Despite some ups and downs, it’s tough to deny that the Buffalo Sabres are on an upward trajectory as an organization. It’s not hyperbolic to suggest that the 2022-23 squad is the closest thing to “playoff caliber” that we’ve seen since the last time Buffalo qualified for the postseason in 2010-11.

That’s not exactly a high bar to clear, so fans are fair to question just how close this team is to ending the league’s longest active playoff drought. Out of sheer masochism, I decided to peruse the on-ice metrics around the 2009-10 and 2010-11 teams to get a gauge for how comparable they were to the 2022-23 numbers.

What I found inspired what I feel is an interesting conversation on goalie metrics and the degree to which they can impact team standing points.

Enter The Time Machine

The year is 2010. The song “Single Ladies” by Beyonce just won song of the year at the Grammys, Barack Obama was in his first full year as President, and all of your friends owned the same quarter-zip sweater from The North Face (you know the one).

The Sabres were back in the playoffs after a two-year hiatus, finding themselves atop the Northeast Division with 100 points. It might come as a surprise to learn that this team was extremely comparable to the one we’re currently witnessing, with one major difference.

The 2009-10 Sabres

The last time Buffalo held the divisional crown, the team was, quite frankly, average. In all situations, Buffalo ranked 14^th (2.59) in expected goals-for and 16^th (2.69) in expected goals-against. Similar to the 2022-23 squad, they were a plus-shooting team, as evidenced by their raw GF/60 of 2.79, which was good for 9^th in the NHL that year.

When comparing teams from different years, league rank is more relevant that the actual xG numbers. In today’s NHL, an all-situation xGF/60 rank of 2.59 would be tied for the third-lowest mark in the league. Conversely, an all-situation xGA/60 of 2.69 would be the fourth-best mark in 2022-23.

With this in mind, I’ve included a chart below which shows the actual year-over-year team metric in numerical form. The number in parenthesis is where the metric ranked among other teams in the league that season (an important contextual factor, as previously mentioned).

As you can see based on the league-ranking chart, the 2009-10 and 2022-23 Sabres are pretty comparable offensively from an expected-goals standpoint. Both are plus-shooting teams (though the 2022-23 squad is converting goals to a greater degree than over-expected).

Defensively speaking, the xGA marks are a bit less similar but still comparable. Both teams ranked in the bottom half of the NHL. The real difference, however, exists in the raw GA/60 metrics. As you can see above, the 2009-10 Sabres ranked 4^th in the NHL in GA/60 while the 2022-23 team currently ranks 28^th.

Why the stark contrast? You guessed it, goaltending.

When a team ranks among the top-five goals-against squads in the league, you expect them to be strong defensively. That is unless they have a Vezina Trophy-caliber netminder doing most of the heavy lifting, which the Sabres did that year with Ryan Miller. His GSAx rate of 25.03 in 2009-10 ranked second in the NHL. As a team, Buffalo’s team mark of 24.79 ranked third.

As a group, the Sabres netminders essentially saved a goal above-expected every 3.3 games. That’s significant, but just how significant? The answer to that question can be partially revealed by examining the Sabres’ metrics the following season.

The 2010-11 Sabres

From an expected-goals standpoint, the 2010-11 Sabres were extremely comparable to the 2009-10 team, particularly on defense. Offensively, there was a slight year-over-year improvement that took place, but nothing earth-shattering. The team’s on-ice metrics were largely similar from one year to the next.

Despite a modest team xG improvement, the Sabres didn’t fare quite as well in the standings, managing 96 points compared to 100 the year prior. This change is reflected a degree in the raw GA/60 metrics which show that the Sabres allowed .31 additional goals-against per 60 (basically a goal every three games or so) compared to 2009-10.

If the xGA metrics are identical, how did this happen? That’s right, it’s goaltending again. As previously mentioned, the Sabres’ team GSAx rate in 2009-10 was spectacular. The following year was not nearly as impressive, however, as the team rate plummeted to -5.66 on the season. That’s a delta of 30.45 expected goals.

Over an 82-game season, that’s a massive shift. It’s the primary reason why a team with generally improved team metrics managed to produce fewer points in the standings. In a vacuum, the -5.66 GSAx number isn’t all that detrimental over a full season. This is evidenced by the fact that the Sabres’ xGA/60 and actual GA/60 numbers were very similar in 2010-11 (and both ranked 16^th in the league at the time).

It’s the overall year-over-year delta that matters here. Looking at these two seasons side-by-side, that is the statistic that sticks out in explaining how an improved offensive team and a near-identical defensive team managed to lose points in the standings. If the xGF numbers didn’t increase the way they did, it’s not unreasonable to suggest that they’d have missed the playoffs entirely.

What Does This Mean for the 2022-23 Sabres?

Goaltender performance is the most significant factor in explaining how the 2009-10 Sabres over-achieved in the standings, while the 2010-11 Sabres probably underachieved (albeit only slightly).

Based on where each of these respective teams ranked leaguewide in each of the metrics in the charts provided, the 2022-23 Sabres are extremely comparable to the 2009-10 squad in terms of team on-ice xG rankings. Unfortunately, they have had near polar-opposite success in net.

As of today, the Sabres are on pace to produce an overall GSAx mark of -18.6 on the year. Craig Anderson is the only netminder on the roster who has produced above-replacement level results, but not enough to offset Eric Comrie and Ukko-Pekka Luukonnen’s respective struggles. That serves in stark contrast to the 24.79 GSAx number the 2009-10 Sabres experienced between the pipes.

While there is some added nuance that could be applied here, it does paint a picture of just how much goaltending can elevate, or conversely sink an otherwise middle-of-the-pack team. The 2009-10 skater group was comparably productive to the 2022-23 skaters. The GSAx delta of over 43 goals (based on the 2022-23 projected pace) is the fly in the ointment.

Buffalo is currently on pace for a 79-point campaign. It would be irresponsible to declare that an elite goalie would make the current Sabres a 100-point team (as they were in 2009-10), but based on the historical data, it’s far from inconceivable.

This brings us to the topic of goaltender WAR, and how we can apply it to this analysis. To date, the Sabres goaltenders have cost the Sabres .81 wins above replacement. There’s an important distinction here. It must be understood that “above replacement” and “above average” are not the same thing.

This merely suggests that, as a whole, the Sabres have produced below-replacement goaltending this season. They’re not unique in this regard. Seven other teams in the league have experienced this.

Among the top 32 goalies in terms of minutes played in the NHL this season, the median WAR rate is between 1.1 and 1.4. So, if we arbitrarily apply the idea that the median starting goalie WAR is 1.25, we can speculate that if the Sabres had a “league average” starter, they’d have likely produced 2.06 additional wins to date 2022-23 (5.82 wins over a full season).

That would equate to roughly four standing points, which would still leave Buffalo on the outside of the playoff picture (as it currently stands), but not by much. Their standing-point percentage would be .551, which would rank them 10^th in the Eastern Conference (projected 90-91 point pace over a full season).

Closing Thoughts

I know that was a lot of data to digest. When it comes to goalies, in particular, the data is imperfect. That being said, it does give us something of a framework for understanding just how much the Sabres’ play in net has cost them this season.

One additional point to be made here is how public data tend to under-value odd-man rushes as it pertains to xG. The 2022-23 Sabres are a high-event team. I believe that goalies are at times disproportionately punished (statistically speaking) in that type of environment. Odd-man rushes in particular are an aspect of the game that requires some fine-tuning in most public xG models.

Based on both the underlying, (and what I remember from watching the 2009-2011 teams over a decade ago) the Sabres are way more high-event now than they were under Lindy Ruff. So, the idea that the 2022-23 Sabres would be a fringe playoff team with a league-average GSAx rate may be slightly ambitious.

There’s also the possibility that Comrie could come back from injury and provide positive results. The team’s goaltending trajectory isn’t necessarily set in stone. Still, the addition of a reliable starting netminder needs to be a top priority for this organization between now and opening night in 2023-24. No single roster spot has cost them more points in the standings (or has the propensity to continue to do so).

This isn’t as easy as it sounds. We just saw an example where even an elite goalie like Ryan Miller experienced a massive GSAx drop from one year to the next. Goalies are volatile, and even one with a historically strong track record could drop off a cliff in a given season. No other position is so prone to random (and dramatic) spurts of success and/or failure.

Data, scouting, and luck all need to factor into the equation. Just because it’s a difficult task, however, doesn’t mean the Sabres shouldn’t be exhausting all efforts to accomplish it.

Advanced Metrics and Charts courtesy of Evolving Hockey

Photo Credit: Mike Stobe/NHLI via Getty Images