We’ve reached the end of the USL season – though Nashville’s been done for nearly a month – so let’s continue wrapping things up by a graphical representation of the players’ 2018 performances.
A few notes here:
- Field players only. I’ll consider doing something for keepers in the future, but it doesn’t seem like the sort of thing that makes sense without broader comparisons.
- I used a cutoff of 600 minutes played (because otherwise sample size errors would be even greater than they ended up), which removed Jordan Dunstan, Ramone Howell, and Robin Shroot from consideration.
- I also took out Michael Cox and David Edgar, because they played the majority of their minutes with other teams (St. Louis and Ottawa, respectively), and the way the USL website presents the data, there’s no way to separate that out. Neither would have played over the 600-minute threshold for NSC, anyway.
- That leaves a pool of 17 field players.
- Keep in mind that some of these factors are an indication of quality, others are a description of style. “Was in more duels” is not necessarily synonymous with “better,” just a different type.
- That said, I’m not happy with a couple of the metrics representing the sort of thing I wanted them to. Specifically, duels are not as indicative of a defensive mindset as I’d thought (particularly because aerial duels went mostly to Tucker Hume on longballs, etc.). I’d re-calculate the data, but I got way too deep into the process before realizing it, so it’ll have to wait for another time.
- Since I’m using limited software here (Google Docs, actually), the wheels are a bit tougher to interpret, with no raw numbers. Everything is scaled from lowest on the team (0) to highest on the team (1), without regard for how it’d stack up to the rest of USL. For example, Brandon Allen had the best finishing rate on the team (30.3%), so he’s represented by a 1. There were plenty of USL players with higher marks (such as Cincy’s Danni Konig at 37.9%), but they’re outside of the sample size.
- The stats are divided into four categories, starting with usage in the upper right, and going clockwise through shooting, passing, and defense. Each category includes four metrics, though as mentioned above, I’m not super-happy with how representative they all are of what I’m going for.
Here we go:
Primarily offensive players
Forwards, wide midfielders (minus Taylor Washington, who played wingback and fullback more than he played as an offensive-minded midfielder), and central attacking midfielders. Not sure whether to stick LaGrassa here because he also played significant amounts as a central defensive midfielder, but given his time as a winger and second striker, I guess I will.
Winn’s role as an offense-minded winger was one that worked out pretty well for him as a distributor, especially. He barely edged out Kris Tyrpak for the mantle of “greatest percentage of his passes were key passes.” His finishing could use some work, and he was mostly a non-entity defensively.
Allen’s role as a poacher and finisher cannot be overstated. Of course, there’s a bit of a confounding factor here: four of his ten goals on the season came from the penalty spot, and two of them came with the Bethlehem Steel before his transfer.
Were it not for his season-ending injury, Moloto would have been one of the ironmen of this team. His conversion rate on shots was well-documented as being too low (though, as I’ve enumerated plenty of times in the past, that’s probably a product of feeling like he had to do too much with a whole new team, especially early in the year). His shots on-target rate indicates bad luck played a part, too. He was also one of the key creators for this team.
LaGrassa played multiple roles for the team, as described above. His offensive numbers certainly indicate that he spent much more time in that CDM role (which I believe to be true, though I haven’t gone back and checked). His win rate on duels and tackles is certainly pretty good.
Jome, like LaGrassa, played multiple roles, though his were a little less diverse: left winger, left fullback, and a little bit of central defensive mid. He pretty much got benched after getting a key red card.
Hello, Mr. “tries shit.” If Mensah had been at full fitness earlier in the year, this team’s (often deserved) reputation for being a bunker-counter squad with little creativity in the final third might have been different. Mensah’s conversion rate wasn’t great, but to a certain extent, having him out there was not only a way for him to score, but to open things up for teammates.
Extremely similar graph to Winn’s, save for the fact that Tyrpak didn’t join the team until August and only got into five games. A whole season with him available would certainly be interesting (though he and Winn have overlapping skillsets, to an extent).
The “shoot only” version of an offensive player. You’d actually like to see at least the passes per 90 be higher, given that he’s a hold-up striker. If the key pass version of a hockey assist existed, though, he’d be much higher. Also: the graph that made me realize duels don’t belong in the “defensive actions” category.
Primarily defensive players
The rest of ’em. As you can figure from the above, there’s some overlap in the LaGrassa/Washington/Jomes of the world.
The only player on the team (or at least among these 17 who got enough playing time to count) who didn’t register a shot. Solid defender and ground-coverer, and the majority of his key passes were crosses in from the wing.
A lot of minutes played, solid defensive numbers (remember, we shouldn’t be holding a lack of volume in duels against him), and decent action going forward with key passes. Given that he played both centerback and fullback, the pass numbers generally get a little more impressive (aside from long passing, which you expect more of from a centerback).
The most offensive of NSC’s central defensive mids, Reed made an offensive impact with line-breaking passes (that long pass mark is pretty nice, especially when considering how many of those passes turned into key passes, and how accurate Reed’s passing was overall). He didn’t get forward much until later in the year, which you’d like to see more of with a team that’s a bit more comfortable with each other next year.
James didn’t play a ton to get much data on him. Non-entity offensively (unsurprising given that much of his time, especially late in the year, came as a third centerback sub). Was a very good ball-winner, though.
Some eyebrows were raised about Doyle’s selection as the team’s defender of the year, but the graph is pretty impressive to me. Tons of blocks and clears, did a great job winning tackles, wasn’t a liability with the ball at his feet (completing a lot of passes despite simply booting many of them upfield), and was pretty much an ironman.
I’m actually fairly surprised Bourgeois’s long passing rate wasn’t higher, because there was a stretch in the middle of the year where it seemed like he was just instinctively banging it upfield. He would have been one of the minutes leaders if not for a mid-season injury, he would have had a ton of minutes, too. Glad to see him get a couple goals in there, as well.
Played multiple positions, scored on one of just seven shots on the year. Wasn’t super-involved on or off the ball, based on the graph, but was good when called upon.
Hello, weird graph for a central midfielder. Akinyode was very good defensively (upper left quadrant) and got plenty of playing time (upper right). The bottom two portions are where it gets interesting: he was a non-entity offensively – aside from one absolute banger against FCC, of course – and his passing chart shows a guy who was similarly not involved either getting forward or moving the ball into the offensive third. “Guy who doesn’t mess up with the ball at his feet” is certainly an asset for a team, but I’d like to see more (or, if he’s not going to produce going forward, a couple fewer situations where he was jogging back in defense while his guy scored or set up a goal).
Woodberry actually had the ball at his feet a lot for a centerback. He was fairly solid blocking shots and clearing them with regularity (perhaps there’s something to be said for that), though the other centerbacks had a bit more. Anecdotally, he did have a game-losing own-goal, of course.
What we learned
Aside from “let’s make sure we understand what part of the game duels demonstrate before chopping up the data,” I think a lot of what we see here either follows with what we saw on the field (“Ropapa tries to make things happen,” “Akinyode may not be physically capable of a pass longer than eight yards”), or taught us something that we might not have otherwise realized (“Hume’s shooting was actually more important to the team than his hold-up play,” “Winn and Tyrpak were far and away the most important setup men”).
Again, some of the graph is on a scale of “bad to good” while other parts are simply stylistic measures, so there’s a bit of mining you can do with these.
If you have any suggestions for how to make the graphs more enlightening, or a question/suggestion/etc. otherwise, let me know in the comments or drop me a note on the social channels. I’m all ears, and trying to get as much information displayed in an interesting and informative way as possible.