On last week’s episode of Expected Buffalo | The Podcast we talked about how the Buffalo Sabres, in their current state, would be projected as a mid-80 point team in public models. One of the reasons we cited for it was the likelihood that they would see regression in their scoring.

After that episode, I thought to myself that fans might not fully understand what that means or how impactful it would be in the standings. Therefore, I decided to break out some data points that regression models would use in generating projections on scoring.

The Set- Up

The data that you’re going to see here is based on the last five years (2018-19 to 2022-23). I also used all situations scoring data. Focusing on just 5 on 5 wouldn’t be helpful with how important power play scoring is for some teams. It’s best to get the full scoring picture for this discussion. Also, when it comes to the Sabres, the power play is something we’ll get into.

I also focused on shooting percentage as the primary area for the year-over-year change. I could have looked at goals scored over expected, but shooting percentage was the cleanest way to look at it. It’s also a stat that everyone understands.

Scoring on a year-over-year basis is something that has a good amount of volatility in it. The good scoring teams will usually find themselves at the top of the league consistently but they’ll fluctuate between 4th one year and then 10th the following season. That doesn’t seem like a big change, but in some cases, it’s a difference of 15-20 goals. That could directly result in a standings change of an estimated three to four wins. Every situation would be different, but I’m comfortable with that assumption.

The Data

The first thing I did was look at the year-over-year variance of every team from one season to the next. I started with the change from the 2018-19 to the 2019-20 season. Then 2019-20 to 2020-21 and so on.

Below were the results on average variance percentages.

To make sure everyone understands the math here, a team with a 10% shooting percentage in 2021-22 that had an 11% variance regression would have had an 8.9% shooting percentage in 2022-23.

This starts to give us a glimpse into the variability in scoring. As I said, this isn’t the case for every team. I pulled together the chart below looking at two teams with some consistency in their shooting and two teams that have had inconsistency.

The Toronto Maple Leafs and Tampa Bay Lightning have been two of the top-scoring teams for the last five years and their shooting has seen little change year over year. At the same time, the Minnesota Wild and Sabres have seen a lot of variability year over year in their shooting.

With this information, I took the average variance percentage over the last five years. The result was 8.55%. Once I had this data point, my next step was to look at the top-scoring teams year over year. Perhaps those teams did not see as much variability from one season to the next.

The Sabres finished last season sixth in shooting percentage, therefore I took the top six scoring teams in each of the last five years and looked at their shooting percentage variance. Below is a look at the teams that were the top shooting teams from the 2020-21 season and what the variance was in their shooting the next season.

Here is the same thing for the 2021-22 top teams and the variance to the 2022-23 season.

I wanted to see how common it was for these top teams to see an improvement in their shooting percentage. The answer was that it wasn’t. In these two views, only two of the 12 teams saw a year-over-year increase in shooting percentage.

I considered that perhaps the league overall saw a reduction in shooting percentage that could have contributed to this, but nope. In Evolving Hockey’s data, the league shooting percentage has increased every year since the 2020-21 season.

The other finding looking at the top teams was the five-year average variance was similar to the percentage when I looked at all teams at 8.22%.

The Sabres

So let’s start to apply this data to the Sabres. This past season they had a 10.98 shooting percentage in all situations. If we apply an 8.22% regression that the top shooting teams had over the last five years, the Sabres would drop to a 10.01% shooting next season.

If they took the same amount of shots as they did this year (2,668), they would have scored 267 goals. That would be a 26-goal reduction compared to this season. Having said all of this, I don’t think that the Sabres’ shooting percentage would drop that much.

However, for the sake of argument, let’s say they go from sixth to dropping just outside the top 10 in shooting percentage next season. The Los Angeles Kings were the 11th-ranked shooting percentage team at 10.3% in all situations. That would be a 6.2% shooting regression for the Sabres. A reasonable number.

Once again, if they took the same amount of shots, they would score 274 goals. That’s a 19-goal reduction from last season. This was the genesis of the poll that I put on Twitter earlier this week.

Question for no particular reason at all.

Idk the exact answer but if I said the Sabres scored 19 less goals last season how many more games do you think they would have lost?

— Chad DeDominicis (@CMDeDominicis) July 9, 2023

Realistically, I believe a 19-goal reduction is worth a six to eight point reduction in the standings. In other words, three or four games lost.

The next obvious question you jump to is well what about goaltending improvement and offseason moves that would make up for that? We’ll get to that another day. The goaltending data article is next in the queue.

I want to remain focused on the shooting because I feel the spot where we’ll see the regression is obvious when you look at the Sabres’ numbers. That area is the power play. Without a change in the system, I’d be shocked if they didn’t see a change in their shooting with the man advantage.

Last year, the Sabres shot 16.06% on the power play. That ranked fourth in the NHL. They shot 11.76 goals above expected on that unit, which was the second-highest in the league in Evolving Hockey’s model. Only the Edmonton Oilers were higher.

In the last three years, that’s the fifth-highest a team has shot over expected on the power play. It’s an obvious place of regression unless something changes. They’re not going to shoot that well while also being ranked 27th in shot quality as they were last season with that system.

It’s also worth noting that overall Tage Thompson, Alex Tuch, Dylan Cozens, and Victor Olofsson all had career-high shooting percentages last season. I wouldn’t expect all four of them to repeat that. Their path of having someone ascend in scoring as those four came down a little is on the shelf until Christmas. Jack Quinn’s injury removed any true argument for them improving or even remaining flat on their shooting without another move.

Conclusion

I didn’t put this together to try to scare anyone into thinking the Sabres are not going to be good next season. I’m not trying to take the air out of the balloon. Instead, I’m trying to help people understand that this is an area that should be accounted for as a drop next season and why that’s the likely outcome.

The job of Kevyn Adams was to upgrade his roster enough to make up for that anticipated regression and then some to get this club into the playoffs next season. As of today, I’m not sure that he has done that and that’s why the models will predict the Sabres to be in that mid to high 80s in point projections.

The wild card in all of that will be Devon Levi. We’ll look at some goaltending data next.

Data via: Evolving Hockey

Photo Credit: Bill Wippert/NHLI via Getty Images