We are on the cusp of the start of the 2021 Six Nations tournament with the opening round kicking off this upcoming weekend. With all teams dealing with the ongoing challenges associated with the global pandemic along with injuries to/absences of key personnel, this year’s tournament could be completely up for grabs. Given how tight things could potentially be for all teams, every little edge gained will be huge. Knowing something like where an opponent might be vulnerable defensively could be valuable to some teams – especially if they aren’t aware that they’re vulnerable?
In a previous post we determined that within the Super Rugby competition, MTI (Missed Tackle Impact) is a defensive metric that is a stronger correlate to winning compared to other traditional statistics such as missed tackles and tackle completion percentage. This post can be found here. Will this same trend for MTI translate to another competition like the Six Nations? There are a couple of reasons why MTI might not measure defensive performance as well:
The style of play in Super Rugby is very different than they style adopted by teams in the Northern Hemisphere.
There are only 15 games played in each Six Nations tournament. Thus, MTI will be calculated from a small data set so it could be noisy and thus potentially less predictive.
RUGBY NERD ALERT:
I am about to go into some superficial detail about statistics along with some methodological adjustments that were made for this study. If this is not your thing, you can jump directly to the pretty pictures towards the end of this post.
For this study, we looked at all missed tackles which occurred in the Six Nations tournament from 2014-2020. From this data, MTI, missed tackle totals, as well as tackle completion percentage were calculated. In the Super Rugby study, we looked at the correlation of MTI to winning. However, this is problematic when looking at Six Nations, because the teams play so few games. For example, if we look at all of the teams that have won 3 games over the past 7 tournaments (Table 1), when we compare the “points for” and “points conceded” for all of these teams, we see that they are all very different from each other even though they won the same number of games.
TABLE 1: Points For and Points Conceded for all teams that won 3 games in the Six Nations tournament (2014-2020)
Thus, instead of looking for correlations to winning, in this case it would be better to look at a measure that better captures discernable differences in team performance for these teams. Since MTI is a defensive metric, points conceded seems like a logical candidate with which to compare. Table 2 summarizes the findings of comparing MTI, missed tackles, and tackle completion percentage to points conceded.
TABLE 2: Correlation coefficient (r) and Coefficient of Determination (r2) of MTI, missed tackles, and tackle completion percentage when compared to points conceded for all teams in the Six Nations tournament (2014-2020)
As a refresher, the linear correlation coefficient (r) measures the strength and direction of the linear relationship between the tackle metric and points conceded. MTI has a very strong negative relationship with point conceded. This means that teams that have a more negative MTI score will tend to concede more points. Missed tackles has a moderate positive relationship with points conceded, and tackle completion percentage has a moderate negative relationship.
The coefficient of determination (r2) is a measure of how much the variance of points conceded can be predicted from the tackle metrics. Thus, this is a measure of how well the regression line represents the data. Based on this, we can determine that ~68% of the variation in points conceded can explained by changes in MTI. While it isn’t perfect, MTI is a much stronger predictor of points conceded compared to missed tackles (r2 = 0.2238) and tackle completion percentage (r2 = 1801).
NOTE: ANTI NERDS CAN RE-ENGAGE WITH THE CONTENT AT THIS POINT
When we graph the relationship between MTI/Missed Tackles/Tackle% versus points conceded, we get a better visual representation of the strength of the relationship between these variables. In Figure 1 we see the relationship between MTI and points conceded. We already know it has the strongest correlation and you can see that the data points are more tightly clustered around the line of the graph. When we compare that to the scatter of points for missed tackles (Figure 2) and tackle completion percentage (Figure 3) we see that the data is more widely scattered around the line.
FIGURE 1: Scatter plot of Points Conceded versus MTI in the Six Nations tournament (2014-2020). Coefficient of determination (r2) = 0.6781.
FIGURE 2: Scatter plot of Points Conceded versus Missed Tackles in the Six Nations tournament (2014-2020). Coefficient of determination (r2) = 0.2238.
TABLE 3: Scatter plot of Points Conceded versus Missed Tackles in the Six Nations tournament (2014-2020). Coefficient of determination (r2) = 0.1801.
CONCLUSIONS:
Missed Tackle Impact (MTI) can be used as a measure of team performance in the Six Nations competition. Since the data set is small within a given competition year (15 games), one should exercise caution when trying to identify trends in team performance - especially in the earlier rounds of the competition. Even though the MTI data may be noisy after a few rounds, it is still a better predictor of points conceded when compared to missed tackles and tackle completion percentage. The results in this study suggest that 5 rounds of competition in the Six Nations tournament is sufficient to evaluate defensive performance using the MTI metric.
Do you think MTI would be an effective metric of defensive performance in your favourite rugby competition?