PLAYER PROFILE: Dirk Kuyt
Wed 14 Nov 2007, 12:47
As we approach and overtake the 1000 minute mark of the season we finally have enough data to start analysing the results and drawing conclusions from what we are seeing. While looking through the attacking, defensive and total goal value figures for each player and team this week (as we prepare to bring you our first third of the season report cards) we noticed an unusually low rating for a player who finished last season above average in all three measures and inside the top-20 in attacking goal value. That player was Liverpool’s Dirk Kuyt and he is the subject of this week’s player profile.
Before Last Season
Dirk Kuyt arrived at Liverpool on 18th August 2006 from Feyenoord with many thinking he was the 20-goal player conventional wisdom believed the team needed. The main reason for this assumption was his prolific goal scoring record in the Eredivisie where he had averaged 23.67 league goals per-season during the last three years. Unfortunately, lost in translation (literally) was the strength of the league Kuyt would be entering compared to the league he had just left. Conventional wisdom is that the standard of many of the European leagues are comparable, but nothing could be further from the truth and although there were a few scribes that mentioned that it would be harder for the Dutchman to score in the Premier League many still underestimated the change in quality.
Using historical data it is possible to translate performance statistics from one league to another in much the same way as exchange rates work between the currencies of different countries. For example from our calculations the historical translation for a player’s goal scoring rate moving to the Premier League from the Eredivisie is approximately a 26% drop off. As you can see that is a significant difference and one that can make many a Dutch recruit seem a transfer bust.
Using this translation on Kuyt’s goal scoring from the top Dutch league gives a better representation of his true ‘Premier League’ ability before last season even began. Over his last three years for Feyenoord he scored 71 league goals at a rate of 0.70 per-90 minutes. Using our translation method his Eredivisie goal rate equated to a Premier League equivalent of 0.516 goals-per-90 minutes. Of course this is just a baseline level for his past performance and many other factors should be considered (Age, team, experience, playing time, will they be the teams 1st, 2nd, 3rd or 4th forward etc) before projecting for a future season. However, even without any further adjustment and Kuyt playing every minute of Liverpool’s 2006/07 league campaign the most goals one could have realistically expected from the Dutch international last season was 19 based on his translated goal scoring rate.
After translating all his performance statistics and accounting for many other factors our SIMprofilerTM projection model had him producing goal/assist/points rates in 2006/07 of 0.485/0.406/0.891 in 2657 minutes - approximately a 14 goal 12 assist season.
The 2006/07 Season in Review
Kuyt’s season begun well, scoring his first goal in his second start of the campaign, and netting five times in his first nine appearances resulting in a goal-per-90 minute rate (0.689) - close to those he produced in Holland. The remainder of the season was more of a struggle in front of goal as the Dutchman only scored a further seven times in over 1900 minutes (0.327 goals-per-90 minutes rate) that included a one goal in 876 minute stretch from mid November through December.
What separated Kuyt apart from other forwards in the Liverpool team and the league as a whole was his general fitness levels and his tireless work rate that enabled him to be an excellent link man between midfield and attack and also create opportunities for his teammates. His assist contribution was much more consistent and wasn’t subjected to the same drought as his scoring production. In fact during the same 876 minute period mentioned above, Kuyt maintained his effectiveness and ‘team’ play by contributing five assists to continue to create goals even if he wasn’t the one ending scoring moves.
Once the season was over Kuyt had established himself as the team’s top forward and even led the team in goals with the overall stat line below:
26 Starts 2578 minutes 12 goals 10 assists 22 points with respective rates 0.42/0.35/0.77
His final raw totals and production rates weren’t too far removed from what SIMprofilerTM projected or from what his translated Feyenoord performance baseline suggested. However, many were disappointed because his raw stat totals were so much lower than his mammoth Eredivisie numbers. The truth of the matter is that thanks to the vast difference of quality between the two leagues, and the reduction of playing time from his Feyenoord days, even if Kuyt had played exactly to his previous level of performance he would have scored only 2-3 more goals.
His final figures, therefore, were not that far from what could and should have been expected, with his ability to create and bridge midfield and attack making him all the more valuable overall leading us to write this for his scouting report in THE FOOTBALL REVIEW 2007:-
“To assess Dirk Kuyt just by how many goals he scored would be missing the point about what the Dutchman offers. He finished 14th in goal rate, 10th in assist rate and 11th in points-per-90 minutes, among forwards with over 1000 minutes. His aGVA was 8th, his dGVA 4th and his tGVA 5th in the league for his position. That last statistic probably best illustrates his worth. Once you include everything he offers he becomes one of the most valuable forwards in the game. Excluding his work rate, not one of his abilities is outstanding but when you add up his ability to score, his ability to pass, his ability to bridge midfield and attack, his ability to closedown all over the pitch, his ability to create space, and the many others he does possess, you get a 10 goal – 10 assist striker that will do anything for the team. In this day and age of prima donnas and sulky strikers, that is valuable indeed.”
Pre-season Projection
The pre-season projection given in this year’s book had Kuyt producing at a better rate, but in less time meaning the early signs were that the Dutchman might score fewer goals this season:
2256 minutes 11 goals, 12 assists, 23 points at 0.456/0.474/0.930
This prediction was generated in June and before Torres or any other of Liverpool’s summer signings were signed, sealed and delivered and, therefore, still had Kuyt pegged in as Liverpool number one striker.
Transfer deadline final projection
Once Torres, Voronin, Babel and others were entered Kuyt’s projection changed significantly. With the arrival of Torres we down graded Kuyt to Liverpool’s 2nd forward and re-ran SIMprofilerTM giving us this output:-
Table 1.1 Dirk Kuyt’s percentile forecast for the 2007/08 season as of 1st September 2007
|
|
Rates |
|
Raw Totals |
| Percentile |
Minutes |
G90 |
A90 |
P90 |
G |
A |
P |
| 85th |
3206 |
0.663 |
0.684 |
1.192 |
24 |
24 |
48 |
| 80th |
3074 |
0.553 |
0.617 |
1.061 |
19 |
21 |
40 |
| 75th |
2999 |
0.507 |
0.557 |
0.995 |
17 |
19 |
35 |
| 70th |
2773 |
0.469 |
0.499 |
0.917 |
14 |
15 |
30 |
| 65th |
2529 |
0.433 |
0.459 |
0.851 |
12 |
13 |
25 |
| 60th |
2452 |
0.371 |
0.428 |
0.835 |
10 |
12 |
22 |
| 55th |
2346 |
0.353 |
0.364 |
0.798 |
9 |
9 |
19 |
| 50th |
2192 |
0.337 |
0.318 |
0.725 |
8 |
8 |
16 |
| 45th |
1825 |
0.300 |
0.284 |
0.650 |
6 |
6 |
12 |
| 40th |
1617 |
0.278 |
0.262 |
0.636 |
5 |
5 |
10 |
| 35th |
1509 |
0.227 |
0.230 |
0.613 |
4 |
4 |
8 |
| 30th |
1310 |
0.218 |
0.196 |
0.553 |
3 |
3 |
6 |
| 25th |
1202 |
0.185 |
0.156 |
0.521 |
2 |
2 |
5 |
| 20th |
968 |
0.149 |
0.077 |
0.484 |
2 |
1 |
2 |
| 15th |
885 |
0.105 |
0.000 |
0.395 |
1 |
0 |
1 |
| Mean |
2192 |
0.41 |
0.41 |
0.82 |
10 |
10 |
20 |
Table 1.1, above,displays Dirk Kuyt’s percentile forecast for the up coming season. Due to how the SIMprofilerTM model is constructed using a set of ‘highly’ similar players we are able to produce a forecast distribution that gives a prediction of the player’s performance at various levels of probability.
The model produces playing time and attacking rate statistics at percentiles ranging from the 15th to the 85th to give an indication as to the possible variability in a player’s performance. These can be useful in determining the chances of either outstanding or abysmal performances in the up coming season as well as the production levels one might have more confidence in occurring.
This output obviously has its benefits over a rigid single forecast line and gives a better feel for a player’s maximum and minimum level of performance that could realistically be expected. It can be useful in determining which players even if their performance collapses will still be an above-average performer and which players even if they serious improve will still be no better than average standard.
The mean projection (and that given in the book) is simply the weighted average forecast from the model and incorporates all the potential values into one single forecast line.
For Dirk Kuyt the arrival of better striking partners meant a slight downward shift in predicted production. Both his playing time (64 minutes less) and his rate stats (0.046/0.064/0.11 less) were projected to decrease compared to his pre-season output. Even with this slight reduction in the forward’s production rates he is still forecast to be a productive player thanks to his ability to create goals as well as score them. For this reason he is still highly likely to be an above-average forward in terms of goal creation as detailed in his points-per-90 minutes probability distribution (below in Figure 1.1).
Figure 1.1 Dirk Kuyt’s P90 probability distribution
Dirk Kuyt’s points-per-90 minutes probability distribution, Figure 1.1 above, gives the estimated range for his predicted performance in attacking points for the upcoming season - generated by SIMprofilerTM. The distribution (the solid blue line) gives his expected performance in P90 at various levels of probability based on his baseline performance and set of most ‘similar’ players. Also given on the graph is the points-per-90 minutes rate for the league-average forward (the pink flat line). Where the two lines meet gives an indication as to the chances that a player’s production will be above or below the average level for his playing position.
The graph shows that Dirk Kuyt has a 72% chance of finishing the season with an above average points-per-90 minute rate for his position. As you can see from his percentile forecast (Table 1.1 above) this is mainly due to the greater stability predicted in his assist rate that has a 65% chance of being above average for his position compared to that of his goal rate (52% chance of being above-average). The forecast also shows that only at the upper limits of his probability distribution the Dutchman could be expected to score at the 20 league goal level that some predicted and expected when he arrive in August 2006. It is therefore not out of reach of his abilities, but it is highly unlikely given the evidence of his past level of ability.
Current 2007/08 Pace
Kuyt’s current pace, however, is the real worry. So far this season the former Feyenoord frontman has only managed two goals and two assists in 742 minutes for production rates of 0.24/0.24/0.48. That is close to approximating his 35th percentile performance (Table 1.1 above) and way below his mean forecast. When you also consider that the two goals he has scored where both from spot kicks he is currently goalless from open play and only been a part of two goals in just over eight full games.
Having reviewed and scrutinised Kuyt’s 2006/07 season performance it is clear that he can occasionally go into long goal scoring slumps, as described earlier, but last season he was able to overcome that problem by still being able to create for others. His current problems are also much more in focus right now because they have come to start a season and not been lost in the middle of a season that started so brightly as his performance last season.
Fans of the Dutchman should fear not because he will get thing turned around and with a 72% chance predicted to be an above-average forward it should be in a large way. It is highly likely that he will get somewhere near the 10 goal, 10 assist level SIMprofilerTM projected and that is all that should and could be expected from the Dutchman with an understanding of ageing curves in modern football and interpreting performance statistics from other leagues. Unfortunately, Dirk Kuyt will forever have unrealistic expectations attached to his career at Liverpool thanks to the artificial decline in production that occurs when a player moves from a lesser league to stronger competition that currently gets lost in translation.