By: Chinmay Vaidya
We've heard the term "regression to the mean" used in sports. It's similar to the law of averages in a way and basically follows the thought process that eventually, every team's performance averages out. For example, if a team shoots 70 percent from the floor in the first half of a basketball game, regression to the mean would suggest they would shoot significantly worse in the second half to bring their overall percentage for the game to around their season average.
With 11 seasons of IPL data at our disposal, we can make accurate conjectures on whether teams eventually regress to the mean over the long run.
I started by looking at the finishing position of each team over the 11 editions of the tournament. Some of this data is skewed because certain IPL years had more teams than others. Some franchises became defunct. Others were suspended for two seasons. For the sake of simplicity, I looked specifically at the eight franchises that will be competing in IPL 2019. I assigned Rising Pune Supergiant's two seasons to the Chennai Super Kings and Gujarat Lions' two seasons to Rajasthan Royals to account for the suspensions. Since Sunrisers Hyderabad started the year after Deccan Chargers went defunct, they absorbed Deccan's earlier finishes. Delhi, unfortunately, got stuck with a 10th place finish in the lone season where the IPL had 10 teams.
Another hiccup was the changing playoff format. There was initially two semifinals and then a final, meaning fourth place wasn't assigned for the first two seasons. To keep things fair, I assigned third place finishes to both losing teams from the semifinals in the first two seasons. Here are the average finishing positions for each franchise in the IPL.
Chennai Super Kings - 2.36
Mumbai Indians - 3.45
Kolkata Knight Riders - 4.36
Royal Challengers Bangalore - 4.54
Sunrisers Hyderabad - 4.63
Rajasthan Royals - 4.81
Kings XI Punjab - 5.72
Delhi Capitals/Daredevils - 6.18
Obviously, winning titles and getting a first-place finish helps. The top three teams have all won multiple titles. CSK has only missed the playoffs once, which came under the Rising Pune Supergiant name. However, it does look like each team averages out its finishes in an eight-team league. The talent distribution is fairly even and thus, any team could win the competition in any given season. The average finishes reflect this fact, given that five of the eight teams average a finish near 4.5, which is the exact middle of the IPL standings. CSK is boosted by its long run of success, but the other teams are fairly consistent in eventually regressing back to the mean. But does winning a title change things?
For those unfamiliar with the IPL format, there's an auction held every offseason where franchises bid on players who need new contracts. There's occasionally an auction where a massive amount of players become free agents, but usually there's only two or three big names available. This means good teams tend to retain most of their key players for longer times, so shouldn't they consistently be good?
To account for this, I looked at the title team's finishing position in the season before it won the title and the season after it won the title and calculated the average. This would show year-to-year regression to the mean for title teams, even though the roster would be fairly similar. This data is somewhat incomplete because we don't know how the defending champions Chennai will finish in IPL 2019. The Royals won the first IPL tournament and thus have no finishing position before they won the title. Here are those results.
Average Standings Position Before Title Season: 4.5
Average Standings Position After Title Season: 4.3
This shows teams fluctuate significantly year to year, creating the excitement for IPL every season. A title team, with essentially the same roster, was in the middle of the pack on average the year before capturing the trophy. In the same way, a title team reverted back to league-average levels the season after winning it all. This isn't exactly great news for Punjab, Bangalore and Delhi fans given the franchise's respective struggles, but the next set of data is what brings fans back to the IPL every single year. There's a reason this tournament is the most successful T20 league in the world.
I tracked how many new playoff teams entered the fray from one season to the next. The early results were somewhat disappointing as three of the first four seasons featured only one new team entering the playoff stage. However, the competition has become much closer in the last six seasons with at least two new playoff teams each year.
The average number of new teams entering the playoff stage is 1.8, which we can round up to 2. That's half the playoff field changing every year, on average. With only eight teams in the league, eventually every club is going to regress to the mean and find itself in the postseason.