My Lax-Elo rating algorithm was one of the earliest features put into the LacrosseReference infrastructure. It’s simple, it’s elegant, and it does a pretty good job of forecasting the probability that one team will beat another team.

Of course, there is no reason that Elo ratings should be restricted to team strength. In fact, any situation where you have a series of one-on-one match-ups is amenable to Elo. Faceoffs, for example. Although wings play a role, you essentially have a series of one-on-one matchups between FOGOs. This means that you can apply Elo mechanisms to faceoffs to come up with a ranking of FOGOs that is superior to win rate.

Superior you say?

Explain Elo to me again

Elo is a ranking system, originally created for chess, by Arpad Elo (thanks to a reader who recently suggested that I should write it “Elo” instead of “ELO”). The mechanism is that for each event (game, faceoff, chess match), the loser forfeits a number of rating points to the winner. The number of points is based on the relative strengths of the opponents prior to the match-up. In situations where there is a margin of victory, that can affect the number of points transferred as well.

Elo systems have been implemented for many sports, especially on Nate Silver’s 538 site.

The envelope please…

But I have to say, I was pleasantly surprised to see the symmetry involved. Across more than 3,000 career faceoffs, they currently sit just 1 Elo point apart. Or, if you want to look at career faceoff wins, Ierlan is at 1,156 and Baptiste finished with 1,157. Of course, given that Ierlan has another year of eligibility, it is very possible he will end up besting the Denver star.

Of course, it’s also possible that he will not. The tricky part of using a career Elo system is that Ierlan has 3 years of history to match or exceed if he wants to cement his status as the top FOGO. If Ierlan turns in a season more in line with a top-10 FOGO, rather than a Baptiste-like season, he’ll probably end up 2nd all-time (since 2015 I should say).

The next tier of FOGOs, per Elo, is a bit more interesting. Alex Woodall certainly had a stellar career for Towson, so I wasn’t surprised to see him near the top of the list. Zach Cole has been great for Saint Joseph’s since he came on the scene in 2019.

But I was a bit surprised to see Jon Garino there at #4. For starters, he wasn’t even the top FOGO on any of his Maryland teams. Second, between 2015 and 2017, he only registered a total faceoff record of 147-116 (55.9%). That’s fewer faceoffs than almost everyone else in the top-20.

It probably also has to do with the fact that Garino’s faceoff story is virtually synonymous with the run he had helping the Terps to the 2017 title. Across 4 NCAA tournament games, he won 27 out of 38 faceoffs, including 10 out of 14 in the title game against Ohio State. The narrative seemed to be “unknown FOGO comes out of nowhere to help a team down the stretch” rather than “4th best FOGO of all-time does his thing.”

The fact that Garino-at-4 is a surprise may simply come down to the fact that to-date, our only method for comparing FOGOs is win percentage. Compare Garino’s career numbers with the two guys on either side of him in the Elo table. He won 55.9% of his faceoffs, Alex Woodall won 62.4% and Zach Cole has, so far, won 66.8% of his chances. In fact, to find someone on the Elo list with a career win percentage lower than Garino’s, you have to get down to fellow Terp, Justin Shockey, at No. 11.

By conventional metrics, the narrative I laid out above isn’t wrong. And that is why I’m excited about the faceoff Elo model; it gives us a chance to objectively account for the difficulty of the situations these FOGOs are put in. And as a result, we can appreciate the skill that someone like Jon Garino brought to the position. (I’ll leave it to the Maryland fans to debate whether Charlie Raffa, career fElo: a respectable 48th, should have been the primary FOGO.)

Expected Win Percentages anyone?

Rankings are fun, as we all know, but how else might Elo ratings for FOGO’s make their way into the LacrosseReference ecosystem?

Well, for starters, I can now calculate a win probability for every FOGO in every faceoff, which means that we can calculate expected faceoff wins. And that means we can calculate excess faceoff wins as the number of faceoffs someone wins above what we would expect them to win given their opponent(s). Excess faceoff wins is an interesting metric because it can help assess whether a FOGO is improving or not.

Let’s say you are up against TD Ierlan in the (please lacrosse gods) 2021 season. Unless you are someone like Zach Cole, you are probably not going to be expected to win more than 20 or 30% of your attempts. Maybe you’ve been working on a new technique, and you manage to win 40% of the faceoffs against Yale. That’s really impressive, and perhaps a clue that the tactical changes are working. Otherwise, you’d win 40% of the faceoffs and say: “well, I did what I could.” That’s no way to learn.

I have also been playing around with the faceoff data by separating out faceoffs where the FOGO picked up the loose ball vs faceoffs where a wing grabbed the GB. Using the faceoff Elo ratings should give me a sense for what kind of faceoff win rate a team should have, given a certain opponent if only the FOGOs ever won a GB. By seeing what the actual win rate for the team was, we may be able to identify some sort of signal as to how good each team’s wing play on faceoffs is.

Anytime I can separate performance between the components that are included, I’m happy.

But for today, I’ll just have to leave you with the top 5, and the knowledge that there is now a better way to gauge who the best FOGO’s of all-time since 2015 are.