Welcome to Beyond the Basics!

My name is Zack Capozzi, and I run LacrosseReference.com, which focuses on developing and sharing new statistics and models for the sport.

The folks at USA Lacrosse Magazine offered me a chance to share some of my observations in a weekly column, and I jumped at the chance. Come back every Tuesday to go beyond the box score in both men’s and women’s lacrosse.

I submitted my Tewaaraton breakdown last week and you all … had some thoughts. There were two main undercurrents from the comments I got. The first was around opponent strength. Several of you rightly pointed out that the teams represented on the Watch List have played vastly different schedules. And while the committee certainly would have considered this, the fact remains that disparate schedules can make it difficult to compare players head-to-head.

I have a model that adjusts team ratings to account for opponent strength, but it’s somewhat easier to do this at the team level than it is at the player level. With teams, you have a straightforward metric to use as the input to a model; efficiency or shooting percentage for example. If the team’s efficiency dips against a given opponent, you can be fairly certain that the opponent was largely the cause.

With players, there is a lot more uncertainty. And what do you even model against? Total production? Individual efficiency? What about the fact that some teams may decide to try and shut off a specific player, whereas another team may not. It’s a problem I plan to try and tackle over the offseason, but my instincts say that the opponent-adjustment model is going to be much less precise for players than it is for teams. And a less precise model means that it may not be as useful as I’d like it to be.

POSITION-FULL LACROSSE

The second class of comments was around the different criteria that are (or should be) used to evaluate players who play different positions. Certainly, because different players have different roles, we should be measuring them based on the context of what they are being asked to do.

But how to quantify that? We can use the position designations listed on the team rosters, but that’s not always a true reflection of a player’s role. We can try and derive a role from their statistical profile (percentage of team shots, etc.), but that’s not going to be accurate in every case (even less so for low usage players).

Still, while it’s not an easy problem, it’s less of a puzzle than a player-specific opponent-adjustment model. While we know that the position given to a player on a roster may not be reflective of reality in all cases, it’s a good starting place. At the very least, it can be a good way to understand the statistical profiles of different positions and who bucks the norm.

So, let’s start there. As an example, let’s look at the share of a team’s shots taken by attackers, midfielders and defenders. The median attacker in Division I women’s lacrosse takes 9.4 percent of her team’s shots; the median midfielder takes 5.6 percent. The average defender takes just 0.4 percent of her team’s shots.

The difference in roles is starker when you look at assists. The median attacker has generated 11.0 percent of her team’s assists. For midfielders, it’s just 3.4 percent. And remember, the midfielder bucket is going to include players whose primary contribution is on the defensive end. If there were a way to exclude those players from the sample, I’m betting that attackers and midfielders would be pretty close to parity on shots, perhaps not so much on assists.

And just to sense-check my logic, we can look at caused turnovers, too. The average defender has 0.9 caused turnovers per game. The average midfielder creates 0.51 turnovers per game, and the average attacker produces 0.24 turnover per game.

OFFENSIVE DEFENDERS

With those baselines set in your mind, we can also dig into the players who buck the trend the most. For example, which defenders have generated the most offensive value? That would be Payton Barr from East Carolina. For the year, Barr has 16 goals on 49 shots. She’s added two assists for good measure. The 49 shots means that she’s taken 10.1 percent of the team’s shots this season.

The rest of the top 10 highest grossing defenders are:

• Brinley Anderson (UMass)

• Layton Nass (UMass Lowell)

• Christine Fiore (Navy)

• Ellen Monahan (Boston U)

• Sammie Morton (Denver)

• Marge Donovan (Princeton)

• Shannon Boyle (Hofstra)

• Kailee Lammers (Denver)

• Casey Sullivan (Jacksonville)

In general, most of these players have done their damage scoring goals. Brinley Anderson is shooting 67 percent (10 goals on 15 shots). But that’s not true in all cases. Layton Nass, Christine Fiore and Casey Sullivan all have at least four assists to go with whatever goal-scoring they’ve accomplished

ATTACKING DEFENSIVELY

The opposite perspective is finding the attackers who have generated the most value as defenders. For this, we use our trusty defensive EGA metric, which boils down to penalties, ground balls and caused turnovers. The leader of this pack is Abby Hormes. In addition to her off-the-charts offensive production, Hormes has the highest defensive EGA mark among players listed as attackers. She’s averaging 0.67 dEGA per game, mostly on the strength of the 55 ground balls she’s picked up this season.:

• Julianna Sanchez (Campbell)

• Megan Matthews (Campbell)

• Maddy Spratt (American)

• Kirsten Flannigan (Wofford)

• Elisabeth Ralph (Wofford)

• Chamberlaine Bell (Bryant)

• Kay Rosselli (High Point)

• Cassidy Burns (Bryant)

• Margaret Stephan (Elon)

I was a bit surprised to see that this group wasn’t here because of a propensity for causing turnovers. Of the top 10 players listed, no one had more than one caused turnover, and three had zero. I was expecting a team that was good at pressuring opponents on the clear would have to have some attackers registering caused turnovers. In retrospect, a failed clear is probably ending up as a ball out of bounds or a ground ball by an opposing defender, so it would be difficult for an attacker to ever be awarded a caused turnover.

INDIVIDUAL PLAYER EFFICIENCY

For the last crosstab, let’s go back to an old favorite: individual player efficiency. Usage-adjusted-EGA is my metric for measuring how efficient a player is. It takes overall production (EGA) and divides by a player’s play share to come up with a value that describes their production per touch. While attackers tend to have higher EGA because they take a lot of shots (see above), uaEGA penalizes those attackers who are volume shooters but may not have as high of a shooting percentage.

Conversely, midfielders who do a bit of everything and do everything well tend to be favored by uaEGA more than in the raw EGA formula. I’m never a fan of using points to quantify a player’s value (assuming that’s all you use), but I think it’s even more the case for midfielders. uaEGA though, since it accounts for their contributions outside of scoring AND it accounts for their lower play share, is a much better alternative.

And what does uaEGA think about this year’s crop of midfielders? Well, it is very, very impressed with USC’s Kelsey Huff. Her efficiency rating is a 97 out of 100. That is phenomenal across any position group, but it makes her stand out, especially in the midfielder bucket.

The nine other most efficient midfielders are:

• Leah Marshall (Iona)

• Emma Tyrrell (Syracuse)

• Kendra Harbinger (UMass)

• Kari Buonanno (Princeton)

• Chloe Schaeffer (Mercer)

• Rachel Clark (Virginia)

• Kendall Sims (Boston U)

• Skylar Simmonds (Bryant)

• Ava Vasile (Vermont)

I feel like I need to put Huff’s 2022 in a bit more context here. Leah Marshall, who has been the next-most efficient midfielder, is closer to Ava Vasile in 10th than she is to Huff’s mark. That is remarkable.