year their outcomes were -2 wins, zero and plus five wins.
- I’m only going to look at teams that really stuck out statistically in each season (plus the Tigers).
Outperformers: Manly (plus four close wins), Penrith (plus four).
Underperformers: Cowboys (minus four), Souths (minus four)
Three of these four did what they ought to in 2015.
- Manly dropped from second in 2014 to ninth in 2015, in part because the eight points they won ‘net’ in close games (won six, lost two) became zeo in 2015 (won four, lost four).
- Penrith dropped from fourth to 11th, going from 8-4 in close games to 4-6.
- The Cowboys climbed from fifth to third, going from 3-7 in close games to 7-2. They also won the premiership in 2015, of course - although I didn’t include finals games.
- Souths were the outlier, as they finished third in 2014 despite a 1-5 close game record, and seventh in 2015 despite a 5-1 close game record. Obviously other factors at play - roster change etc.
Outperformers: Cowboys (plus five), Souths (plus four).
Underperformer: Raiders (minus four).
All three followed the trend from 2015 to 2016.
- Cowboys only dropped from third to fourth despite going from 7-2 in close games to 4-5, so a bit hard to call a big win for the theory here.
- Souths dropped from seventh to 12th, going from 5-1 in close games to 3-4 in the process.
- Raiders improved from 10th in 2015 (2-6 in close games) to 2nd in 2016 (3-3).
Outperformers: Bulldogs (plus four), Dragons (plus four), Sharks (plus four), Storm (plus five).
Underperformers: Knights (minus four), Roosters (minus eight).
Thoughts on what to watch for in 2017:
- The Roosters should be a lot better. Losing eight more close games than you win (0-8) is unique for a team in the three years I looked at the data - and by some distance: no-one else did worse than minus four.
- As for the Knights, well, it’s hard to only win one game in a season without having some close ones. I guess you could say they were unlucky not to jag another win or two (they also had a draw, which I didn’t include in my data), but of course they’d still have been spoonists even with a handful of extra points.
- If you had to pick three teams to drop next year I reckon the Bulldogs, Dragons and Sharks would be popular picks - and the fact that all three were pretty lucky in close games seems to back that up.
- The Storm probably outperformed expectations this year (I don’t remember a lot of people picking them for minor premiers). They’re obviously a good side but they also caught a fair tailwind this year.
First of all, in no season were the Tigers’ close game results lopsided enough to draw massive conclusions. They went plus thee in 2014 (4-1) and finished 13th, minus one in 2015 (2-3) and finished 15th, and minus three this year (1-4) for ninth. The good news is that the trend was still followed (a bit lucky in 2014, then a worse ladder position in 2015) - so with ‘normal’ luck next year they should be a couple of ladder points better off.
Of course, there’s nothing to say they won’t be unlucky again in 2017, or that other factors (like not signing any good players!) will overwhelm mean reversion on close game outcomes.
I can give info for any team for any of these three seasons - I’ve highlighted the outliers for each year. The best performer of all over the three year span was the Bulldogs (+9, based on +3, +2 and +4), and the worst was the Roosters (-1, -1, -8).
Well, hope that’s of interest to someone. Just an idle lunch hour…