Last Sunday, Roda announced the firing of manager Ruud Brood. After starting the season with seven points from four matches, only two wins from thirteen followed, and Roda now hold the 14th place with 18 points from 17 matches. Saturday’s 3-4 loss at the hands of N.E.C. Nijmegen, who are currently bottom of the table, was the trigger to fire manager Ruud Brood. Indeed, results do not look pretty, do they? But here are three simple graphs to illustrate why Brood’s firing seems a weird decision.
TSR and PPG
Regular visitors of 11tegen11 know my fancy of a stat called Total Shot Rate, adapted in football through the work of the brilliant James Grayson, and originally coming from ice hockey analysis. TSR is simply the fraction of shots a team takes in a match, computed by diving shots for by total shots. TSR is an excellent marker for team performance, as the next graph shows.
This graph shows Roda’s points-per-game (PPG) and TSR in 34-match moving averages. This simply means that each point in the graph represents the average of the preceding 34 matches. The two lines make an almost perfect dance, always moving in the same direction, visually illustrating the relevance of dominating the shot count in football matches.
We can see the Roda did better under the previous manager, Harm van Veldhoven, until the summer of 2012, and the first year under Brood Roda declined to a TSR of 0.375, coinciding with a relegation threatening PPG of 0.90. Still, Brood managed to avoid relegation through the play-offs after finishing 16th. Now, halfway into the new season, Roda’s TSR is nicely creeping back up to the usual 0.450 level, with a corresponding uptick in PPG to 1.2 over the past 34 matches.
The next graph shows Roda’s TSR in slightly more detail, with a 10-game moving average added to the thicker 34-match line.
Perhaps now, we can see better why Roda departed with Van Veldhoven, who finished the 2011-12 season in dangerous 0.350-0.375 TSR land. However, since Brood’s arrival at the club, a steady increase in TSR has been established, an Roda look well underway to finally playing above 0.450 level, for the first time in nearly two years.
The next graph shows Roda’s PDO, again with 34- and 10-match moving averages. PDO is a useful stat that combines a team’s shooting percentage (goals per shot) and saves percentages (saves per shot conceded). For convenience, the number is multiplied by 1000. The essential thing is that long-term, PDO tends to revert back to around 1030 for top teams and 970 for bottom teams, since the natural spread in Sh% and Sv% is limited.
Long-term, Roda nicely stays within the expected zone, with PDO’s of around 1010-1020. However, there is a lot of short term movement in PDO. This has everything to do with football being a low scoring sport. One goal can make a difference on short-term PDO, and on short-term results. Long-term, however, PDO and points are very poorly correlated.
Over the past three years, Roda has suffered three dips in short-term PDO. The first, in October 2011, coincided with a drop in TSR and had a dramatic impact on the results. Over the course of six matches, Roda’s 10-match PDO dropped from 1055 to 926, with one win and five losses in that period. But, PDO always recovers, and long-term, Roda’s PDO never dipped under 1000.
The second dip in PDO was a very quick one, in October 2012. From 1033 to 950 in three matches, unsurprisingly bringing three losses. Recovery, as always, was quick, and long-term, Roda hovered nicely around the 1010 mark.
The third dip in PDO is what killed Ruud Brood. Two wins from the past thirteen matches had everything to do with a severe fall in PDO, from 1053 at the end of August 2013 to XXX presently. Ironically, Roda’s TSR (the best predictor for long-term performance!) had just been increasing nicely to above the 0.450 mark.
It looks like Ruud Brood is another PDO victim. Or, to put it in normal language, a football manager who was fired while doing a decent job. Bad luck killed the results, and if Roda had waited long enough for PDO to recover – which it always does – results would have picked up nicely in line the rise in TSR. For now, this Roda vacancy looks like a very good job opportunity, and there is a near guarantee for success.