A statistical look at Yeimar Gomez Andrade

Editor’s note: There has not been an official announcement about Yeimar Gomez Andrade, but it looks like it’s heading that way.

Seattle Sounders target Yeimar Gomez Andrade has an unusually broad statistical profile for a centerback — with above-average to elite rates of attempted (and successful) tackles, clearances, aerial challenges, and (particularly) interceptions. In Argentina, passing has been the only notable hole in his game. A transfer from the Argentine Superliga to MLS would likely change his game — both due to league context and adjustment to a new CB partner. Nevertheless, he would likely be very successful in whatever manner his role is adapted to the league.

Yeimar is a very good ball-winner — an above average tackler, good in the air, with a great interception rate — by the standards of the Superliga. We’ll talk about those other two lines beneath him in the second table at greater length below, along with an in-depth discussion of what his Argentina performance can tell us about a prospective transfer.

With respect to defensive activity, I’ve plotted passing and defensive action rates for Superliga and MLS starters at CB exceeding 100 minutes of play in 2018/2019 and 2019 seasons.

Across three seasons of play in Argentina for Union de Santa Fe since 2017, Gomez has persistently registered more than 13 defensive actions for every 90 minutes on the field — an elite activity rate for Argentina and a level MLS CBs don’t presently reach.

As I noted last week, looking at Joao Paulo’s performance in Brazil, it’s problematic to project player performance across leagues, and the measure of defensive actions per 90’ is heavily influenced by player role. Many teams, like Seattle, mix up CB pairings with players with greater responsibilities in one or the other phase of play (passing; stepping up towards midfield; holding deep; contesting aerials on set pieces) and these variable responsibilities plainly impact the resulting statistics. In Argentina, Yeimar regularly exhibited modestly higher rates of defensive actions than his Union CB partner Jonathan Bottinelli, while Bottinelli took the much greater passing responsibility. With Seattle, Yeimar could fill a similar role as the “stopper” of the CB pairing, with Xavier Arreaga the “sweeper.” The terms are a bit archaic for the present state of soccer tactics, but are well used here — Yeimar the more aggressive pure defender, Arreaga the safety valve/distributor with freedom to push up into the midfield while on the ball.

Roles and tactics are clearly an effective means of emphasizing player strengths and hiding weaknesses. In projecting Yeimar’s potential in MLS, the familiar role should keep his performance relatively consistent between leagues. But what do we do when the leagues themselves exhibit significant play style differences.

Teams in the Superliga generally play more direct and more defensively active games than their MLS counterparts. You can also see this in the chart at the top of the article, with Argentine CBs exhibiting a generally higher rate of defensive activity than MLS CBs. The largest differences are reflected in defensive clearances and aerial challenges. All this is a key reason for my warning that soccer statistics cannot be rigorously translated between leagues. Players may be chosen to apply a particular limited skillset to a need within a league context — while the rest of the team covers for weaknesses in other aspects of play. To quantitatively project player performance between leagues we need three things:

1. A clear data set of player transfers between the leagues. It’s not enough to know the overall league differences noted above — these may allow us to adjust rate statistics to the style of play, but since these numbers are also defined by role (and I would argue role is more important) this wouldn’t be useful without similarly analyzing team fit. We don’t have reference points to convert “skill” statistics (e.g., how difficult is a “tackle” or “dribble”?) without looking at players in both league contexts.
2. Players in similar roles to the transfer target. Since the league styles contrast, as we noted above, success in a particular phase of play (let’s say “midfield tackle attempts” for a reason we’ll get to by-and-by) may not translate to success in others. Other player roles may also have insufficient data for relevant skills (e.g., attacking midfielders may have too few aerials to relate to CBs, or CBs may have too few dribble attempts to relate to attacking midfielders).
3. Performances from recent years. With MLS, the influx of talent to the league with allocation funds over the past several years complicates statistical comparison. Even if we had OPTA event statistics for both leagues pre-2016, these numbers wouldn’t necessarily be useful today.

Take everything together, and we have fewer than 20 players with sufficient playing time in both leagues. We have four CBs (Leandro Gonzalez Pirez, Eric Godoy, Franco Escobar, Jefferson Mena), none of whom make for an ideal comparison to Yeimar (either because they are dissimilar players or lack sufficient data). Still if we were to ignore the above concerns and attempt league conversions, how would it look?

Here I’ve included only those players with at least 20 attempts (which is far too small a threshold, but again, this is a bit of carefree, reckless stat-mongering) at the specified action. CBs are marked with gold diamonds, and other players with blue circles. Linear trendlines are based on all included players. The dashed line represents 1:1 equivalence between the leagues — points above the line show improvement in MLS while points below show decline. The trendline equations are the basis for the “linear regression” statline in the stat table, while the “avg difference” line simply takes the average percentage-point difference between the leagues for the four CB examples and applies it to Yeimar’s numbers.

Transplants from the Superliga to MLS exhibit reduced tackle success and generally improved dribble success and short pass accuracy. The linear trendline is a poor model for the aerials data — Leonardo Jara and Eric Remedi saw their aerial success rates drop dramatically in MLS play. We should be very cautious to apply any of this to projecting Yeimar’s potential play in MLS, but the varied calculations perhaps give us a safe range of expectations.

There are few if any players who play a comparably active defensive CB role to Yeimar in MLS. Ike Opara is perhaps the closest, which is why I have included his stat line here. We would expect Yeimar’s clearance rate to drop due to league context (and possibly his interception rate as well, to a lesser degree). We would expect his pass accuracy to rise given the same balance between short and long passing. We should expect him to be somewhere between good and great in the air, and a below-average to above-average tackler. Overall, he would be an exceptional ball-winner by league standards and a near-ideal fit to partner with Arreaga. If Seattle successfully completes this transfer and Yeimar adapts happily to life in Seattle and MLS, he should be a joy to watch on the field.