Offseason Notebook: Final(ish) projections, plus Oakland and El Paso
Offseason Notebook: Final(ish) projections, plus Oakland and El Paso
Exploring offseason upgrades from the Oakland Roots and El Paso Locomotive and the data model's stab at league tables
With a week (and, surely, some signings) to go, here’s what my model is thinking about the season to come in the Championship and League One:
How does the model work? You start with an individual player like Rhode Island’s Albert Dikwa, the reigning Chanpionship Golden Ball winner. Using a formula to calculate “Goals Above Replacement,” a one-number measure of value derived from xG, xA, passing totals and accuracy, defensive actions, and on-off splits to capture intangibles, Dikwa rates out as being worth about four points across an estimated 30 matches played. Multiply the minutes and value together, repeat that across an entire roster, regress it to get a team points value, and voila.
New signings get extra layers of complexity. Age, league of origin, and my expected minutes estimates are combined to predict a value. League One players carry over their worth with a slight penalty; Championship players that drop down get a slight bonus. Mix it all together, and you’ve got the table above. Odds are then created based on a historical formula for points value and playoff qualification.
What of the predictions at a granular level? Death, taxes, and the Louisville-Tampa Bay duo atop the East. The newcomers aren’t far behind them, and there’s a solid little bubble hovering around the eighth seed. I think the results are unduly harsh for Birmingham and Loudoun, but that Miami number really catches the eye in a negative sense.
In the West, Sacramento is the favorite by a country mile. My model simply loves them, even if there’s absolutely zero chance they sweep the conference by a total approaching 16 points. The race for the playoffs is pure insanity; less than two points separate the Switchbacks in fourth from Tulsa in tenth. Las Vegas is a real unknown here. They’ve got a squad, but there’s lots of depth to add in the central midfield and elsewhere that’ll swing the prediction significantly.
League One is defined by a fairly distinct top four, one topped by a Forward Madison that had a crazy good offseason with proven additions from full back (Ferrety Sousa) to the midfield (Jimmie Villalobos) to forward (Juan Galindrez), though Charlotte is close behind. Chattanooga is expected to edge back into the playoffs, as is expansion Spokane. Meanwhile, my numbers think Jermaine Jones has a roster that’s destined for the basement and doubts a retention-heavy Richmond winter.
These projections won’t be perfect, but they nailed about 75% of playoff teams last year and, at the other end, correctly identified that a revamped Hartford team would crater. The numbers are worth something even if my subjective table is different in meaningful ways.
Has Oakland done enough to get back into the playoffs in 2024? They had the fourth-best offense (52.8 expected goals for) and the twelfth-best defense (45.3 expected goals against) last season but couldn’t get over the hump because of a 10.7% conversion rate that sunk a statistically sound side.
This winter’s moves have focused on the defensive line by and large, though Miche-Naider Chery may be the striker needed to fix the finishing problem. It’s an interesting approach given the personnel who have relocated to the East Bay, but how might the Roots change their look in the season to come?
Stylistic consistency with personnel turnover is on the menu, with Noah Delgado set to run back the 3-4-3 shape introduced by Juan Guerra a few seasons ago. It’s a shape in which the outside backs drive offense, the central midfielders are expected to push up in the press, and the front line is defined by vertical movement.
Working into the half spaces and focusing on verticality are definitional for Oakland. Last year, you’d regularly see Danny Barbir and, to a lesser extent, Emrah Klimenta drive into the channels from the center back spots with the ball at their feet. Note that Barbir (Indy) and Klimenta (Las Vegas) as well as fellow center back Tarek Morad are gone heading into the new campaign.
To accommodate that style of movement out of the back line, the Roots tended to drop their central midfielders low or wide as a counterweight. It was a difficult role for the No. 6/No. 8 types; balancing offensive depth with up-the-pitch pressing aggression for players like Daniel Gomez and Napo Matsoso in the pivot could unsettle the defensive structure. With risk comes reward: the scheme did set up some genuinely exciting bouts of mazy dribbling from Barbir and co.
The progressivity and freedom of the defenders themselves could also come back to bite. Delgado’s side ran with a very high defensive line, one that wasn’t well-shielded by the midfield and rarely got the inside rotation from the wing backs to help make weight. You can see how that created problems above.
Oakland’s secret has always been Paul Blanchette’s insane ability to shut down one-on-one breaks: he had 6.3 goals saved above expected last year which rated in the 90th percentile on a per-match basis, and he was one of 10 regular starters to register a 70% save percentage or better. Still, Blanchette couldn’t patch over the sheer number of high-leverage looks that the system conceded.
So how do the new faces figure in? In terms of intra-USL acquisitions, Niall Logue doesn’t seem to be as natural a fit on the stat sheet but will do quite well in practice. Logue has been stuck in more direct and physical assignments in recent years that limit his numbers, but the 6’4” center back has prototypical size for his position and a long stride. There’s ample technique, too: American Soccer Analysis would’ve ranked him as Oakland’s second-best passer out of the back line last year (+0.75 passing goals added) in spite of Hartford being disastrous.
Camden Riley is the standout. He’s familiar with a very Barbir-esque ball-carrying role, only flipped to the right side of the pitch rather than the left. Riley is the sort of player who’s grown better and better with each passing season between his stints with RGV and the San Diego Loyal, and his extremely keen read for moments to intervene (see the 1.9 tackle attempts per game) is exactly what the Roots need.
You can see Riley doing his thing in the reel above, as well as similar contributions on the left from Justin Rasmussen. A left back by trade, Rasmussen had been an early-season starter for the Portland Timbers two years running before losing his place by summertime. In 2023, he put in just 1,000 minutes between MLS NEXT Pro and MLS, but he’s clearly got the passing skills and crossing boot to emulate the Barbir types.
I’m not quite sold on Rasmussen’s awareness at the back post, having watched as much Portland tape as I could find from his MLS minutes. Still, he’ll benefit from playing in a back three where there are two other central defenders there to find marks. He could also play at the left wing back spot, though Memo Diaz and Baboucarr Njie are more than capable of holding things down.
It is a bit of an odd mix. There are essentially five center backs on the roster, and Bryan Tamacas can cover there in a pinch. If I were Delgado, I’d shift Neveal Hackshaw into the defensive midfield. He’s no stranger to the position throughout his USL career, and his goals-added totals of +1.11 for passing and +0.76 for interrupting last year paint the picture of a do-it-all holding mid if there ever was one.
Of course, the depth is necessary because Tamacas and Hackshaw will, respectively, miss time on Salvadoran and Trinidadian international duty; expect plenty of run for Logue and Gagi Margvelashvili in defense. There are oodles of options in this group.
Is the new set an improvement? My data modeling doesn’t seem to think so, but there’s a lot to be said for Riley as a step up over Klimenta as a defensive stopper and the addition of very, very good depth pieces when duty calls. If they can at least keep Oakland as a mid-table defensive side, they’ll give Chery and co. the runway they need to fix the more important goalscoring issues.
Famously mum on their preseason results, the El Paso Locomotive cut a hugely intriguing profile heading into the new year. What catches my eye is how Brian Clarhaut and co. have re-invented their forward line almost entirely, replacing 7,700 minutes of Lucho Solignac, Aaron Gomez, Denys Kostyshyn, Emmanuel Sonupe, and Petar Knudsen.
Solignac (0.73 goal contributions per 90 minutes in the last two years) has been fairly directly swapped out with Justin Dhillon (0.64 contributions a game in the same span) at the No. 9 spot - literally, with the strikers trading San Antonio FC and El Paso shirts. It’s the rest of the attacking line that’s so fascinating for me; the potential at Southwest University Park is tremendous.
The swaps haven’t been like-for-like. Denys Kostyshyn is fairly one-of-a-kind to begin with, a dominant presence in the right half space with very strong vision. He defined El Paso throughout 2023, often by tilting the pitch (in tandem with defender Marc Navarro) to lean heavily upfield on the right. Emmanuel Sonupe was a purer sort of energy man, a speedy wide forward who often came on late to help chase a game and put up instant offense. Gomez was the most beloved locally, a slaloming second striker who always gave 110%.
Tumi Moshobane and Joaquin Rivas provide aspects of Locomotive stars past. I’m an unabashed fan of both, but Moshobane is especially exciting. He was usually used as a No. 10 or withdrawn right winger last season with the San Diego Loyal, one that contributed to 39 goals on 29 expected contributions in the last four seasons. He’s extremely good about picking the ball up around halfway and carving through half the defense to initiate a counter, not unlike a spunkier Kostyshyn.
Rivas, meanwhile, has Gomez’s nous in the box with a more judicious brand of movement. When the Salvadoran commits to a run, he commits hard and usually pays off the effort. That’s not to say Rivas is lazy; he’s more of a probing type, one that isn’t wont to take opponents on head-to-head but prefers to find pockets. Late-arriving dives at net have garnered 23 goals since 2020, including 10 last year.
That leaves Amando Moreno, a superstar in the USL who hasn’t received nearly enough love in terms of awards and attention. He’s a more willing dribbler (about successful 2.0 take-ons a game) than El Paso have had in years, one with a mean right-footed finish that gives him undeniable gravity in the final third. American Soccer Analysis rated him as the single-best receiver in the USL, someone who added 2.4 goals in 2023 from finding touches in great spots. Getting him from rival New Mexico United is a coup and a half.
My expected shape based on the roster build is a 3-4-2 or 5-2-3, depending on the phase of play. In many ways, it won’t be wholly dissimilar to what Clarhaut broke out at the end of 2023. Dhillon is a nasty presser when he wants to be up top, and flanking him with two capable ball-carriers with good shooting boots via Moreno and Moshobane seems a potent mix. The attack will be lethal, even if Clarhaut goes a different way with the alignment.
I’m thinking back three just because of the numerical depth at the center back spot in El Paso. Bolu Akinyode could cover as the centermost man in a trio, but he’s best off as No. 6, constantly putting in interventions; 5.2 defensive actions a game with Miami last year don’t lie. Elijah Martin is an interesting one who could play anywhere in a back five. That still leaves four dyed-in-the-wool centermen, and this team didn’t ink Noah Dollenmayer to a permanent contract or loan in Brandan Craig, a USYNT stud, to have them sit.
We’ll see how it all shakes out, but I might be higher than anyone else in the USL-sphere on this Locomotive side.
That’s all, folks! It’s almost time for the regular season, and you can find the thousands upon thousands of words of power rankings and analysis every single week on Backheeled after opening day. Elsewhere this week:
I think Charleston has announced that they’re using a 3-5-2 ahead of every preseason game, and I’ll eat my shirt if they don’t start the year in a back four. The lineups are blatantly constructed with two center backs and two full backs; ex-Riverhounds defender Nathan DosSantos is often running on the right, and Mark Segbers has typically been inverted on the left. This is a quibble that approximately zero people care about, but that’s kind of my whole MO.
A fun offseason game for me is to spam my mom with kit photos and ask for opinions. Oakland’s home jersey is a big winner; Pittsburgh’s well-received bridge shirt - which I dig! - did very poorly. If you ask me, Loudoun’s home pinwheel and literal away dark horse are the big champs.
I was traveling this week, so little else in terms of links. Check out the USL Show’s Championship preview to get a catch-all.
Thanks again.