Bob Bradley: Possession, Pressing, and Personnel
/By Kieran Doyle-Davis (@kierdoyle)
Heading into Toronto on a snowy December day in past years may have had visitors buzzing about the state of BMO Field’s snowy pitch in preparation for an MLS Cup final. For now, there is just roster talk. Roster talk, and manager talk. Enter Bradley. Not the old Bradley we’ve come to know and love, but the even older Bradley who is new to us. Toronto FC signed new head coach and sporting director Bob Bradley who departs LAFC after they missed the playoffs for the first time, despite having had the best xGD in the entire Western Conference. While there is some skepticism about Bradley’s hire (missed the playoffs, MLS retread, Swansea blah blah blah), I am really excited to watch this team under him in a way that I just haven’t been about the Armas-Perez version of this team. As such, I thought I’d break down some of the key things Bob Bradley did at LAFC last season and how they apply to Toronto FC, what went wrong, and what I think he might do with the team here. Normally we would call this something of an “Offseason Outlook”, but we can kind of mish mash Toronto and LAFC’s together here.
Fourteen, Fourteen, Fourteen, (Zone) Fourteeeeeeen
Given the glut of attacking talent in Toronto that has looked decidedly not gluttonous for the past nine months, I think it makes sense to start with how LAFC put together an attack. In broader terms, LAFC ran the 2nd best xGD in the league at 0.61 xGD/90, second only to NYCFC, with a league best xG for at 59.22 xG. Add in that best-in-class attackers Diego Rossi and former MVP Carlos Vela played less than half of the available minutes, and that is very impressive. If we think back to 2019, Vela had the highest g+ “burden”** of any player in MLS since 2013, that is the most value created landed at his feet on any given possession, before being passed along and added to by Vela. Vela had the 39th highest total g+ burden of all players in MLS this year, with his total g+ burden dropping from about 82 g+ all the way down to 25. In simpler terms, in 2019, Vela’s teammates put him in incredibly dangerous positions and he made even more of them. In 2021, he wasn’t getting the ball in nearly the same sort of positions, and his output suffered in kind. All this to say, Bradley had an elite attack built around one of the best attackers in league history, and then made it keep working even as he was unavailable and fading fast.
**Reminder: g+ burden is the addition of all possession prevalues for every time a player touches the ball. For example, if Carlos Vela gets passed the ball in an area where LAFC have a 10% chance of scoring, and a 1% chance of conceding on the next page, he faces a g+ burden of 0.09. We can sum this over all touches to try and determine how heavily a player is used by their team. For more info on usage, read here
Here, g+ burden is the sum of prevalues coming into each player and production is just their raw g+.
This attack largely came through a domination of zone 14. LAFC led the league last season in passes attempted, completed, and progressive passes into zone 14. They also led the league in g+ per match generated in zone 14. When you plot LAFC’s high g+ value passes (greater than 0.025 g+) into and out of this magical dangerous pocket, it’s easy to see why. Vela and Rossi often drifted inside from their respective halfspaces, receiving between lines to attack the center of the pitch.
This shows the MLS team with the best g+ difference in each of the 30 zones on the field, attacking from left to right.
As a quick aside, ASA divides the pitch, for the sake of g+, into 30 zones, so the traditional zone 14 (the pocket at the top of the box) is zone 23 in the interactive tables if you would like to look for yourself. Most coaches and analysts refer to that pocket still as zone 14, so I will continue with that terminology.
This shows all LAFC passes coming into and leaving zone 14 with a g+ value greater than 0.025. Look at all the beautiful danger created from there.
Take some of the actions in the embedded video: the winger finds the ball in the halfspace, the center midfielder or striker fills zone 14 as a connection between both sides of the field and makes forward progress while the opposite winger occupies the fullback, then the wingback or fullback overlaps aggressively. In Argentina, they use the word enganche (a hook) to describe an attacking midfielder, a 10 who connects the midfield and attack. For Bob Bradley, anyone can be the hook, so long as they fill the right space.
| LAFC League Rank 2021 | |
|---|---|
| 2nd | |
| 1st | |
| 1st | |
| 1st | |
| 1st | |
| 1st | |
| 5th | |
| 5th | |
| 2nd | |
| 2nd | |
| 24th |
Now, LAFC’s attack wasn’t perfect, Vela and Rossi both underperformed their xG last year as the team put up 51 goals for 59 xG, but not in any clear sustainably negative way. They did, however, put up solid attacking numbers, even outside of zone 14. Bradley’s LAFC led MLS in both non-interrupting goals added difference (+18.25 and a very solid predictor of future expected goal difference), and in passing g+. On the more basic side, they led MLS in shots with 15 per match and had the best xG per shot according to StatsBomb via FBRef and 5th best according to ASA. They were 5th in passes into the final third, second in passes into the penalty area, 2nd in progressive passes, but had the fourth fewest crosses into the box. You will see a similar vomit of league ranks in various metrics when we talk about the LAFC press, but the point is no matter how you slice it, this team was good. Standings be damned!
TFC on the Geigen-check
Much was made about last year’s incoming hire, Chris Armas, and how his radical Red Bull-ness was going to overhaul the organization. Toronto FC was going to press in a way it hadn’t before, full of piss and vinegar and a willingness to get in people’s faces, they were going to play the kids, and it was going to be fun and exciting. That lasted like six games? The kids played and somehow beat Leon in maybe their best performance of the season. Noble Okello, the pressing 10, and a team held together by duct tape had a sniff at a CCL run before coming back to Earth and crashing hard, costing Armas his job. Javier Perez steadied the ship for the rest of the year and did his best to make sure nothing got too ugly as Toronto FC sank to the bottom of the Eastern Conference.
Quietly, Toronto FC tried to press, even after Armas left. They were upper mid table in final third pressures, 7th in opponent passes under pressure. They were also, unfortunately, 3rd bottom in xG conceded and 2nd bottom in goals conceded. It didn’t work. Omar Gonzalez and the rest of the CB corps were regularly caught out by passes in depth when forced to consider whether to step and join the press, or cut bait and protect depth. Chris Mavinga had his worst season in Toronto, regularly sidelined by injury, and unable to appropriately cover for the morass of mistimed steps and unchecked runners ahead of him. As much as TFC tried to press, they were scythed through at almost every opportunity.
The colored trendlines here are shots that occur from specific kinds of passes, take note of how much value Toronto FC afforded other teams on progressive passes. Those kinds of passes advance up the field quickly and directly - think a through ball to break a disorganized press. The green bubbles are the xG conceded by progressive passes for each game of the season and the white trendline is the three-game moving average. This is all to say, TFC were B.A.D.
| LAFC League Rank 2021 | |
|---|---|
| 2nd | |
| 3rd | |
| 2nd | |
| 3rd | |
| 2nd | |
| 2nd | |
| 2nd | |
| 5th | |
| 3rd | |
| 1st | |
| 2nd | |
| 2nd | |
| 1st | |
| 1st |
On the flip side, Bradley’s LAFC side were one of the pressiest outfits in MLS last year. Here’s some data courtesy of our friends at Second Spectrum: LAFC had the second most individual pressures per match at 303, the third most distance closed down at 72.5 km, the second most turnovers generated by pressure with 81 (2nd only to Red Bull’s gargantuan 89), and the third most turnovers generated by pressure in the opposition half. The StatsBomb/FBRef public data backs this up: opponents completed the 2nd most passes under pressure in the league (behind NYRB), opponents played the 2nd most passes over shoulder height (behind RB), they intercepted the 2nd most passes, forced the 5th most passes out of play, 3rd most passes blocked, the most attacking third tackles, the 2nd most defensive actions leading to a shot attempt, and the 2nd most attacking third pressures. There are a lot of numbers there, handily summarized in the table below, but take all of them to say this: LAFC pressed, and they pressed really effectively.
Now, at the same time, pressing to press is not the be all and end all. LAFC were able to press their way to the 8th best xGA in the league. Here’s the same chart from above but for LAFC (look, they don’t have a nickname, I’m sorry you’re stuck with reading LAFC 214 times). Similarly, they were able to deny teams the ability to progress with a respectable 9th best g+ passing against, and 3rd best g+ receiving against. The press was effective, but they were able to recover into a block as well.
Why yes, all of those numbers are much lower than Toronto’s.
LAFC could really defend, and defend in an aggressive way that let them get their attack going, and they didn’t do it with good old fashioned farm boys from New Jersey, they did it with designated players and high priced attackers. The same kind of players who have found it, let’s say difficult, to defend with the group in Toronto. Now, you may notice that everything with LAFC has been about their xG conceded and not regular old G conceded. We’ll talk about it later, but the discrepancy was mostly just rancid goalkeeping and tough luck, which shouldn’t really be a problem moving forward.
So, how does Bradley set up his team to press? While Armas was fairly interested in the Red Bullian 4-2-2-2’s and 4-diamond-2’s to create numerical superiority in central areas, Bradley has shown quite a bit of flexibility with his LAFC sides - particularly in the last year. They started the season in a fairly standard 4-3-3, with one clear pivot in possession, flipping the triangle into more of a 4-2-3-1 structure when initiating the midblock press against some opponents, leaving the single pivot against others. By the midpoint of the year, LAFC were playing a 3-5-2 with a clear number 10 (generally Cifuentes) behind two strikers. While the ins and outs of who stepped where in each system varied (for example, whose responsibility is the opposition fullback when pressing out of a back three or a back four? Is it the same on the first side as on the switch?), there were some key tenets no matter what. Take the video below, here are some representative clips of LAFC’s press, across all the months of the season, in both a 3-5-2 and a 4-3-3. There are 3 key situations to look out for:
How do the midfielders react to someone receiving the ball with negative (facing their own goal) body orientation?
How do the front 6 or 7 players react to a backwards pass to anyone other than the goalkeeper?
How do the secondary and tertiary players react when LAFC lose the ball?
(1) and (2) here are clear pressing triggers for LAFC. They aren’t the only ones you might notice in the video, but you see a lot of similar type actions. The front line blocks passes wide, opposition defensive midfielder shows with negative body orientation and receives, LAFC midfielder steps aggressively. Sometimes this is results in a tackle, but a lot of times it forces a pass backwards which now engages trigger #2. The midfielder stepping out has momentum and can “flow on” such that their cover shadow takes away the opposition defensive midfielder, and they can aggressively close the ball. The opposition build is now under quite a bit of pressure and you see some long balls, some turnovers, some resetting all the way back to the GK.
These aren’t crazy groundbreaking concepts, but they are cohesive and well executed – much in the same way as the LAFC counter-press. LAFC ranked second in the league in successful (turnover, clearance, ball out of play) pressures within six seconds of losing possession according to Second Spectrum. While Greg Vanney’s Toronto FC wasn’t the most pressy team you will have ever watched, they did have a fairly robust counter press to win possession back quickly and get the fun possession train rolling again. It is fairly safe to assume that Bradley will apply similar concepts in Toronto, given the amount of possession talent on the roster still.
The Personnel
Toronto FC faced their fair share of roster maladies last season. The only real striking options in Jozy Altidore and Ayo Akinola both missed most of the season due to injury, while Fabrizio Romano has a “Here we go!” ready to hit tweet on Altidore to Turkey before the start of next season. Former MVP and heir to the “Build an attack around Carlos Vela” throne Alejandro Pozuelo was similar. The only designated player who did play consistently was Yeferson Soteldo to fairly middling success. There are some green shoots from Jayden Nelson and Jahkeele Marshall-Rutty but neither really got the type of run you would have hoped for in such a down year. Both teenagers profile out as better than replacement level by g+, which is a pretty good starting point for players so young. Sneaking under the radar is Jacob Shaffelburg, who very quietly put up 0.63 xG+xA per 96 with a fair few minutes at left wingback. Keep an eye on where he ends up playing next year. The fact of the matter is that Toronto FC have to significantly improve in attack for Bradley to really take advantage of the talent in the squad. LAFC faced similar availability issues last season, Carlos Vela has only played about 1500 minutes over the last two years and Diego Rossi left on loan partway through this year, but LAFC have four players who all would’ve been TFC’s most xProductive attacker. This is not the same glut of attackers available to Bradley in Los Angeles.
Here is the joint attacking wheel of the de facto front 3 for each club. Given injuries and availability etc. these front 3’s played virtually zero minutes together, but more gives you a sense of the gulf in attacking output between each group.
In midfield, the literal prodigal son (the son part, not the prodigal) Michael Bradley looked serviceable last year and certainly better value on TAM than on a DP deal, but there are still significant questions over how you build a midfield around a single pivot who can’t really press evade and can’t really anchor a press himself. On the bright side, Osorio and Delgado have both been really positive and serviceable midfielders in possession based systems in the past. Osorio in particular has thrived as both a wide attacker drifting in from the half spaces, and as a pressing eight between the lines. Ralph Priso offers another mobile option with dynamic ball winning and serviceable possession retention. But much like the front line, a lot of these pieces just were not very good last year. When you compare to the glory and grace of God that are Cifuentes, Atuesta, Latif, Ginella, Mark-Anthony Kaye, etc. it really isn’t close in terms of talent. There is certainly a lot of work to do from Bill Manning and Bob Bradley to overhaul this roster.
In defense, there’s some clear talent that should fit the Bradley ethos. Laryea and Auro are both solid attacking fullbacks. Laryea in particular has proven himself a really dangerous attacker on the overlap, even when playing out of position on the left. Auro provides useful passing as more of a midfielder and fullback hybrid than the wide attacker that Laryea is. On the left, Shaffelburg is a demon at attacking space and whipping balls across the face of goal, while Lawrence givesmore balance in possession and more defensive presence. While at fullback Toronto FC might have the edge compared to Bradley’s old stomping grounds, at centerback that flips on its head. Gonzalez and Mavinga, the de facto first choice pairing, had really rough seasons despite starting strong. Luke Singh faded quickly, and none of Dunn, Zavaleta, or any other prospects look quite ready to fill a significant role at the back. LAFC were similarly hodge-podge in terms of bodies, but got really good minutes out of Fall, Segura, and Murillo across the season. This is, and has been, the most important area for recruitment in Toronto for three consecutive seasons.
High value passes and box entries from Toronto FC wingbacks in 2021.
High value receptions given up by Toronto FC in the box. The centerbacks were a huge problem for The Reds last season.
Last but not least, let’s talk about goalkeepers. Most of LAFC’s xG pummeling last year came off the fact that for 75% of the year they were playing bad goalkeepers. Sisniega can play but was decidedly not good before he got injured (G-xG of roughly two), and Tomas Romero was one of the worst goalkeepers in the league at +6.5 G-xG in his 1800 minutes played. Jamal Blackman came in and steadied the ship, and sure enough the crazy goals conceded (mostly) stopped. Toronto FC are in a weird position in goal, but less weird than that. Alex Bono and Quentin Westberg have battled it out for the starting role for the past two seasons and it’s fairly clear that Westberg is just better. Statsbomb has Westberg as a league average shot stopper, which is fine, but he is a fairly positive and progressive distributor and a much more active sweeper when compared to Bono. Bono is a much better cross taker, but the more conservative sweeping, more agricultural distribution, and worse shot stopping make Westburg a much better fit for Bob Bradley’s style here.
Data courtesy of StatsBomb via FBRef. StatsBomb’s freeze frames on shot data generally provides a more complete picture of goalkeeper shot stopping data.
Welcome to Toronto
All in all, this is an interesting marriage. There are a host of fascinating off field narratives to this team and this scenario, but the on field ones are maybe more exciting. Bradley should feel hard done by at the end of his time in Los Angeles. They crushed the league like five years in a row and walked away with one Supporters Shield and not a whole lot else. Toronto FC are sliding towards where they were 10 years ago instead of five years ago. With a roster rebuild on their hands and a relative superstar aging out of their prime in Alejandro Pozuelo, Bradley might be just the right coach for the job.