# Analytics goes ball hockey

Wednesday nights at the American Indoor Hockey Center on Mineral Springs Road in West Seneca, New York – a suburb of Buffalo – groups of dudes get together and play ball hockey. Some of the players are former high school or college players, some can barely handle the ball and some are good enough to take their game to the North American Championships in Philadelphia, Pa. The average age of the dudes is muddy. Ones who just left college can really move, but some of the late 30s and early 40s ballers still have it.

The Whig Party, a team named after an inside joke, is one of the league’s top clubs. In 2012-13, they went 20-5-3 against the league’s other 13 teams and won a their first championship. Things were not always that good for the Whigs. When they first entered the league in 2008, the group of newbs went 1-25-1.

Bill Cirocco is a pretty serious hockey player. He did not play in high school or at any sort of level, but got into rec leagues later and plays anything with a stick and net. Bill is, like many of us folks, a fan of the numbers.

Like his amateur career in hockey, he does not have any background in hockey stats. But…did I mention he takes his league seriously? Bill began keeping stats for his league based on goals, assists, plus-minus and penalty minutes and has attempted to create “advanced stats.”

He explains:

Say we play one of the other top teams in the league. They average 4.96 goals per game, and 3.28 goals against per game. We score 6 goals on them, and allow 3.

This yields an “Offense Score” of 2.72 (6 GF minus 3.28 Average Goals Against for the Opposition). and a “Defense Score” of 1.96 (Opponents 4.96 goals per game average minus the 3 we allowed). I then combine these 2 numbers for a “Team Score” of 4.68, which is a pretty good number compared to the league average.

I keep track of these numbers for all 14 teams in the league, for every game throughout every season. I use excel to make everything a lot easier and then compile all the data in charts so that it’s easy to look at how good teams actually are.

I can now look at sums of every teams offense, defense, and team scores to get an idea of how good our competition is. I included the one workbook I use to keep track on all this information and auto-calculate everything just for reference.

Assuming you’re still with my at this point, my question is just about if you’re aware of a different way to present this data. So here’s the 14 teams, and our offense and defense scores, and yes, I know some of the team names are random/silly.

 NAME OFF DEF TEAM The Whig Party 23.36 47.36 70.72 Devils 11.20 51.84 63.04 Brew Jays 40.00 20.04 60.04 Mighty Ducks 26.45 21.08 47.53 LWO 22.20 12.84 35.04 Haters -1.64 24.40 22.76 Multiple Scoregasms -6.24 20.68 14.44 Ogie Ogilthorpe 26.36 -13.36 13.00 Ebenezer Ale House 1.40 4.40 5.80 Crowley Can’t Score 6.12 -5.80 0.32 Big Sausage Pizza -19.04 -39.80 -58.84 Jabber Jaws -29.16 -37.84 -67.00 Trashers -49.11 -43.71 -92.82 Das Boot -43.24 -69.72 -112.96

Bill sent us a 173 page document. Really. When I clicked to download the document, my computer nearly shut down and started on fire.

Within the pages of the 173 pages – that I did not read in its entirety – are numbers on points his team has scored against teams he qualifies as easy, moderate and difficult based on the goal differential figures.

He ranks the teams’ offenses and defense based on his Quality of Competition goal numbers and determines whether an opponent’s strength is their offense or defense. He qualifies wins as “offensive” or “defensive” wins and divides by the total number to find whether a team gets more of their wins by playing exceptional on D or with their scoring.

For individuals, he normalizes the raw numbers with his easy, moderate, difficult system and by determining if goals were scored in close games or blowouts. Clearly, the goals and assists against difficult teams in close games are weighted most heavily.

 # Opponent Whig GF Whig GA Opp. Avg. GA Opp. Avg. GF Opp. Goal Diff. Offense OR Defense DR Team TR Better Unit 1 Haters 2 2 3.2 4.1 22 -1.2 24 2.1 12 0.9 20 DEF 2 LWO 3 3 3.7 5.0 34 -0.7 20 2.0 13 1.4 18 DEF 3 Sharktopi 12 2 7.1 2.2 -123 4.9 2 0.2 22 5.1 5 OFF 4 Brew Jays 4 3 3.2 5.8 66 0.8 11 2.8 6 3.6 11 DEF 5 Big Sausage Pizza 9 0 5.8 3.3 -63 3.2 6 3.3 2 6.5 3 DEF 6 Ogie Ogilthorpe 7 3 4.6 5.2 17 2.4 8 2.2 11 4.7 6 OFF 7 Ebenezer Ale House 3 2 4.0 4.3 7 -1.0 22 2.3 9 1.3 19 DEF 8 Trashers 9 0 5.8 2.0 -94 3.2 5 2.0 13 5.2 4 OFF 9 Multiple Scoregasms 3 2 3.4 4.0 16 -0.4 19 2.0 17 1.6 15 DEF 10 Mighty Ducks 5 5 3.3 5.0 42 1.7 10 0.0 26 1.7 14 OFF 11 F.S.B. 2 2 4.5 4.3 -5 -2.5 27 2.3 9 -0.2 26 DEF 12 Jabber Jaws 10 0 5.8 2.9 -71 4.2 4 2.9 5 7.2 2 OFF 13 Devils 0 3 2.0 4.9 72 -2.0 26 1.9 20 -0.1 25 DEF 14 Haters 4 1 3.2 4.1 22 0.8 12 3.1 3 3.9 9 DEF 15 LWO 4 3 3.7 5.0 34 0.3 14 2.0 13 2.4 13 DEF 16 Sharktopi 9 4 7.1 2.2 -123 1.9 9 -1.8 28 0.1 21 OFF 17 Brew Jays 2 3 3.2 5.8 66 -1.2 23 2.8 6 1.6 16 DEF 18 Big Sausage Pizza 5 1 5.8 3.3 -63 -0.8 21 2.3 8 1.5 17 DEF 19 Ogie Ogilthorpe 3 4 4.6 5.2 17 -1.6 25 1.2 21 -0.3 27 DEF 20 Ebenezer Ale House 4 0 4.0 4.3 7 0.0 15 4.3 1 4.3 8 DEF 21 Trashers 12 0 5.8 2.0 -94 6.2 1 2.0 13 8.2 1 OFF 22 Multiple Scoregasms 4 1 3.4 4.0 16 0.6 13 3.0 4 3.6 10 DEF 23 Mighty Ducks 6 3 3.3 5.0 42 2.7 7 2.0 18 4.7 7 OFF 24 F.S.B. 9 6 4.5 4.3 -5 4.5 3 -1.7 27 2.8 12 OFF 25 Jabber Jaws 3 1 5.8 2.9 -71 -2.8 28 1.9 19 -0.8 28 DEF 26 Devils 0.0 15 0.0 23 0.0 22 TIE 27 Haters 0.0 15 0.0 23 0.0 22 TIE 28 LWO 0.0 15 0.0 23 0.0 22 TIE

For some of the games, there are stats you could draw Corsi numbers from – for some games those stats do not exist. There are also goalie saves and shot against stats.

 Game #10 9:30 Mighty Ducks NAME Y/N POS G A P +/- SOG MS BS FOW FOL Scott Escobar Y C 1 0 1 0 3 0 0 16 8 Mike Parks Y D 0 0 0 0 0 1 1 0 0 Andrew Maas Y RW 3 1 4 0 5 0 0 0 0 Steve Staebell Y D 0 0 0 0 1 0 1 0 0 Bill Cirocco Y LW 0 0 0 0 3 1 0 0 0 Nick White Y LW 0 0 0 1 2 4 1 0 0 Ryan Snyder N Jamie Annunziata Y D 0 1 1 0 2 0 1 0 0 Mark Williams N Joe Kurczewski Y RW 0 0 0 -2 2 3 0 0 0 Matt Stewart Y C 1 1 2 1 5 2 1 10 9 Gordon Kus Y D 0 0 0 0 1 2 3 0 0 TOTALS 10 5 3 8 0 24 13 8 26 17 Name Shots Against Goals Allowed Save % Jeff Munkelt 20 5 0.750

There is no message or lesson to this story. We just thought it was crazy and cool. Hopefully you did too. If you have ideas to help Bill with more stats, leave them in the comments.