NHL weekly betting guide: True moneylines for every game

Early this season, we tried to determine how much Aaron Ekblad affected the Panthers' win probability in any given game. The market being slow to downgrade the Panthers without Ekblad, who averages over 24 minutes per game, led to a profitable stretch of betting against Florida.

Last Monday, another Stanley Cup contender got bad injury news when the Avalanche's Nathan MacKinnon suffered an upper body injury that will keep him out for four weeks. So how do we determine how much a star like MacKinnon matters to moneyline prices and Colorado's implied win probability?

The Avs' next game came against the red-hot Bruins in Colorado. Our rating on each team has been really solid, only occasionally finding value backing or fading each team. As a result, with MacKinnon, the line was going to look something like Bruins -115 / Avalanche -105.

With MacKinnon announced out, the line opened Bruins -160/Avalanche +140, and eventually closed Bruins -175/Avalanche +150. This gives us our first look at what the betting market thinks of the Avalanche's chances to win a game without their star forward:

That's an 11% change in win probability with MacKinnon out. The Avalanche lost that game 4-0 to Boston, getting dominated at even strength.

Next up, the Rangers visited Denver on Friday and we projected a fair moneyline of +109/-109, so a line of Rangers +100/Avalanche -120 would be expected. Instead, New York was lined at -150 and Colorado +130. Let's look at the difference in implied win probability:

Look at that. Another 11% adjustment without MacKinnon, and a loss to the Rangers.

Basically, it took two games for us to find out what effect a missing MacKinnon has on Colorado's chances to win on a game-to-game basis. If the Avalanche continue to get dominated, though, we'll have to adjust them down, and if they rally without their star forward, we can lower "The MacKinnon Effect" from 11%.

The recipe

We started the campaign using regular-season point totals as a baseline for rating teams since it's still our best measurement. Throughout the campaign, we'll adjust club ratings using on-ice metrics to remove the cognitive bias of win-loss records (which can be skewed by outliers like special-teams results, poor goaltending performances, and other unreliable events).

The cheat sheet

There are no bad bets at the right price, but how do we know what a good price is?

The following includes my fair price on the games (true moneyline) and the moneyline price I'd need to bet either side. I only need a 1% edge for a favorite if we're getting better than a fair price on the team more likely to win. For the underdog, I'll need 4% or better to make it a bet. For games I project to be closer to a coin flip, a 2.5% edge is enough for a worthwhile wager. I also have a 5% win probability consideration for a team playing in the second game of a back-to-back with travel and a 3% consideration for a team on the second leg of a back-to-back without travel. When it comes to injured players, I do my best to estimate the impact on their team's win probability.

When the betting markets open up the night before, you can compare those prices with our "price to bet" column to see if you're getting any value with either side's moneyline. There's also a possibility that a moneyline moves into a bet-friendly range at some point between market open and puck drop.

Matt Russell is a betting writer for theScore. If there's a bad beat to be had, Matt will find it. Find him on Twitter @mrussauthentic.