Pity Martinez is better now than he’s ever been

Pity Martinez as Rex Grossman

Immediately upon hitting the scene in Atlanta, Pity Martinez frustrated fans. He dribbled often and he was dispossessed often. While he would occasionally find a killer incisive ball to create a chance for a teammate, it was common to see him misplay an easy pass out of play or into an opposition counter attack. Most noticeable of all, he liked to shoot from really far away. And all of these shots missed because shots from really far away are shots that do not get past the keeper very often.

There was near unanimous frustration with Martinez that stemmed from the team’s results, from a mournful longing for Miguel Almiron, from the team’s style of play, from the player’s body language, but mostly from his lack of attacking output. It was rough, and to some extent, it culminated in the infamous Fox Sports Argentina interview in which Pity revealed publicly his disapproval with the manager’s approach to the game.

Somewhere in there, we started hearing a different slant from some corners of the media and the social media: the idea that Pity Martinez was indeed an elite talent, but that Frank de Boer’s tactics were cramping his style - that he needed to play in a team executing on a more improvised/free-flowing high pressure approach to the game. There was a small but loud portion of the community that almost suggested “Futbol in Argentina is just different, you wouldn’t understand.”

And I can’t stress enough how interesting the early to midseason Atlanta United dilemma was. While on the one hand, you had a player who was resented by large sections of the fan base for not living up to expectations, on the other hand, you had a manager who appeared to be resented by that very player for footballing differences AND also resented by large sections of the fan base (by many of the same fans that resented the player!). (!)

The Good Old Days

It was at this point that I started thinking about the original preseason Pity Martinez data scouting piece that I wrote. You don’t need to read it again - it’s long and no one needs to hold me accountable for the things I have written (although it is extremely pretty, thanks Joe). For your convenience, here is the scouting report from that post in bullet point format:

• Pity Martinez put up very, very high volumes of shots and shot assists at River Plate, and few players attempted/completed more 1v1 dribbles in Argentina. He had also crossed the ball more than any other Atlanta United player.
• MLS has a history of players coming from Argentina and increasing the volumes of shots and shot assists they contribute after making the jump (the MLS boost you might call it).
• Pity has always turned the ball over a lot and played speculative passes.
• Pity has always taken a high percentage of his shots from outside the box and that isn’t good. There were signs (weak signs in hindsight) that perhaps he was starting to move them closer to goal.
• I warned that a bad possible scenario would be a “Mephis Depay at Man United under LVG” type “high usage” player whose previous stats were boosted by constantly “using up” his team’s possessions with speculative shots and take-ons.
• But all in, I gave my vote of confidence that he was creating a ton of shots for himself and others, and that this was in Argentina, so his output in MLS would probably be even greater.

Looking back, I think most parts of that analysis, especially the warnings, hold up pretty well, and I had limited access to good data on his time at River. Having watched him now for over half a season and having seen the more advanced stats from MLS trickle in, it was time to see if I couldn’t look back at Pity’s history in Argentina with a better analytical arsenal.

Shooters Shoot

I want to go back to my preseason warning about his shot selection, because after some digging I think I have some equivalent expected goals (xG) data from Argentina that I didn’t before. The data in the rest of the article includes information from the Superliga Primera Division regular season competitions and does not include any data from Copa Libertadores tournaments. I won’t fully rehash what “expected goals” is but quickly, think of it as a count of shots weighted by each shot’s likelihood of scoring (shots closer to goal have a higher xG), and know that the rate at which a player’s generates expected goals is a better predictor of their future goal scoring output than the current or past rate at which they score goals.

Below is a graph of Pity’s 4 game moving average xG per 90 from open play from the 2016/2017 Primera season to today along with his xG per shot (a measure of a player’s shot selection) plotted as the gold line w/ dashed lines referencing the average MLS attacking midfielder’s xG.

Some takeaways:

1. The first 3-4 months of the 2019 MLS season look very similar to Pity’s 3 seasons in Argentina, from a shooting perspective. Pity averaged over 2 shots per game, all the while his shots averaged a 5% chance of going in based on historical shots from similar situations and locations (0.05 xG). And to be clear, this is a completely terrible figure for a playmaker. This is a player who, while perhaps always immensely talented, had a real problem with taking speculative shots from long range and he needed some coaching very badly. As a comparison, the average attacking midfielder in MLS takes shots with an average xG of twice that (gold dotted line in the above chart), presumably because it’s more natural to shoot when you’re more likely to score.
2. As the spring months gave way to summer and into late Summer in 2019, specifically right around the time Pity mouthed off to South American media, his shot selection started to improve drastically. In 291 minutes in July he put up 2 shots per game, with an average xG per shot of 0.17 (shots 17% likely to be scored based on historical data) and in August his numbers continued to improve. He has begun to shoot from inside the box rather than just wherever he received the ball. It should not be a surprise that having an attacker that consistently finds shots inside the box is critical for scoring goals at above average rates, so this is a great development for the club.

If Pity has in fact transitioned full-time from a guy who pops off from range on his way 0.13 expected goals worth of shots per game to a guy who gets into the box and generates closer to half an expected goal worth of shots per game (and this may well not persist), then Atlanta United might be in for some exciting times, and I have to imagine Frank de Boer (and perhaps even his approach to the game) deserves a little credit for this.

Pity Martinez hasn’t historically done assists

Pity Martinez was never a great setup guy. During his time at River, he put up a healthy 1.4 and 1.7 key passes (shot assists) per game from open play in the 2016/2017 and 2017/2018 seasons, respectively, but the average value of the shots he was assisting dropped from 0.21 to 0.12 over that period. In all, he averaged around 0.16 expected assists (xA) per 90 from open play, which is comparable to an average MLS attacking midfielder, not a South American Player of the Year.

At River Plate, the high volume of shots that Pity was assisting was often illusory as most of these shots were his teammates firing speculative attempts from long distance, just as he fired off speculative attempts from long distance when his teammates passed to him. The 2016/2017 year looks the healthiest but for the year plus leading up to his transfer, it really doesn’t look bright. This is something my preseason data scouting post obviously missed, and it’s why the analytics community likes xG and xA more than they do total shots and total shot assists (key passes) as units of account for their ability to cut through misrepresentations from pure volume. It’s a learning experience.

But enough looking backwards,... specifically at Pity’s history as an attacking midfielder at River Plate, because by the numbers, it was frankly less impressive than you were led to believe (by people like me and many others). Also included in the above chart is his time in MLS, which has been much more dynamic. Even starting the season out, Pity was creating high value chances for his teammates, and while this has been a bit noisy throughout the season, the most recent late summer months suggest he’s starting to put it together again.

Who is Pity Martinez?

Stats aren’t everything. But other than having watched a few highlight reels where Pity takes on a 35 year old defender in the Superliga and rips a long shot into the corner of the net, I’m kind of at the point where I’m not sure what his dominance in Argentina actually looked like. And in fact, the best predictive attacking numbers we have, suggest the Pity you’re seeing now is pretty similar to the Pity that played for River Plate, and perhaps that he has improved in a couple of facets this year with his ability to set up teammates for good shots. In the last couple of months, something has caused him to cut out the bad shots. If we look at the full trend line of Pity Martinez’ career at River Plate and Atlanta United (I’m missing the 2016 season admittedly), it appears as though in the last couple of months, we have been seeing the best Pity Martinez there’s ever been, and I think that’s a significant finding here, perhaps one that should rearrange the context around which we discuss the guy and his efforts.

You can say that the team has switched to the 3-5-2 at the very same time Pity’s form has peaked — and you’d be right, and then you’d probably suggest some idea that perhaps all Pity needed was for the team to run and gun like the days of Tata Martino — but that would fail to account for the fact that Pity’s propensity to shoot from bad places looked very similar for the first half of 2019 as it did in several seasons in Argentina. Similarly, his prowess in setting up open play chances for others, while notable in the 2016/2017 Primera season was absent the entire 2017/2018 season into 2018/2019, and has really only awoken at moments in 2019 in MLS. I should also caution that soccer is hard, and a month or two of good form, even against some of the best teams in MLS, is just not enough to be sure of anything.

Conclusion

A wholesale hindsight revision of the Pity Martinez transfer analysis might suggest that Atlanta United mightily overpaid for a high volume, high usage attacking player in his prime, and that this may not have been the most prudent thing to do as Tata Martino was on his way out of the club. As bloggers/fans/tweeters, perhaps the lesson is never trust the youtube highlights. However, the data suggests (and it’s early still) that the long-awaited “MLS Bump” that so many Argentina stars experience upon arrival in the states is finally coming good for Pity. In the later summer months, his output in creating shots for himself and others has peaked significantly above his historical trends, and perhaps most importantly, after a career of “F&*k it I’m going deep!” shot selection, he has finally settled into some kind of discipline about the types of shots he is taking - which is beneficial for the overall team attack. Atlanta United has too many dangerous attacking players to be ending solid strings of possession with 5% chances when they could find better ones, and Pity might finally be stepping up to do his part at cutting out the garbage shots and more importantly, making runs and carries into the box to find some juicy shots for himself and his comrades. In the end if Pity Martinez is not an MVP caliber player in MLS, he still looks to be on a trajectory to being a guy no one wants to face in a playoff match, much like Tito Villalba.

...

Post Credits Teaser Scene: Turnovers

I find it very easy to fall into the trap of watching Martinez play, and watching him lose the ball over and over again, and think “What in the hell is going on. What happened to the South American player of the year?” I think the truth, is mostly, that this is just who he is, and he’s always been a guy who turns it over a lot - remember Yamil Asad?. The chart below shows me that there’s nothing new to the turnover aspect of his game that’s popping up in 2019 on it’s own. I think he can improve in this area, but this is the guy they went out and signed. Perhaps his inclusion in a front 2 partnership with Josef will limit the risk of these turnovers to the team in transition.