Game State and the 2019 Packers: why the regression may not be as severe as expected

Some of you may like soccer. Others of you may really hate soccer because of the United States (but please take time to see our golden boy, Christian Pulisic, tear apart the English Premier League). Regardless of your views on football, there is a concept that I think can be helpful in describing the 2019 Green Bay Packers and, in turn, helps us forecast a bit for the 2020 Packers.

Before entering American football, I have to explain some soccer statistics and concepts. First, we will start with the expected objectives, or xG. xG is a fairly simple statistic: it measures the probability that any shot will turn into a goal. The closer and more focused you are on the target, generally the higher the xG. A five-yard tap-in has a better chance than a 25-yard shot. If a shot has a 20 percent chance of being a goal, the xG for that shot is 0.2. If you have a 5 percent probability, the xG for that shot is 0.05.

In each game, you can add up the xG total for each team to see which team “made the best shots / took the best shots.” Soccer is a low scoring event game. Absolute burners like the Liverpool / Arsenal Carabao Cup match last fall still only had 10 goals, which is very low compared to something like basketball. However, football is also a relatively low-scoring event game. With this, both have a high degree of variation within the individual games.

The panoramic view of expectations

This brings me to a larger image in xG. Using xG for an individual game is only slightly useful. There is still too much variation. But if we drift away over the course of a half season or a full season, xG and its sister statistic of Expected Goals Allowed (xGA) can tell us more about a team than their actual goals / allowed goals.

In general, the best teams will have the highest expected target differential or xGD. This works most of the time. For example, in the 2017-18 season of the English Premier League, Manchester City racked up the staggering 100 points, racking up 32 wins with just two losses and four draws. They also racked up a hilarious xGD of +58.3. There are times when it doesn’t work as well, such as the 19-20 season of the Premier League. Liverpool have already secured the title with 93 points, 21 ahead of Manchester City. However, the city did not get worse: they have a +53.7 xGD, and Liverpool’s is only +28.5. What’s going on?

This is not intended to be a deep dive into soccer statistics, but a concept of soccer may be applicable to soccer here. Liverpool has only faced eight deficits throughout the year. They spend the vast majority of their time tied or leading the game. The next closest is Manchester City at thirteen. What teams do when they are active also matters. Liverpool does a phenomenal job of controlling the games. When they are up, they don’t drop points. Part of the reason why his xGD might not be as outstanding as Manchester City has something to do with the fact that the City team is a steamroller of attacking beauty, but it also has a lot to do with the fact that Liverpool doesn’t spend that much time needing to put the pedal down.

Liverpool is not even the best example of this. José Mourinho’s teams in Chelsea often “park the bus” after gaining an advantage, and spend the remaining time defending themselves, essentially spending entire halves playing “prevent defense”.

But what about pigskin?

That finally brings me back to American football. And I want to talk about the Green Bay Packers and their own “xG”. It’s no secret that the Packers’ point differential in 2019 was pretty mediocre for a team with a 13-3 record. I am not here to argue that this was a truly talented 13-3 team. It was not. They had a pretty easy schedule of quarterbacks who helped a lot and were good at one-point games, both of which are primaries for the regression in 2020. Their point differential ranked fifth in the NFC last year at +63. The gap between Green Bay and team # 4, Minnesota, was 41 points. The Packers’ Pythagorean record, operating outside the differential was 9.7-6.3. The Packers were a hilarious 9-1 in one-score games.

To summarize: A lot of data showing Green Bay’s record last year was a fluke. And that doesn’t seem to be wrong, but how correct it really is is quite important.

The big accountant for all of this data is a look at NFL games by probability of winning, rather than points, from FiveThirtyEight. Using the probability of winning instead of the point differential, they projected Green Bay for a 12-4 record. That is still worse than his actual record, but no more than three wins worse.

Controlling the game

I presume that this difference between the two has a lot to do with the script of the game. The Packers often did a great job getting to the leaders early, ranking fifth in first-quarter points per game and sixth in opponents’ allowed points in the first quarter per game. From there, it was often a “wait for life” situation. To illustrate this, here are the ESPN victory probability tables for each of the GB regular season games of a score in 2019 (Note: a score is defined as eight points or less):

While GB had an incredible number of single-score games, based on the probability of winning, the games weren’t as close as scorelines show. Green Bay was lucky to get both games out of the Lions out of a hat, but there are also games against Carolina and Washington where the final score feels closer than the games for probability of victory.

So why did I spend several paragraphs talking about soccer and providing you with a bunch of photos for you to see? To talk about the state of the game. When a team is winning, they are more likely to play conservatively than when they are losing. Approval rates decrease with an advantage. Passing is more efficient than running. If you do less efficiently, the game is likely to get closer. As you can see below, Green Bay was not afraid to throw the ball in situations where the game was still within reach.

However, when GB gained an advantage, they ran the ball a bit more. Green Bay’s passing rate in two scoring games (9+ points) in the 1st and 2nd down fell to 14th. What Green Bay did was not the football equivalent of “Park the Bus,” but they spent a fair amount of time in those settings. Green Bay spent the sixth most offensive plays in the NFL with an advantage. However, Green Bay was not as good at playing offense with an advantage, particularly in the passing game. Green Bay’s 47% success rate in passing with a two-point lead was well below the league average of 52%. His number of yards per attempt also seemed poor at 6.6 compared to the league average of 7.6 in those situations. Rodgers averaged -.04 EPA per play in those situations as well. The Packers’ offense was bullshit once it went ahead. For comparison, Rodgers’ EPA per play on pass attempts in one-shot situations was +.11, and when he was behind by scores of more than two it was +.12.

I don’t know if Rodgers is intentionally too conservative with an advantage, Matt LaFleur’s game becomes too conservative with an advantage, a reflection of not being so good, a myriad of other possibilities or a combination of all that, but the Packers need power. step on the opponent’s throat when they have them down. Parking the bus may have worked in 2019, but it’s asking a lot from the Gods of soccer year after year.

The Packers certainly weren’t a true 13-talent winning team last year, but they may have been a little better than their Pythagorean records. The prediction of the use of “closed games” based on the probability of winning has not yet been seen, but it may indicate that, although some regression is inevitable in 2020, that the fund will not fall on this team.