The Case Against Protecting a Lead With a 3rd CB
The Case Against Protecting a Lead With a 3rd CB
Taking a Look at Data Trends in Revs Games With a 3rd CB Substitution
A lot has been made of the Revs recent draw vs FC Dallas.
New England, who sit in last place in the Eastern Conference, held a 1-0 lead late into the match. More than that, they looked like the better team.
They were, to that point, out-shooting Dallas by a margin of 12-3 and out-possessing them by a margin of 56-44%. After a disappointing loss earlier in the week, a full 3 points was sorely needed for a squad looking to work their way back toward the playoff picture.
Then, in the 81st minute, Caleb Porter subbed on center back Jonathan Mensah for midfielder Tommy McNamara. A move clearly intended to pack-in and help secure the win.
Fast forward to the end of the game and the Revs found themselves with points left on the table, conceding a penalty kick in stoppage time after enduring wave-after-wave of pressure from the visitors.
It’s become a common story for the Revolution since they semi-righted the ship in June: Look like the better team - take a lead - sit back - sub in defenders - try not to lose.
In fact, in eight of their last 10 matches, Caleb Porter has subbed on a center back in the second half. Or at least a player who has primarily played center back.
On four occasions Andrew Farrell entered the match late, often directly taking the place of a defensive midfielder or outside back. Farrell has historically played centerback, but his versatility allows him to play other positions when needed.
Mensah has also made five substitute appearances in the last 10 matches, always subbing in for a midfielder - always as a 3rd center back.
That’s the scenario I would like to talk about today. The late addition of a 3rd center back, substituted into the game in order to try to secure a result.
The eye test seems to indicate that it’s maybe… not the best idea. But New England has been successful while using that tactic. In 5 games, with a late sub of a 3rd center back, the Revs are 4-0-1.
The Case For New England Subbing Central Defenders Late
Listen, I’m going to focus a lot on why I don’t think this is a valuable strategy for the Revolution.
But it makes sense.
You want to make sure you don’t concede a goal? Throw another defender onto the field. Put another obstacle in front of your goal.
Increasing the number of bodies in the defense can help gum up the works for opposing attacks. Adding a center back often provides an additional benefit on set pieces, as aerial dual win percentage is a stat traditionally dominated by CBs.
Committing to a low-block and playing compact defensively isn’t a new strategy. And it can work.
You might also make the argument, if you so choose, that even if it doesn’t really work it’s the only viable substitution strategy for Porter to make - given the Revolution’s lack of attacking depth.
The Revs have dealt, in recent weeks, with injuries to Carles Gil, Dylan Borrero, Tomas Chancalay, Ema Boateng, Nacho Gil, Giacomo Vrioni — the list goes on. That’s a large amount of attacking talent that remains unavailable for selection, testing the Revs attacking depth.
In some way, the only sensible substitutions are defensive minded subs.
The Case Against New England Subbing Central Defenders Late
But in a different, much more real way - you might be better off doing nothing at all.
Every game plays out differently, so I won’t go so far as to say it’s a bad strategy. There may be times where it’s the correct move.
But we’ve seen the Revs sub on a third center back in five of the last 10 games and, while it has only cost them once, it hasn’t looked great.
For one thing, it represents a tactical shift into a different formation. Player’s roles and responsibilities change, positioning changes, communication changes. These players are all pros, and we should expect that they can handle the nuances of switching formations mid-game.
However, New England doesn’t start the game with three center backs on the field very often, and in their most recent match against Philadelphia we can see why…
But let’s focus on the five games in which a third center back entered the field - to help close out a lead.
New England ended up winning four of those matches, but conceded a late goal in three of five That’s likely due to the cascading effects of taking a more defensive posture.
When the Revs add a third center back and pack it in, the first thing you may notice is how little they see of the ball.
During those five matches, the Revs went from holding an average of 49.76% possession with two CBs on the field to just 28.72% after subbing on the third CB.
This isn’t totally surprising.
If the goal is play compact and set up in a low block, that naturally invites more space and therefore easier possession from your opponent. And with fewer outlets in the midfield, you do tend to see fewer possession passes from the Revs, and more blind clearances.
Now, conceding possession is fine if you can limit your opponents ability to shoot by playing smart and disciplined defensively.
But…
That hasn’t really happened for the Revs. After subbing on a third center back they allow their opponents to shoot nearly twice as often. With two center backs they allowed a shot once every ~six minutes, on average. That drops to once every ~three and a half minutes with the addition of a third CB.
Again, not terribly surprising.
New England has had difficulties preventing shots this season, allowing the second most shots-on-target per 90 of any team in MLS.
Don’t get me wrong: conceding shots is fine, if you can ensure they are low quality chances taken from poor positions on the field.
But…
That hasn’t really happened for the Revs.
New England already has the fifth highest expected goals conceded value in all of MLS. In the five game sample when a third center back enters, they appear to concede xG at a much higher rate.
Before the substitutions, they would have been expected to concede a goal once every 64 minutes. That dipped to one xG every ~29 minutes when bunkering with three CBs.
To further hammer home this point, if we extrapolate the xGA per minute onto a per-game scale, that tactical shift looks pretty bad.
Sample size is, of course, a huge caveat here. So take it with a grain of salt, but a pace of 3.13 xGA/90 when defending a lead doesn’t feel like a sustainable plan.
If I were being totally fair, the addition of a PK (against Dallas) during that small a sample size is doing a lot of heavy lifting. So lets take PKs out of the equation and look at non-penalty xGA and see if that helps the three CB bunker’s case.
I mean… it helps a little?
The Revs still give up non-penalty expected goals at twice the rate when adding a third center back. It’s 2.2x the rate when including penalty kicks.
All that said, conceding expected goals is fine if your defenders and goalkeeper are great at preventing goals.
You all know where I’m going with this. They’re not great at that.
The Revs have the 7th highest team GA-xGA, and the 9th highest GA/xGA. That’s to say, over the course of the season, that their opponents are scoring more often than xG would estimate. And that they are doing so, against the Revs, at the 9th highest rate in the league.
So What’s Your Point?
At the end of the day, every game plays out differently. There may be situations where subbing an additional defender to help hold a lead is the right choice.
But in recent history, when New England has employed that strategy late in games, it hasn’t worked out as well as it could.
After adding a third center back to the mix, the Revs conceded a ton of possession, which in turn lead to a higher number of shots. Those shots were from areas of the field that constituted a substantial increase in opponent’s expected goals pace.
And, while it only came back to bite them in one of the five games I looked at, New England did concede a late goal in three of those five contests.
With all the injuries the Revolution have had to deal with this season, it’s hard to say what they are when fully healthy.
But I think we can say that they aren’t a particularly successful bunker-and-counter team with the players that are currently available, and employing that strategy might be hurting their chances.