# Here Come the Hawks, Part I

Anthony Dalke is a died-in-the-wool Blackhawks fan from Chicago. He is also an MSBA student at the Mendoza College of Business. Follow him on twitter @AnthonyDalke

Life has been pretty good for Blackhawks fans. Right now, the team stands tied with Nashville for first place in the Central Division (with the Preds enjoying a game in hand) and boasts the league’s best goal differential, an impressive 28% better than – you guessed it – next-best Nashville. What more could Chicago faithful ask for?

Well, when measured against the vaunted eye test, the Hawks have displayed a trend: Instead of looking like the team carrying the torch of the “Detroit Model”, consistently driving and keeping play in their offensive end and utilizing their goalies only as a true last line of defense, their netminders have seemed to work harder this year, facing more shots and, consequently, shouldering a greater portion of the overall defensive load.

This brings us to two questions:

1. Do the underlying numbers paint the same picture?

2. If so, does it matter?

Let’s start with the first question. Taking my cues from HP contributor and uber-stat expert Stephen Burtch, I pulled 5-on-5 score-adjusted, as opposed to close, data for Corsi For Percentage (CF%), Corsi Against Per 60 Min. (CA60), and Total Corsi Events Per 60 Min. (CP60), all adjusted for rink count bias and home advantages, as well, from War-on-Ice.

From a CF% perspective, things look good: the Hawks sit behind only the Lightning and Kings this season. Furthermore, their year-over-year trends seem acceptable:

However, when we narrow it down to CA60, things get a little more concerning:

As you can see, this chart contains error bars referring to the standard deviation, 2.73, of this data sample. It tells us a few important things: 1. The team’s number of 50 this season stands just over one standard deviation above the sample’s average of 47.2; 2. The yearly figures have quite a bit of noise; and 3. In their two recent Stanley Cup seasons, they have neared or exceeded one standard deviation below the average.

We can utilize some basic statistics calculations to glean additional detail. For instance, a normal distribution calculation reveals the Hawks’ 50 CA60 this season falls around the 85th percentile of our sample data. This means if we charted out the probabilities of the numbers in our sample, we would expect 85% of the data points to fall below Chicago’s CA60 this season. Although that seems rather high, it does technically fall below the 95% threshold statisticians typically require to chalk up an observation to statistical significance, as opposed to random variation. The CA60 would have to edge up to 51.7 to earn that “distinction”.

So, what conclusions can we draw? Although the Blackhawks have managed to continue controlling possession at strong overall rates, they have done so by allowing more shots than normal and simply taking even more. And even though their shot allowance rates haven’t climbed into crisis territory, they do considerably exceed both the team’s recent averages and the levels they registered during their Cup seasons in 2009-10 and 2012-13.

You might be asking yourself, “Okay, but their top-line numbers have stayed so strong this season, so why should we care?” Well, for one thing, the team’s On-Ice Save Percentage has risen to 92.6% this season, up from 92.1% last year and 92.0% the previous, when they most recently got to kiss Lord Stanley’s chalice. Corey Crawford, Antti Raanta, and Scott Darling have respectively posted Even-Strength Save Percentages of .921, .941, and .944, combining for .933 as a group, placing Chicago third in the league.

For comparison’s sake, here are the team’s ESV% rankings the prior four seasons – the years since Crawford snatched the starting role: 24th, 8th, 14th, and 7th . Because this shows so little repeatability (we all know what they say about goalies having fickle natures!), the team cannot reasonably expect to rely on continued strong goaltending to remain atop the standings.

Furthermore, far more detailed research from the incomparable Jen Lute Costello (@RegressedPDO on Twitter) suggests shot prevention, not accumulation, correlates strongest with playoff appearances and success. The Corsi Against data discussed above lend further credence to this notion. We can speculate as to the exact causes of this phenomenon – for instance, perhaps the teams that allow lots of shots but still post strong overall possession numbers have to rely on those maddeningly unpredictable goalies and, thus, more often fall victim to the whims of PDO.

Whatever the reason, the bottom line remains: recent history shows the teams that make the playoffs tend to excel in shot prevention. Therefore, although the Blackhawks and their fans don’t yet have cause for alarm, this season the team has disconcertingly strayed from the shot-suppressing ways that have helped it achieve so much success the past five years.

For the next part of this series, I’ll take a stab at parsing out possible causes of these trends, focusing on shot quality allowed by the team and the performances of selected players.

All Corsi data compiled from War-on-Ice.com, with individual Save Percentages from NHL.com and Even-Strength Save Percentages from hockeyanalysis.com*

## 3 thoughts on “Here Come the Hawks, Part I”

1. Anthony, thanks for a novel look at Blackhawks Big Data, from a fellow Hawks supporter. Two questions:

1. Doesn’t the relatively great number of shots that the Hawks take probably contribute to the higher number of shots against them? The logic being, taking shots puts the puck out of Hawks players’ control and, often, puts it in the opposing team’s control, which increases the changes of the opposing team getting a shot on goal?

2. If so, do you have any statistical evidence to suggest that if the Hawks (or any team) take fewer *low percentage* shots while continuing to take acceptable-percentage shots, they would probably decrease their shots-against % and, as a result, find that sweet spot between taking too-few shots to win and taking too many low-percentage shots that often can lead to odd-man rushes or other negative stats in the defensive zone?