It was the biggest match of the season so far, at least at the top of the table: Liverpool, leading the Premier League by seven points, traveled to second-place Manchester City in hopes of ending the title race early. At least, that was the narrative coming from the media. Manchester City were narrowly victorious, but does this mean the title race is back on?
Before the match, the bookies gave Liverpool roughly a 60-65% chance of winning the league, with Manchester City at about 30-35% and Tottenham around 5%. It’s worth pausing for a moment to consider what that meant. Since the Premier League shrank to 20 clubs in 1995-96, only five other clubs ever had a seven-point lead or better after 20 matches: Newcastle United in 1995-96, Manchester United in 2000-01, Chelsea in 2005-06, Manchester United in 2012-13, and Manchester City last season. All of them took home the title except my beloved Newcastle. If we indulge in a little statistical barbarism and say the baseline chance of a club with a seven-point lead after 20 matches going on to win the title is 80%, then the bookies may have thought Manchester City was an especially strong contender this season.*
Indeed, going into the match, the bookies had the odds of a Manchester City win at about 50%, with 25% each for the draw and a Liverpool win. In the Premier League so far this season, clubs have won about 44% of their home matches. So the idea that Manchester City had a higher-than-average chance of winning at home against the only club above them in the table was striking to say the least. Liverpool may have been leading the table by seven points, but the bookies’ models – yes, legend has it they use models, too – may have made Manchester City the stronger club.
In the end, Manchester City won 2-1, thus grabbing about 1.25 points more than the bookies expected (three points versus 50%*3 + 25%*1 + 25%*0 = 1.75 points). And now, the bookies have Liverpool around 50-55% for the title and Manchester City at 40-45%. Tottenham’s odds are basically unchanged. So there’s been a considerable swing from Liverpool directly to Manchester City, and just from an extra 1.25 points relative to expectations.
What else could have caused such a swing? If the bookies’ models were based on measures of underlying performance like expected goals, did Liverpool actually perform much worse – or Manchester City much better – than the scoreline suggested? According to at least one notable xG jockey, the answer was no; Liverpool actually edged the match by his measure. If anything, the supposed strength of Manchester City’s squad failed to manifest itself as much as a model might have predicted.
More likely, the betting markets have a bit of froth in them today as punters process their excitement after a pretty dazzling game of football. I wouldn’t be surprised if the odds shifted back in Liverpool’s favor before the next match in the league, which is still eight days away. With only one match away remaining against the top six, the title is still theirs to lose.
* Either that, or Liverpool supporters were betting like the blazes on their own club, since bookies’ odds sometimes have to respond to supply and demand as well.