Evolving Hockey 2024 Playoff Preview: Conference Finals

We’re to the point now where the remaining teams are all good. At least that’s what the expectation is in the Conference Finals, right? It’s May Madness baby, the Final Four. And this year, I suspect, we’re in for an absolute treat. The Rangers killed analytics for good, the Panthers made Boston look fun again while properly dismantling them, the Stars are my begrudging favorite, and the Oilers have the best player in the world. If the real world was a Monte Carlo simulation, only the Rangers wouldn’t be here, and even that feels somewhat predestined given the unfortunate recent playoff history of the Canes and how all the numbers were sure they’d win the Cup. I suppose hockey remains Hockey, after all. I can’t wait to watch some good Hockey. Let’s go Stars (I can’t believe I’m being forced to say this because they rely so heavily on skater defense and that’s a principle thing).

Here’s how last round played out:

Conference Home Away Round 1 % Outcome
East NYR CAR 31.4% – 68.6% 4-2
East FLA BOS 70.7% – 29.3% 4-2
West VAN EDM 29.1% – 70.9% 3-4
West DAL COL 68.0% – 32.0% 1-4

We’re close.


Eastern Conference

New York Rangers (M1)     vs.     Florida Panthers (A1)

37.9%                         62.1%

The Rangers did the impossible: they beat the (public stats) models. Aka the Hurricanes. I don’t know what it is, and we’ve already talked about this at length here, but historically, the Hurricanes were *the* team to go far by modern hockey analysis, and the Rangers were set up to lose big. The models want it to be one way, but it’s the other way. At least if you assume a 7-game series is enough to determine the better team. Either way, the Rangers did it. Their series with Carolina was extremely close, at least in vibes – they nearly went to a final game. As usual, the Hurricanes dominated the underlying metrics at even-strength, out-attempting the Rangers by 85 shots, averaging over 2.5 goal expected more, and scoring 2 more actual goals. But the Rangers have a great powerplay and an incredible goalie in Igor Shesterkin that matched the perimeter-happy Canes team quite well. They enter the Eastern Conference as fairly large underdogs by our model, but I guess at this point we just have to understand that this Rangers team does not play by the rules. They also have the best players in the ’23-24 NHL playoffs (Kaapo Kakko and Alexis Lafrenière).

After a surprisingly dull and, quite frankly, bad showing from both Toronto and Boston in the first round, the Bruins showed up with a completely different game plan. It didn’t really matter after game 1, since the Panthers were clearly ready for it, but it made for a much more interesting series. After a messy showing in the early games (there were a lot of goals, that’s how we define that), Florida tightened things up quite a bit, with the help of an ageless of Sergei Bobrovsky and 2024 Selke winner Aleksander Barkov. The recent Rangers teams have come to be known for their reliance on a good powerplay, but the Panthers absolutely peppered Boston with the extra man. With more than 100 shot attempts on the powerplay (compared to Boston’s 31), people may just start asking what happens if they don’t get a powerplay. The answer there is they’ll dominate at even-strength or score shorthanded. With the Hurricanes out of the playoffs, the Panthers move into prime position to become the analytics darlings of the Summer. Shawn can’t wait to watch Gustav Forsling and tell us how much we would be talking about him if we were Panthers fans.


Western Conference

Dallas Stars (C1)     vs.     Edmonton Oilers (P2)

53.1%                         46.9%

The Stars and Oilers were our favorites here in the second round – Dallas at 68% and Edmonton at 71% – so, we’re pretty happy this is the Western Conference final we get because it means we were right (we’re never wrong). It took another long series for the Stars, who along with Boston and Vancouver, have played the most games in the playoffs at 13. For me, Game 6 may have been the best game of this summer’s playoffs. At even-strength, you could possibly argue Colorado was the better team, out attempting Dallas in 4 of the 6 games. The Stars’ special teams on both ends made up the difference, however, with two short-handed goals, even generating about half an expected goal more than the Avs on the penalty kill in game four. The Stars weren’t as stingy defensively this round, but that may be expected against the firepower Colorado brought. Their task may be even taller with this Oilers team. Of note, Jake Oettinger has played in every game so far and leads all goalies in time on ice by 130 minutes. His 3 goals saved above expected has gotten the job done, so far, but it will be interesting to see how or if Dallas fits in any rest days for the Lakeville native. Of note, Hintz is questionable for game 1, but we’ve gone ahead and slotted him back in the roster for these projections.

What we thought may be a messy series from two ardent, maybe even rabid, fanbases turned out to be a very cordial affair. Other than the standard bickering of two Canadian rivals… It took seven games, and the Oilers only out-scored the Canucks by a goal through those games, but it felt a bit more lopsided than your standard seven game match. Edmonton out-attempted Vancouver by 75 shot attempts at even strength and generated about 3 expected goals more on their powerplay than they allowed from Vancouver’s. As we noted in our writeup heading into this series, the Canucks were hit with a very unfortunate injury to Thatcher Demko, and rookie Arturs Silovs played quite well, but you can only ask so much of a 23-year old goalie with nine games played in the NHL. Our model had the Oilers as heavy favorites heading into round 2, but their Conference Final opponent will prove a true challenge. While Vancouver wasn’t necessarily a bad team defensively, they didn’t bring the same skater ability that the Stars do on the other side of the ice. The Oilers enter the third round with at least two Conn Smythe favorites and the Stars with arguably less than one. Interesting, isn’t it?

Leave a Reply