By: Lydia Murray | Follow Me on Twitter @lydia_murray12
Perhaps my favorite type of hockey analytic graph is the good/bad//fun/dull graph. Once you know what you’re looking at, they’re easy to read and display a ton of information. However, not everyone knows exactly what they mean, and so I thought it would be good to write an introductory article to them before I start really analyzing them.
The number one thing that seems to throw people off with these graphs is what exactly each classification means. Everyone knows the concept of good and bad, but what are the others telling us? What exactly does good and bad tell you about the team’s performance relative to their opponents? I’m going to explain each of these classifications relative to actual graphs first, as I think it makes it much easier to understand. I’ll then provide a general overview of how to read a team and player graph at the end. So, without further ado, let’s dive into the explanations of these graphs.
A Brief Explanation of How the Graph is Set Up
Before I begin explaining which each classification means, let’s go over how the graph is set up. The x-axis of each graph increases from left to right, while the y-axis is inverted and decreases as you go up. This is primarily because it makes the graph more pleasing to the eye, and therefore easier to read.
If you did not invert the y-axis, you would have “good” and “dull” on the bottom of the graph, while “bad” and “fun” would be on the bottom. This is confusing to the brain, as we are taught to look at graphs where the more to the upper-right of it you are, the better it is. Meanwhile, the more to the lower-left you are, the worse it is. So, it makes sense to invert the y-axis, as it’s the only way you can have “good” in the upper right corner, and “bad” in the lower-left corner.
Good/Bad/Fun/Dull Relative to Team Expected Goal Rates
Starting off with a team graph, in the above embedded tweet, you’ll find a good/bad/dull/fun graph of a team’s expected goal rate. This graph concerns teams that made the 2020 Stanley Cup playoffs and is as of the games on August 30th, 2020. The teams you see that are grayed out are the ones that have been eliminated from the playoffs at the time of this graph, while the ones that have color are still in.
Moving on to the explanation, the x-axis of this expected goals rate graph concerns the expected goals for per 60 of the team (xGF60). That’s the number of goals each team is expected to get per 60 minutes. The y-axis is the expected goals against per 60 of the team (xGA60). That’s the number of goals let up by each team per 60 minutes. The x-axis is from 1.6 to 3.2 xGF60, while the y-axis is from 3.2 to 1.6 xGA60 since it’s inverted. I will be referring to the xGF60 as goals per game (GPG) from this point on, as it’s easier, and unless the game goes to OT, 60 minutes is the length of a game (as you already know). I’ll be referring to xGA60 as goals against per game (GAPG) for the same reason.
Good and Fun
A team is considered “good” if they score between 2.270 and 3.2 GPG, while only letting up between 2.293 and 1.6 GAPG. Obviously, this is the part of the graph teams aim to be in, as it means they’re expected to score more goals than they’ll let up, which obviously means they’ll win the game.
A team makes its way into the “fun” part of the graph if they score between 2.270 and 3.2 GPG, while letting up between 2.293 and 3.2 GAPG. Fun is a relative term here. It means the teams are fun to watch, but at the same time, fans are probably having minor heart attacks while watching the games. It means the games are high-octane offensively, but typically not good defensively. It’s not necessarily a bad thing to be in this section as long as a team is expected to score more goals than they’re expected to let up; however, for obvious reasons, it’s certainly better to aim more for the “good” section.
Bad and Dull
“Bad” teams are ones that only score between 1.6 and 2.270 GPG, while letting up between 2.293 GAPG and 3.2 GAPG. It goes without saying that this is the section of the graph to avoid at all costs. It means you’re expected to get scored on more than you’re expected to score, which obviously means you’ll lose the game. It’s deeply unfortunate albeit unsurprising that the Bruins find themselves a little bit in this section, especially given the Game 3 fiasco.
Finally, we have the “dull” teams. These teams are the ones that are scoring between 1.6 and 2.270 GPG, while letting up between 2.293 and 1.6 GAPG. Dull teams are the ones who are usually rather boring to watch, as they aren’t scoring a lot of goals and they aren’t having many scored on them. Like the fun section, it’s not necessarily a bad thing to be in it; however, it’s much better to be in the good section. Dull teams are often playing in tight, low-scoring games, which again isn’t necessarily bad, but it shouldn’t be the goal either.
Good/Bad/Fun/Dull Relative to Player Expected Goal Rates
Moving on to a player graph, in the above embedded tweet, you’ll find a good/bad/dull/fun graph of players’ expected goal rate. This graph concerns players in the 2020 Stanley Cup playoffs and is as of the games on August 30th, 2020. I’ve configured it so just Bruins players are highlighted, as those are really the only ones we care about here at BNG.
Good and Fun
A player finds himself in the “good” portion of these graphs if he’s expected to score between 2.315 and 4.5 GPG, while only being on the ice for between 2.293 and 1.0 GAPG. Just like with the teams, obviously this is the section a player wants to be in. However, it is not realistic to assume every player will get here. It all depends on usage. So, these graphs should never be considered without context as to which line or pairing each player is on, and therefore their usage. If they’re performing as expected, the context of their usage matters more than their placement on the graph. You can’t expect a fourth-liner to be in the “good” section, for example, just as it’s a horrible sign if a first-liner is in the “bad” section.
For example, you’d expect top line players like Patrice Bergeron, Brad Marchand, and David Pastrnak to all be in the “good” section, or at least at the top half of the “fun” section (more on that one shortly). Thankfully, they all are. That’s how you expect them to be: great offensively as well as defensively, and so on the ice for more goals for than against.
As for the fun section, a player will land himself here if he scores between 2.315 and 4.5 GPG, while also letting up between 2.293 and 4.5 GAPG. Much like with the teams’ graph, fun is a relative term. It means a player is on the ice for a lot of goals for, but also for a lot of goals against. It’s fun to watch, but also not good for fans’ blood pressure. Also, just like with the “good” section, usage matters more than their placement on the graph, provided they’re performing as expected.
Playing style also has an impact on where players place on the graph, primarily in a section such as “fun” or “dull.” This is because if a player is high-octane offensively, but not overly solid defensively, they’re bound to end up in the “fun” section because the team should score a ton while they’re on the ice, but they’re also going to get scored on a lot. This is also true of a first-pairing defender like Charlie McAvoy, who’s used a lot in offensive situations as well as important defensive situations. Naturally, he’s going to be on the ice for a lot of goals for but also a lot against. It’s just the nature of his usage and style. It’s not ideal for a player to be in this section, but it’s usually fine (albeit likely terrifying), especially if they’re expected to be on the ice for more goals for than against.
Bad and Dull
“Bad” players are ones that only score between 1.0 and 2.315 GPG, while letting up somewhere between 2.293 and 4.5 GAPG. While this is the section to avoid at all costs on the teams’ graphs, that’s not necessarily true for a player graph. While obviously nobody wants to be on the ice for more goals against than for, it is completely unrealistic to expect that no one will be. The context in terms of usage is huge for player graphs. Fourth-liners are extremely likely to be in the “bad” section of the graph, but that doesn’t mean their bad players. It just means they’re used in a more defensive role and so they don’t score a lot, which leads them to be on the ice for more goals against than for.
So, it makes sense for fourth-liners like Sean Kuraly, Chris Wagner, and Joakim Nordstrom to be in the “bad” section. They don’t chip in much offensively, and they’re used in a lot of minute-eating defensive situations. So, they’re going to find themselves getting scored on more than they’re scoring, which is fine, because that’s what their usage lends itself to and that’s what’s expected of them. However, if you had someone like Bergeron, Marchand, or Pastrnak in the “bad” section, that’d be a problem because they’re expected to score more than they’re scored against, and it’s what their usage lends itself to.
Finally, we have “dull” players. These are the players that are on the ice for between 1.0 and 2.316 GPG, while they let up between 2.293 and 1.0 GPG. Players fall into this category if they’re typically on the ice when the game is boring to watch. Just like with the “fun” section, usage and playing style matter more than their placement though. A player who’s solid defensively but doesn’t add a ton of offense is likely to find himself here. It’s not necessarily a bad thing. As long as they’re performing as expected, it’s fine, just not exactly exciting.
A Final General Overview of These Graphs
The concepts of the explanations I provided above can be used for any good/bad/fun/dull graph. I simply used that as an example as I feel it’s easier to grasp if you see it spelled out for a specific situation. But, here’s a final general overview of each concept.
For team graphs, “good” means the team is expected to do signficantly more good than bad, while “fun” means the team is going to do a lot good, but also find a lot of bad happening. For a team to be “bad,” they have to be just that, and do a lot more bad than good, while “dull” just means that they aren’t doing much of anything good or bad.
As for the player graphs, if a player is in the “good” section, either they’re doing a lot more good than bad themselves, or their team is expected to do so on the ice. “Fun” means they’re doing a lot good but also a lot bad, or their team is doing that while they’re on the ice. For a player to land himself in the “bad” section, either he has to be doing a lot more good than bad himself, or his team has to be doing so while he’s on the ice. Finally, a “dull” player is one who is not doing much good or bad himself, or his team isn’t while he’s on the ice.
I hope this helps explain these graphs. For those who like analytics, they tell a lot about a team or a player in an easily digestible way, so I think they’re really useful. If you still have questions, feel free to reach out to me on Twitter, I’m always happy to try and answer them! Once the playoffs are over for the Bruins, keep an eye out for a series of articles using these graphs that compares the Bruins’ performance in the regular season to their performance in the postseason!