2026-05-28, 07:32 PM - Word count:
With only the promotion/relegation matches yet to play in S25, I am just about finished with my rookie season. I played right defense for the Cairo City squad, and overall, my campaign was extremely fun and successful. Cairo City was successful as a team, too, and I'm hoping we can finish out the season with one more surprise on Saturday!
After that match, the offseason will give me plenty of time to continue developing my skills and expanding my game, but it's important to head into that time with specific goals in mind. I know I want to be an exceptional defender that can also deliver big plays to help my team succeed, but it's still not clear exactly what I should improve to achieve that goal. What are the best SSL fullbacks good at doing, and what behaviors should I try to emulate? Discovering these habits now will help ensure my development is purposeful.
To answer these questions, I analyzed the 22 fullbacks in the SSL minor league and their performances in S25. They range from 412 TPE to 1439 TPE, with the median at 937 TPE. In S25, their average rating fell between 6.66 and 7.88, with the median at 7.02.
In an attempt to find similarities between these 22 players and identify a smaller set of archetypes that categorize them, I performed a hierarchical clustering analysis. Initially treating each player as its own individual cluster, I executed an algorithm that incrementally paired the two most similar clusters until an entire relational hierarchy was constructed. To keep things simple, I limited the inputs into the algorithm to only 7 important stats that (I believe) are a reasonable simplification of behavior out on the pitch: expected goals (xg), expected assists (xa), dribbles, progressive passes, interceptions, successful headers, and tackles won. There are other things fullbacks are asked to do well, of course, but these are sufficient for performing a high level analysis of archetypes.
The output of the clustering is shown below, with the clusters generally organized with higher average rating toward the top and lower average rating toward the bottom.
![[Image: qI0AFbw.png]](https://imgur.com/qI0AFbw.png)
The bracket structure at the right holds all the information about player relationships. Players linked together more closely to the left of the bracket share the most in common, while players linked together across multiple combinations and branches are least similar to each other. I also took a slice of the hierarchy where there were 9 distinct clusters, and I colored the fullbacks according to these 9 clusters. The black and white color grid in the center of this image indicates whether that player had a high mark (black) or low mark (white) in each of the 7 input variables, but the next plot is a clearer visual of the 9 clusters and their contributions, where each line shows the cluster's average in the 7 input statistics, all plotted on normalized axes.
![[Image: CxldfD4.png]](https://imgur.com/CxldfD4.png)
There is a lot to unpack here, so I'll refrain from outlining every small detail. There are a few interesting things to note, though. First, there were three players that really did their own thing and live in a cluster by themselves. Bob Berendsen in black led the group in expected goals and was also very good at passing, but he had few interceptions and avoided connecting on headers. Lucas Peioxoto in dark green was high across the board with the best expected assist mark of the group. Jakob Fernsterhausen in light green had very low activity in passing, creating chances, and winning tackles, but was extremely successful intercepting and heading the ball. The two blue groups were well-balanced defenders that preferred to bring the ball into the front half by dribbling, while the orange cluster focused more on sending the ball ahead on the pass. My cluster marked in gold (for Cairo City) really didn't have any distinguishing feature other than not being deficient in any one area. The purple and pink groups at the bottom correspond with the 6 fullbacks that had the lowest average rating. The purple group struggled winning possession across the board, while the pink group was extremely honed in on making tackles with little on the offensive side.
The last piece of this analysis is identifying the attributes that were most influential in defining archetype. The plot below scans across the physical, mental, and technical attributes for each cluster.
![[Image: nR64OGf.png]](https://imgur.com/nR64OGf.png)
There is, again, a lot to unpack in this plot, but all the information is here to motivate a particular build plan. Want to play more like Jakob Fernsterhausen in light green and dominate in heading as an outside defender? The physical attributes, especially jumping reach seem to distinguish Jakob from the rest. Want to be better at passing up the pitch like the orange cluster? Focusing on work rate, decisions, and positioning instead of dribbling might imitate that behavior. The assist leaders in dark green and black put a heavy emphasis on jumping reach, anticipation, dribbling, and passing. The purple group that lacked defensive contributions lagged behind in the physical attributes across the board. These are just a few observations connecting the performances of each cluster to their build choices; there are many more trends to discover with deeper study of the data.
In closing, it's worth noting that this data comes from just 15 matches of 1 season and only 22 players, so none of these findings should be taken as fundamental truths, especially since traits, tactics, and teammates, among other variables, also contribute to differences in behavior. I hope it is clear that player builds are very diverse even just within this positional group, different builds can get the same results in certain metrics, and multiple archetypes can create value for their clubs in various ways. In no way is this meant to be a prescriptive guide, just a tool to help influence performance in an intentional way.
After that match, the offseason will give me plenty of time to continue developing my skills and expanding my game, but it's important to head into that time with specific goals in mind. I know I want to be an exceptional defender that can also deliver big plays to help my team succeed, but it's still not clear exactly what I should improve to achieve that goal. What are the best SSL fullbacks good at doing, and what behaviors should I try to emulate? Discovering these habits now will help ensure my development is purposeful.
To answer these questions, I analyzed the 22 fullbacks in the SSL minor league and their performances in S25. They range from 412 TPE to 1439 TPE, with the median at 937 TPE. In S25, their average rating fell between 6.66 and 7.88, with the median at 7.02.
In an attempt to find similarities between these 22 players and identify a smaller set of archetypes that categorize them, I performed a hierarchical clustering analysis. Initially treating each player as its own individual cluster, I executed an algorithm that incrementally paired the two most similar clusters until an entire relational hierarchy was constructed. To keep things simple, I limited the inputs into the algorithm to only 7 important stats that (I believe) are a reasonable simplification of behavior out on the pitch: expected goals (xg), expected assists (xa), dribbles, progressive passes, interceptions, successful headers, and tackles won. There are other things fullbacks are asked to do well, of course, but these are sufficient for performing a high level analysis of archetypes.
The output of the clustering is shown below, with the clusters generally organized with higher average rating toward the top and lower average rating toward the bottom.
The bracket structure at the right holds all the information about player relationships. Players linked together more closely to the left of the bracket share the most in common, while players linked together across multiple combinations and branches are least similar to each other. I also took a slice of the hierarchy where there were 9 distinct clusters, and I colored the fullbacks according to these 9 clusters. The black and white color grid in the center of this image indicates whether that player had a high mark (black) or low mark (white) in each of the 7 input variables, but the next plot is a clearer visual of the 9 clusters and their contributions, where each line shows the cluster's average in the 7 input statistics, all plotted on normalized axes.
There is a lot to unpack here, so I'll refrain from outlining every small detail. There are a few interesting things to note, though. First, there were three players that really did their own thing and live in a cluster by themselves. Bob Berendsen in black led the group in expected goals and was also very good at passing, but he had few interceptions and avoided connecting on headers. Lucas Peioxoto in dark green was high across the board with the best expected assist mark of the group. Jakob Fernsterhausen in light green had very low activity in passing, creating chances, and winning tackles, but was extremely successful intercepting and heading the ball. The two blue groups were well-balanced defenders that preferred to bring the ball into the front half by dribbling, while the orange cluster focused more on sending the ball ahead on the pass. My cluster marked in gold (for Cairo City) really didn't have any distinguishing feature other than not being deficient in any one area. The purple and pink groups at the bottom correspond with the 6 fullbacks that had the lowest average rating. The purple group struggled winning possession across the board, while the pink group was extremely honed in on making tackles with little on the offensive side.
The last piece of this analysis is identifying the attributes that were most influential in defining archetype. The plot below scans across the physical, mental, and technical attributes for each cluster.
There is, again, a lot to unpack in this plot, but all the information is here to motivate a particular build plan. Want to play more like Jakob Fernsterhausen in light green and dominate in heading as an outside defender? The physical attributes, especially jumping reach seem to distinguish Jakob from the rest. Want to be better at passing up the pitch like the orange cluster? Focusing on work rate, decisions, and positioning instead of dribbling might imitate that behavior. The assist leaders in dark green and black put a heavy emphasis on jumping reach, anticipation, dribbling, and passing. The purple group that lacked defensive contributions lagged behind in the physical attributes across the board. These are just a few observations connecting the performances of each cluster to their build choices; there are many more trends to discover with deeper study of the data.
In closing, it's worth noting that this data comes from just 15 matches of 1 season and only 22 players, so none of these findings should be taken as fundamental truths, especially since traits, tactics, and teammates, among other variables, also contribute to differences in behavior. I hope it is clear that player builds are very diverse even just within this positional group, different builds can get the same results in certain metrics, and multiple archetypes can create value for their clubs in various ways. In no way is this meant to be a prescriptive guide, just a tool to help influence performance in an intentional way.
![[Image: F8SELs4.png]](https://imgur.com/F8SELs4.png)
![[Image: LeBbUkk.gif]](https://i.imgur.com/LeBbUkk.gif)
![[Image: ySrStzC.jpeg]](https://i.imgur.com/ySrStzC.jpeg)
![[Image: M7SNVLm.png]](https://i.imgur.com/M7SNVLm.png)
![[Image: Seth-Mc-Neil-Sig2.png]](https://i.ibb.co/BKHRWXxn/Seth-Mc-Neil-Sig2.png)
