NHL Over/Under Betting Strategy: Goalies, Pace, Empty Nets, Overtime Maths and Building Your Own Total

Updated 20 Oct 2026 · By the BetHubCA Editorial Team

Hockey totals look simple. The board says 6.5, you think the Oilers and Leafs will light it up, you take the over. Then two hot goalies stand on their heads, it's 2–1 with five minutes left, and you're praying for an empty-netter. Been there. The good news is that NHL totals are one of the markets where a bit of structure really pays off, because so much of the answer is known before puck drop: who's in net, who played last night, how each team plays at five-on-five. So grab a coffee and let's walk through a real over/under strategy, step by step, the way I'd explain it to a buddy who's tired of guessing.

Before we start: the legal betting age is 19 in most provinces and 18 in Manitoba and Quebec, and this site isn't for residents of Ontario or Alberta. Cloudbet, the book we link to, is licensed in Curaçao and is not authorized by any Canadian province. Every line, price and stake below is an illustrative example for learning the method. None of it is a live price or a pick.

How an NHL total is posted

An over/under on a hockey game is a wager on the combined goals both clubs score. You don't care who wins. A 4–3 game and a 7–0 blowout both count as seven goals. The interesting part in hockey is that the number itself barely moves. Most games sit at 5.5, 6 or 6.5. What moves is the price attached to each side.

Here's how a slate might look (illustrative only):

Because the number is sticky, books shade the price instead. On the Winnipeg game, the over is "juiced" to 1.83, which tells you the book thinks six or more goals is a bit more likely than not. If the market leaned even harder, you'd eventually see the total jump to 6 or 6.5 with fresh prices. When you read hockey totals, always read the number and the price together. Over 5.5 at 1.83 and over 5.5 at 2.05 are very different bets.

The Ottawa game uses a whole number. If it lands on exactly six goals, the bet is a push and your stake comes back. We'll see shortly why overtime makes that push less common than you'd think.

Break-even percentages: the number that keeps you honest

Divide 1 by the decimal price and you get the win rate you need just to break even:

So the strategy question is never "will this go over?" It's "is the chance of the over higher than the break-even on this price?" If you rate the Winnipeg over at 53%, that's a pass at 1.83, even though you think it's more likely than not. Our odds calculator converts any decimal or American price for you.

Step one: build a rough projection

You don't need a fancy model. You need a consistent one that you write down before you peek at the odds, so the market can't anchor you. Here's a simple four-part routine.

1. Start with each team's scoring and allowing rate

Use a sensible window, like the last 20 to 25 games blended with the full season, so one hot week doesn't fool you. Average what each team tends to score with what its opponent tends to allow. Say the home side scores about 3.3 a game and the visitors allow about 3.1. That's (3.3 + 3.1) ÷ 2 = 3.2 expected for the home team. Do the same the other way: the visitors score 2.9 and the home side allows 2.7, so (2.9 + 2.7) ÷ 2 = 2.8. Your starting projection is 3.2 + 2.8 = 6.0 goals.

2. Adjust for the goalies actually starting

This is the single biggest adjustment in hockey. Season scoring numbers blend every goalie a team has used. If tonight's starter is the backup, and the backup has been leaky, your number should go up, often by a few tenths of a goal. If a struggling starter has been replaced by a hot one, it can go down. Starting goalies are usually confirmed by the morning skate or a beat reporter a few hours before puck drop. Wait for that confirmation when you can.

3. Adjust for schedule and travel

The second night of a back-to-back, a long road trip, or a late flight across time zones tends to mean tired legs and, quite often, the backup in goal. That can push goals up on the tired team's end. A team coming off a few days of rest gets the opposite nudge. Keep these adjustments small, in the 0.1 to 0.3 range, unless the goalie also changes.

4. Adjust for special teams and style

Two clubs with strong power plays that also take a lot of penalties create more five-on-four time, and five-on-four is where goals come fast. Two tight-checking teams that sit back after scoring first produce long stretches of nothing. Look at shot volume and shot quality at five-on-five, not just goals, since goals alone bounce around a lot over short samples.

After all four steps, maybe your 6.0 becomes 6.3 because the visiting backup is starting on a back-to-back. That's your number. Now, and only now, open the odds.

Step two: turn your number into a probability

Goals in hockey are rare events spread across 60 minutes, which is why a Poisson distribution gives a decent first estimate of how often a game lands over or under a line. It's not perfect (more on that below), but it's a big upgrade on gut feel. Here's what it gives for regulation-time goals at different expected totals:

Notice two things. First, a small change in your projection moves the probability a lot: going from 5.8 to 6.2 expected goals moves over 5.5 from about 52% to about 59%. That's why the goalie adjustment matters so much. Second, a whole-number total of 6 has a big push chance, roughly one game in six, which changes the maths on those lines compared with 5.5 or 6.5.

A quick worked example. Your number is 6.2, and you see over 5.5 at 1.91. The table says about 58.6% before overtime is considered. On a C$20 stake: 0.586 × C$18.20 − 0.414 × C$20 = C$10.67 − C$8.28 = about +C$2.39 expected per bet. If your number were 5.8 instead, the over hits 52.2% and the same bet is roughly break-even at best. The difference between those two outcomes is often just "who's in net".

Overtime and the shootout: the quirk most bettors miss

At most books, the full-game total includes overtime, and if the game goes to a shootout, the winning team is credited with one extra goal for settlement. Check your book's house rules, but if yours works this way, it creates a neat bit of maths.

A game tied after 60 minutes always has an even number of goals: 1–1, 2–2, 3–3. Whatever happens next, exactly one goal gets added. So:

• A 2–2 game after regulation always finishes with 5 goals. Under 5.5 is safe, and over 5.5 is dead.
• A 3–3 game after regulation always finishes with 7. Over 6.5 cashes.
• A whole-number total of 6 can't push from a 3–3 regulation tie, because it becomes 7.

That's why the push rate on a total of 6 in real full-game settlement is lower than the regulation-only Poisson table suggests: some of those six-goal outcomes are 3–3 ties that turn into sevens. It also means that in a low-scoring 1–1 grinder, under 5.5 is guaranteed even if it goes to a shootout. A fair share of NHL games, often around one in five, need extra time, so this isn't trivia. It's part of the price. For more on how books treat extra time across different hockey markets, see do NHL bets include overtime.

Empty-net goals: why late scoring is lumpy

When a team trails by one or two late in the third, the coach pulls the goalie for an extra skater. Two things can happen, and both add a goal: the trailing team ties it, or the leading team scores into the empty net. Coaches have become more aggressive about pulling the goalie earlier, so empty-net goals are a real part of how hockey totals land.

What this means for you:

• Close games are over-friendly late. A 3–2 game with four minutes left has a decent chance of becoming six goals one way or the other.
• Blowouts are under-friendly late. A 5–1 game usually goes quiet: nobody pulls a goalie, and the winning team often coasts.
• The Poisson model underrates this clustering. Real hockey scoring is a bit more lumpy than a smooth model assumes, so treat the table as a starting point, not gospel.

Reading the price, not just the number

Because the total itself rarely moves, price movement is your best clue to what the market thinks. Suppose a game opens at 6 with even prices around 1.91 on both sides. By evening the over is 1.80 and the under is 2.05. The market has decided it's more of a 6.2 or 6.3 game than a flat 6.0. Usually that's news: a backup goalie announced, a top defenceman scratched, or sharp money hitting early.

Your job is to compare your own number with where the price ended up:

• If your projection is 6.4 and the over has only moved to 1.85, you may still have value.
• If your projection is 6.1 and the over is now 1.77, the market has gone past your number. Pass or look at the under.
• If the price moves sharply against you and you can't find the reason, that's a sign you may be missing information. Be humble.

It also pays to check a couple of books before you bet. On a 5.5 total, getting 1.95 instead of 1.87 moves the break-even from 53.5% to 51.3%, which over a season is the difference between a small profit and a small loss.

Team totals and period totals

Team totals

Sometimes you have a strong view on one team but not the other. Maybe a club is starting a goalie who's been getting shelled. A team total lets you bet that club's opponent to score over 3.5 without needing the other side to cooperate. Using the same Poisson idea for one team:

If you project a team for 3.4 and its over 3.5 is listed at 2.40 (break-even 41.7%), your 44.2% estimate suggests a small edge. Team totals also carry a bit more margin at many books, so you need the gap to be clear.

First-period totals

The first-period total is usually 1.5. First periods tend to be a little tighter than the game as a whole, as teams feel each other out. If you project 1.7 goals in the first, over 1.5 is about a coin flip (50.7%). At 1.9 expected, it's around 56.6%. Period totals settle on that period only, so overtime and empty nets late in the third don't help or hurt you. That makes them a cleaner bet for people who hate the late-game chaos, but the margin is often higher, so shop around.

Situations worth a closer look

None of these are automatic bets. They're spots where the market sometimes reacts too slowly or too much, so they deserve a second look:

• Late goalie news. When a starter is scratched in warm-ups, prices can lag for a few minutes. Only act if you trust the source.
• Backup on a back-to-back against a strong power play. This combo often gets an over adjustment, but check how far the price has already moved.
• Two elite goalies in a rivalry game. Casual money loves overs in big Canadian matchups. If the line drifts up because of public interest rather than news, the under may be the better price.
• Arena and schedule quirks. Early afternoon starts and long road trips can flatten pace. Keep adjustments small.
• Injury to a top-pair defenceman. Losing a big-minutes defender can raise the opponent's chances more than losing a forward does.
• End-of-season motivation. A team out of the playoff race may play looser, while a team fighting for a spot often tightens up. Be careful, though: this is the kind of story that everyone has already heard.

Live totals: where patience gets rewarded

In-play hockey totals update after every goal and drift lower as the clock runs down. A few practical habits:

• Know what the clock does to the line. If a game is 0–0 after the first period, a pre-game 6.5 might now be 5.5 or 4.5 live. Your pre-game projection, scaled for the time left, tells you whether the new number is fair.
• Shots and chances matter more than the score. A 1–0 game where both teams are firing 15 shots a period is very different from a 1–0 game with eight shots total.
• Penalties change everything for two minutes. Books adjust live prices during power plays. Don't chase a price that already reflects the advantage.
• Watch for stream delay. Your broadcast may be several seconds behind the book's data. Books also suspend markets during dangerous moments, so don't count on beating the line by reacting to what you see.
• Remember the empty net. Late in a one-goal game, live overs carry a hidden boost from the pulled goalie. Late in a three-goal game, they don't.

Cash-out offers on live totals usually include extra margin. Use them when your situation genuinely changes, not to calm your nerves.

Staking: how much to bet on a total

Even a good totals bettor loses a lot. Hockey is a low-scoring, high-variance sport, and one bounce off a skate can flip a result. Flat staking keeps you sane: pick one unit size, usually 1% to 2% of your betting bankroll, and use it for every total.

Losing streaks of six to eight bets happen to everyone, even with a real edge. If your unit is small, a bad week is annoying instead of devastating. Avoid stacking totals into parlays: at 1.91 per leg, a four-leg totals parlay pays about 13.3 times your stake, but needs everything to go right, and the book's margin compounds on every leg.

Common mistakes with NHL totals

• Betting before the goalie is confirmed. You're guessing at the most important input.
• Ignoring the price. Over 5.5 at 1.77 needs 56.5% to break even. That's a very different bet from 2.00.
• Trusting a small sample. Three straight overs say very little. Look at shots and chances over 20+ games.
• Forgetting overtime settlement. Know whether your book counts the shootout as one goal before you bet on totals near 5.5 and 6.5.
• Chasing with live overs. If you're down on the pre-game under, piling into a live over to "get it back" usually just doubles the margin you pay.
• Betting every game. On a full slate, you might find one or two prices worth taking. That's normal.

A simple game-day routine

1. In the morning, build your projection for the games you care about, using scoring, allowing and five-on-five shot data.
2. Write down your numbers before you check the odds.
3. Wait for confirmed starting goalies, then adjust.
4. Compare your number with the total and price at two or three books.
5. Bet only where your estimated probability beats the break-even by a clear margin, say two or three percentage points.
6. Record the bet, your number, the price you took and the closing price.
7. After 50 or more bets, check whether you're beating the closing price more often than not. That's the best early sign your method works, long before your profit and loss tells you anything.

If you also bet goalscorers, our NHL anytime goalscorer strategy uses the same expected-goals thinking, and the puck line guide covers the spread side of hockey. If you follow three-down football too, the CFL over/under tips show how weather and scoring rules change totals in a very different sport. For a wider look at patterns and line movement, see NHL betting trends.

A word on tax and the law

For most recreational bettors in Canada, gambling winnings are generally not treated as taxable income, but there are exceptions for people who bet in a business-like way. Our guide to tax on betting winnings explains the basics. On the legal side, each province decides how online betting is run, and the rules differ. Read is online betting legal in Canada before you sign up anywhere.

Keep it fun

Totals make a random Tuesday game between two non-playoff teams worth watching. That's the fun part. The not-fun part is chasing losses at 11 p.m. on the West Coast. Set a weekly budget before the first puck drops and stop when you reach it. If betting stops feeling like a hobby, free confidential help is there: Gambling Support BC at 1-888-795-6111, and Jeu : aide et référence in Quebec at 1-800-461-0140. You'll find more options on our responsible gambling page.

My take

A good NHL over/under strategy comes down to a few habits. Build your own number first. Make the starting goalie your biggest adjustment. Turn your number into a probability and compare it with the break-even on the actual price. Understand that overtime adds exactly one goal to a tied game, and that empty nets make close games lumpy late. Shop for the best price, bet flat, keep records, and pass on most of the board. It won't win every night. Nothing does in hockey. But it gives you a real reason behind every bet you make.