Betting Fallacy 2
FALLACY TWO - THEY'RE ON A GOOD RUN, SO THEY'RE GETTING BETTER/THEY'RE ON A BAD RUN, SO THEY'RE GETTING WORSE
It isn't always true - certainly, not as often as many people think.
What would you make of a team who had had not lost any of their last eight games? You would probably think they were pretty good. What would you make of a team who had not won any of their last eight games? You would probably think they were pretty bad.
Actually, a perfectly ordinary mid-table team are likely to run up both of these sequences at least once in a season.
A mid-table team will usually win 36 per cent of their games, draw 28 per cent and lose 36 per cent. In other words, they will avoid defeat in 64 per cent of the games they play. The chance of them avoiding defeat in eight consecutive games is one in 36. Likewise, the chance of them failing to secure a single victory in eight consecutive games is also one in 36.
You might not be convinced. I wasn't, either, when I first did the calculations. So I got out some old Rothmans annuals and flicked through the pages of teams who had finished halfway down their divisions. And what I found was that each team I checked did indeed run up at least one sequence of eight or so games without a defeat and another of eight or so games without a victory.
You can be sure that toward the end of each of these good runs bookmakers and punters were beginning to think that the team had got better, and toward the end of each of these bad runs they were beginning to think that the team had got worse.
In fact, neither conclusion was true.
On average, the team who finish top of the Premiership fail to win 37 per cent of the games they play. Even Arsenal, when they went unbeaten throughout season 2003-04, failed to win 32 per cent of the games they played.
In other words, the best team in England will usually fail to win approximately one out of every three games they play. Yet when these draws and defeats occur, the betting markets, just like the sillier newspapers, can overreact.
There is rarely a value-for-money bet on the Premiership before it begins. Sometimes there is once the season has got under way - and when there is, it is almost always on the challenger who has just failed to win one or two games.
As a sequence of good results gets longer, we should progressively upgrade our assessment of the team. And as a sequence of bad results gets longer, we should progressively downgrade our assessment of the team. But not as dramatically as most people do.
Teams in the Premiership and Football League who won six games in a row - many of whom would have been genuinely good teams if assessed on their record over a longer period of time - fewer than half won the next game.
And similarly, of all the teams who lost six games in a row - many of whom would have been genuinely bad teams if assessed on their record over a longer period of time -fewer than half lost the next game.
Obviously, some teams do change significantly during a season. I suggest that if a team start to register results that appear uncharacteristically good or bad you ask yourself whether anything has changed at the club. Have they, for example, changed the manager - and if so, is there any reason for thinking the new one will be any better or worse than the old one? (See, also, Fallacy Five).
Have they brought in some new players or shipped out some old ones - and if so, is there any reason for thinking the team will now be any better or worse than it was before? Was there any reason for thinking the team were under-performing or over-performing in the past - and if so, why is that reason no longer valid?
After three games without a win - two against Manchester United, who went on to finish third in the Premiership, and one against Chelsea, who went on to finish second - Arsenal were quoted at evens to win at home to Liverpool.
This was Good Friday 2004. Arsenal were still top of the league and unbeaten in the Premiership. All you needed for that price to represent value for money was for Arsenal to turn in what, for them, had been a typical performance during that season. And they did.