Tonight's Whiteboard Picks: Philadelphia, Toronto, and Colorado — and the Goalie Keeping One of Them Close

By Ashkon Baharlou

Three games tonight, and the model's read is straightforward: Philadelphia over Pittsburgh, Toronto over the Islanders, Colorado over Los Angeles. Two of the three are clear. The third is close entirely because of who's in net for the Islanders. Here is the pick for each game and the numbers behind it.

One note before the picks. Every number below comes from the Whiteboard model, run over tonight's expected lineups. Each club gets a team score from 0 to 100 that blends its skaters (80 percent) with its goaltending (20 percent), and the win probability comes from the Whiteboard's standings formula with home ice included. Skater percentiles are against last season. Chemistry is the Whiteboard's 0-to-100 rating of how a line or pair plays together, and the unit ratings below are on the same 0-to-100 scale. I log every pick, and tonight's results story will grade them against the finals.

7:30 p.m. ET: Pittsburgh at Philadelphia (the goalie gap)

The model's pick: Philadelphia, 80 percent to win.

The Whiteboard's team scores read 63.7 for Philadelphia to 49.1 for Pittsburgh, and the goaltending gap drives most of it. Dan Vladar's goalie grade is 61.0 on the Whiteboard's 0-to-100 goalie scale, built on 49 games last season in which he saved 0.39 goals per 60 minutes above expected. Arturs Silovs's goalie grade is 12.3, the lowest of tonight's six starters, from 39 games in which he allowed 0.33 goals per 60 above expected and posted an .851 save percentage on rush chances. The model's rush-fit adjustment moves Silovs down another 2.3 grade points and Vladar up 0.5, leaving team goaltending at 61.5 for Philadelphia and 10.0 for Pittsburgh.

The skaters are closer. Philadelphia's skater rating is 64.3 to Pittsburgh's 58.9, with the defense rating 66.1 to 55.0. Travis Sanheim is at the 100th percentile, the top Elite driver on either roster, and his pair with Rasmus Ristolainen carries a 70 chemistry rating, the best unit in the game. Owen Tippett is Philadelphia's other Elite driver at the 94th percentile, with Sean Couturier (88th), Cam York (83rd), and Ristolainen (82nd) among six Strong drivers. Pittsburgh's defense has one strong pair, Samuel Girard and Kaedan Korczak at 66 chemistry, with Girard an Elite driver at the 92nd percentile. The other two pairs sit at 46 and 38 chemistry.

Pittsburgh has the edge in two places. Its second line of Egor Chinakhov, Tommy Novak, and Evgeni Malkin carries 85 chemistry, the highest of any line in this game, and its top line with Sidney Crosby and Rickard Rakell is at 75. Its power-play rating is 68.1 to Philadelphia's 60.5, and its first unit carries 65 chemistry to Philadelphia's 47.

The honest risk: most of this pick rests on one goalie grade. On skaters alone, Philadelphia's win probability drops to 65.7 percent. If Silovs played like an average goalie, a 50 grade, Philadelphia's chance would be 67.8 percent.

7:30 p.m. ET: NY Islanders at Toronto (the closest call)

The model's pick: Toronto, 61 percent to win.

The Whiteboard's team scores read 61.9 for Toronto to 59.0 for the Islanders, the closest game of the night. The two sides pull in opposite directions. Toronto's skater rating is 58.6 to the Islanders' 50.3, ahead in the forward ratings (56.8 to 47.6), the defense ratings (60.5 to 53.0), and the power-play ratings (67.3 to 53.9). The Islanders have the best goalie of the night. Ilya Sorokin's goalie grade is 91.3, from 55 games last season at 0.64 goals saved per 60 minutes above expected and a .982 save percentage on rush chances. Anthony Stolarz's goalie grade is 72.6 over 26 games. With rush fit included, Islanders team goaltending is at 93.5 to Toronto's 75.0.

Toronto's strength is its top half. The line of Jack Roslovic, Auston Matthews, and Kirill Marchenko carries 79 chemistry, Toronto's defense pairs run 65, 68, and 59 chemistry, and the first power-play unit is at 67. Toronto has no Elite drivers but six Strong ones: Matthews (86th percentile), Chris Tanev (85th), Marchenko (83rd), Darren Raddysh (80th), Roslovic (74th), and Jake McCabe (67th). The bottom six is thin: the third and fourth lines carry 36 and 34 chemistry.

The Islanders' best unit is on defense. Adam Pelech, an Elite driver at the 92nd percentile, and Matthew Schaefer, a Strong driver at the 76th, form a pair at 62 chemistry. Bo Horvat is the Islanders' other Strong driver at the 83rd percentile, centering a top line at 54 chemistry. The second pair of Alexander Romanov and Ryan Pulock is the weakest unit in the game at 33 chemistry.

There is one night of fresh evidence on this Toronto lineup, and it cuts both ways. Toronto dressed these same 18 skaters in the same lines last night against Montreal, controlled the game, and lost 3-2: 3.17 expected goals to Montreal's 2.22, 2.27 to 1.54 at 5-on-5, six high-danger chances to two. The top line did what its 79 chemistry predicts: Roslovic, Matthews, and Marchenko were on the ice for 1.09 expected goals for and about 0.4 against at 5-on-5, on-ice shares of 74, 71, and 74 percent, with Marchenko leading the line at 0.58 individual expected goals on four shots. The fourth line did what its 34 predicts and got caved: Brandon Duhaime at a 13 percent on-ice share, Teddy Blueger at 6, and Zack MacEwen on for no expected goals for in 5.4 minutes. The third line was the exception to its 36 rating: Easton Cowan was on for 0.30 expected goals for and 0.06 against, an 84 percent on-ice share, with Cole Sissons at 74 and Nick Paul at 59. On defense, Morgan Rielly and Raddysh (70 and 67 percent on-ice shares) and McCabe and Tanev (71 and 58) held up, while Oliver Ekman-Larsson and Troy Stecher (46 and 41) were outchanced. The second line was mixed: William Nylander scored twice on 0.49 individual expected goals while his line was outchanced at a 42 percent on-ice share, and John Tavares generated 0.89 individual expected goals, 0.67 of it on a power play that produced 0.72 expected goals and no goal. Sergei Bobrovsky allowed three goals on 2.22 expected goals; Stolarz starts tonight instead.

That is one game, and none of it is in the ratings yet: the model's numbers for tonight are built on last season.

The honest risk: Sorokin. On skaters alone, Toronto's win probability would be 70.9 percent. His 91.3 grade is what pulls it to 61.

10:00 p.m. ET: Los Angeles at Colorado (the mismatch)

The model's pick: Colorado, 78 percent to win.

The Whiteboard's team scores read 74.2 for Colorado to 61.6 for Los Angeles, and Colorado's 74.2 is the highest team score of the night. Nine Elite drivers dress for Colorado, led by Nathan MacKinnon at the 99th percentile, Devon Toews and Artturi Lehkonen at the 97th, and Cale Makar at the 96th. The MacKinnon line with Lehkonen and Martin Necas carries 88 chemistry, the best line of the night across all three games. On defense, Toews and Makar carry 78 chemistry and Josh Manson and Brent Burns 69. Colorado's defense rating is 80.2 to the Kings' 56.6, the widest single gap of the night, and its penalty-kill rating 76.9 to 58.1.

The Kings are better up front, the one component they win: their forward rating is 68.4 to Colorado's 66.3. The top line of Artemi Panarin, Quinton Byfield, and Adrian Kempe carries 85 chemistry, and the first power-play unit is at 73, the best of any power-play unit tonight. Byfield, Trevor Moore, and Alex Laferriere are Elite drivers at the 92nd, 92nd, and 91st percentiles. The Kings' top defense pair, Erik Gustafsson and Brandt Clarke, carries 38 chemistry.

Goaltending adds to Colorado's edge. Scott Wedgewood's goalie grade is 77.3, from 45 games at 0.58 goals saved per 60 minutes above expected. Darcy Kuemper's is 50.4 over 50 games. With rush fit, Colorado's team goaltending is at 80.9 to 52.0.

The honest risk: Colorado's third line. Jaden Schwartz, Fedor Svechkov, and T.J. Hughes carry 29 chemistry, the lowest of any line tonight, with Svechkov out of position at center and Hughes unrated at the Whiteboard's 40th-percentile default. If the Kings' forward edge shows up anywhere, it is against that line. On skaters alone, Colorado's win probability is 71.4 percent.

The verdict

Three picks. The model takes Philadelphia at an 80 percent win probability, Toronto at 61, and Colorado at 78.

The real risk, named plainly: two of these three picks lean on goaltending, and goalie grades are last season's data. Silovs's 12.3 and Sorokin's 91.3 goalie grades are the numbers until this season says otherwise. These probabilities are the model's read of tonight's expected lineups, and expected lineups change. Treat the numbers as the read they are.

Tonight's results story will grade all three against the finals.

Check the work. Open the Whiteboard, the free line tool I built, and build tonight's lineups yourself. Move the pieces, watch the ratings move, and bring your version to the results story tonight.

All team scores, goalie grades, skater percentiles, tags, and chemistry ratings from the Last Change Hockey Whiteboard model, run on tonight's expected lineups. Toronto's September 29 game numbers from the Whiteboard's shift-corrected tracking of that game. Win probabilities are the model's read.

Follow @lastchangehockey on Instagram for more Whiteboard breakdowns.