Goalie Debuts
Hockey On Stats

It's game day. You're leaving work, ready to sit down with a drink and watch your favorite team play. To make sure you're prepared, you decide to check out the lineups.

"Shit."

You say this just as you realize that the opposing team is starting a rookie goalie and it's his debut. Naturally, your team is going to make him look like an all-star

Note that I counted only when a goalie started the game. Not when they first appeared in an NHL game, for example after the normal starter was pulled
I would've gone farther back, but the NHL doesn't provide game logs before that.
This chart includes zoom and hover. Flip the toggle below to turn them on! All features work on both desktop and mobile. Hover over a dot to see who played when, how and against who. In future articles, an icon like the one above will indicate that there's interactivity related with the chart.
Generally you'd be looking for a p-value below .05 to deem it statistically significant.

I am a fan of the New York Rangers and often think the exact same thing. Our minds, though, are extremely susceptible to biases and they tend to remember the worser parts of a season rather than the best ones. So, do goalies making their debut actually play like all-stars or, do we selectively remember the times a few played like all-stars and forgot about all the times some played average, mediocre even? This article will aim to address that exact question.

For this piece I looked at every goalie who had a debut smsin the NHL since 1988. sms( btw, these are tooltips so click or hover over these icons for more info) This amounted to about 350 debuts over 27 years for an average of about 12 goalies per year.

Save % vs Goals Against
Average

First off, I'll look at what the stats were in general for these goalies. Meaning, just plotting out Save % (SV%) vs Average Goals Against (GAA). In the graph below you'll see the afformentioned statistics. Interestingly, most goalies seem to fall within the expected ranges and perform somewhat the same at a given GAA or SV%.

Another interesting thing to look at is how a now established goalie (Or a once established goalie) performed in his debut. Today, we know these established goalies are legends and are/were some of the best to ever play but did their debut show that? For example, Roberto Luongo had a 1 GAA, 97 SV% night in his debut for Florida against Boston. But while some performed as you'd expect, others didn't. The best example is Chris Osgood. In his debut with Detroit he posted a 7.5 GAA and 67 SV%. Granted, some discrepancies might arise in the visualization of the data due to the averages in SV% and GAA during different time periods, but that will be addressed later.

touch_app bubble_chart
Switch for interactivity! Flip it on to play around with the graph below.
info_outline

All-Star or Average

Next comes the question of whether or not most goalies tend to play out of their minds good duing their debut. This can be visualized by plotting their SV% and GAA versus the league average of a given year. What we see is that it's kind of all over the place. Goalies that debut tend to perform average as a whole which means we shouldn't be expecting goalies to perform incredibly in a debut but rather, at about the league average.

Upon further inspection of the numbers there's actually about a 50/50 shot, considering all things equal, whether the goalie you're playing will play above the league average or below in reference to both SV% GAA. Since 1988, 159 goalies (50.96%) played above the league average SV% and 153 goalies (49.04%) played below the league average. Also since 1988, 154 goalies (49.36%) had a higher GAA than league average while 158 goalies (50.64%) had a below average GAA. So as you can see, when there's a goalie making his debut against your favorite team, you really have no idea what you're in for. This is especially true because there are numerous goalies who went on to have hall of fame careers and play below the league average in their debut.

bubble_chart

What fanbase can complain?

So we've discovered that the chances of a goalie outperforming the average against your team as a whole is much like flipping but are there teams that can actually complain? Well, below are charts of how many times a franchise (active franchises) has lost or won versus a debuting goalie.

Fanbase that can complain the least

And the award for the fanbase that can complain the least goes to the... Columbus Blue Jackets! They've never lost a game when facing a debuting goalie for a win percentage of 100%. In close second were the Flames who have won 5 more games than they've lost which gives them a win percentage of 66%.

Fanbase that can complain the most

And the one you're probably all looking for, the fanbase that can complain the most about facing a debuting goalie... The New York Rangers! Since 1988 they've lost 88% of their games against debuting goalies for a grand total of 7. More than anybody else in the league. In a not so close second are the Philadelphia Flyers who have lost 4 more games than they've won and have a losing percentage of 66%

equalizer

Can you predict future performance?

If a goalie plays out of his mind, does that mean that he'll be a hall of famer? Should he immediately be the starter and have his development sped up? To answer this particular question I compared a goalie's debut performance to his future career numbers. For both SV% and GAA of course. As you might expect, a goalies performance in his debut is not a very good predictor for how good of a goalie he'll be over the span of his career. The p-value for SV% is an unimpressive .38 and the R-squared is .03.sms GAA tells about the same story but it is relatively a better predictor than SV%. The p-value of GAA is .73, the R squared value is .083. The overall takeaway is that in the grand scheme of things, how a goalie does in his debut, stats-wise, is pretty irrelevant when trying to determine how he'll perform in the future.

bubble_chart

Conclusion

Goalies aren't as incredible as we think when they debut. We just tend to rationalize an unkown situation and account for the worst possible scenario, in this case losing. So next time you're facing a new goalie, take a seat, relax and enjoy the game.

linear_scale

Written by: @Daveedgamboa

Graphs built with d3.js and excel. Data obtained via hockeyreference.