Below is a guest post written by Eric, who is an regular reader of the site. He enjoys looking at the statistical side of NHL betting and he turned his focus to the over/under bets on total goals for games in the Stanley Cup Finals. Feel free to let Eric know how you feel about the article in the comment section below.
I’m writing this article with a simple goal in mind: to answer the question “Is it safe to bet on the total score of Game 1 (and the rest of the series) of the 2015 Stanley Cup?” In other words, “Do I think it is likely enough to put money on over or under 5 goals being scored Wednesday night?”
My (entirely stat based) opinion: No.
Let’s break down why.
First off it makes sense to take a quick look at each team’s season ranking in the relevant simple stats: goals per game and goals against. During the regular season, Chicago scored 2.7 goals per game (17th in the league), and allowed 2.3 goals per game (2nd in the league). Comparatively, the Lightning scored 3.2 gpg (1st!) and allowed 2.5 (12th). So we’ve got the #1 offence against the #2 defense squaring off in Tampa Bay. If we quickly average each team’s regular season stats with each other we can predict 2.75 goals for the Lightning and 2.6 for the Blackhawks. That’s 5.35 goals total, just .35 above the 5 we have to break away from to come out ahead.
However, the playoffs are a completely different setting in any sport, and we’ve got plenty more games to look at. So far the Blackhawks have averaged final scores of 3.35 to 2.88, and the Lightning have averaged 2.95 to 2.25. At the bottom of the Blackhawks spreadsheet I average in these numbers to predict a 5.5 goal game, but at this point we’re doing so much averaging that it’s hard to look at that number seriously. Obviously if we average every game the two have played together we’re gonna end up with the same prediction we have to bet against: 5 goals.
So instead of averaging more numbers together let’s look at some different stats. On each sheet I separated the team’s goal totals for home and away games thus far in the playoffs in the “Home Sum” and “Away Sum” columns. Tampa Bay will be hosting the Blackhawks so maybe we can gain some insight from these columns. In the three series we’ve seen from each team so far the Blackhawks have average a losing game of 3 to 3.33. This stat is skewed a bit by two blowouts in Nashville but at any rate they seem to play more exciting games on the road. Meanwhile, the Lightning have averaged a somewhat less high scoring game of 3.11 to 2.44 in Tampa Bay. This score is still, however, higher than their total playoff average. So I believe it’s safe to say we can expect more action packed games in the Sunshine State. Averaging these numbers gives us a prediction of 3.22 to 2.72 in favor of the Eastern Conference Champions. More importantly, that’s a total of 5.94 goals.
Now that we have all the stats and averages we can handle, how do we interpret and value them? Are these reliable stats? I believe that over the series as a whole they will turn out to be reliable, but as always, every game is different and anything can happen. A big part of setting the average goals for the series will be whether or not we see games go into overtime. Both teams are averaging around 20% more goals in overtime games and this can make or break our bet. It’s also hard to know how much confidence to put in our stats. As the article progressed I started to talk about some stats that aren’t based on a whole lot of games. That last paragraph detailing the teams respective stats at home and on the road are based on a mere 9 games each. Furthermore, all the stats used from playoffs games had standard deviations of around 2 goals each, and 2 goals is a lot when we’re talking about a 5 goal game. So for me anyway, the numbers alone aren’t enough to consider a 6+ goal game (the more probable of the two in my opinion) likely enough to bet on.
My bets will be limited to the line and straight up wins for this series; at least with those wagers it makes more sense to bet with your gut instead of with Excel.