In our first article using this new technique, we are trying to find a shortlist of teams who we think will be a good bet to win automatic promotion from League Two at the end of the 2012/13 season.
To do this, we'll first of all look at the 18 teams who will compete in League Two this coming season who also competed in League Two in the season just gone. In the last 4 seasons, 8 of the 12 teams who finished in the top 3 of League Two had also competed in League Two the previous season. One of these teams, Notts County, had a major transformation in between seasons, so we'll only look at the 7 remaining teams to see if there are any trends we can apply.
One thing we can find out fairly quickly is that all 7 teams finished their previous season with at least 59 points. This means that we can remove all of the teams from Bristol Rovers (57 points) and below from our list of 18 teams. Already we are left with just 8 teams.
Another trend is that the worst average performance rating among our 7 marker teams for their previous season was -0.08 (Brentford). Torquay's average rating in 2011/12 was significantly poorer than this (-0.44) , so we'll remove them from our list too.
When looking at the second half of the previous season for our 7 marker teams, all 7 managed at least 32 points. Last season, Oxford and Cheltenham only managed 31 points in the second half of their league campaign, so although they are just one point short, we need to get our shortlist down further, so both are struck off.
We'll now look at the previous season for our 7 marker teams, but this time we'll ignore games at the start and end of a season. The reason for this is that performances at the start of a season can be misleading when teams are adjusting to major changes (promotion, relegation, new manager, new players, etc). Performances at the end of a season can be misleading too as some teams may have little to play for and may rest players, while others may need to win to obtain a play-off place or to avoid relegation. So, by looking at our marker teams' performances from 1st October to 31st March in their previous seasons we can see that all 7 marker teams managed a points per game figure of at least 1.31. From 1st October to 31st March in 2011/12, Rotherham's equivalent figure was just 1.27, so they are removed from our shortlist. We have now cut our shortlist down to just Southend, Port Vale, Aldershot and Gillingham.
We'll now take a look at the 4 teams that will compete in League Two this season who were relegated from League One last season. In the last 4 seasons, only two teams - Swindon (in 2011/12) and Wycombe (in 2010/11) - finished in the top 3 of League Two after being relegated the previous season, so we don't have a lot to go on. However, in the second half of their relegation seasons, both teams managed a positive average performance rating. All 4 teams relegated from League One in 2011/12 had average performance ratings less than -1, so we'll remove all 4 from our shortlist.
Finally, we now need to take a look at the 2 teams who were promoted from the Conference - Fleetwood Town and York. In the last 4 seasons, only Crawley (in 2011/12) and Exeter (in 2008/09) managed a top 3 finish in League Two after being promoted from the Conference the previous season, so again we don't have a lot to go on. While Crawley had very impressive stats when they won the Conference (105 points and an average performance rating of 3.78), Exeter's stats were fairly average, and were worse than those of Fleetwood and York this season. We'll therefore keep both these teams in our shortlist.
So, our final shortlist for teams who think have a good chance of finishing in the top 3 of League Two in 2012/13 are as follows...
Due to their financial problems, Port Vale may be less likely than the other teams on this list to win automatic promotion, however, if the trends we look at are to continue, then there may be some value to be had in the other teams. In particular, we think there may be some value in an each way bet on Gillingham, who are currently best priced at 25/1.
As we've mentioned before, we haven't had a lot of data to go on to come up with this analysis. Also, this is still an experimental technique we are using and has not been back-tested. For these reasons we don't recommend using this analysis as the basis for any serious bets.