According to the Numbers: Week 5 – Fantasy Hockey Pickups, Drops, Sell-Highs & Buy-Lows
Box scores tell you what happened. Underlying numbers tell you what’s going to happen. Each week in “According to the Numbers”, I dig into SportLogiq’s data to find the players whose results don’t match their process: the waiver-wire adds who are better than their ownership suggests, the underperformers who are due for a breakout, and the hot starts that are about to cool off.
This week, a 90%-rostered defenseman has the biggest gap between his points and his expected points in the NHL, and the Panic Meter takes aim at both a rookie phenom and Sidney Crosby.
New this week: Expected Points, which multiplies a player’s on-ice expected goals by his individual points percentage (IPP) over the last three seasons. It shows how many points he “should” have based on the chances created while he’s on the ice, and how often he’s typically involved in them.
All stats via SportLogiq and are current through games played on October 29th, 2026. Rostered percentages are from Yahoo.
Pickups
Jordan Eberle, SEA – RW (26% Rostered)
The Stat: Eberle ranks 6th among forwards in inner-slot shots per game
Eberle isn’t going to win any style points at 36 years old, but he’s quietly one of the most productive wingers on the waiver wire. He has 10 points (4G / 6A) in 13 games, and six of them have come on the power play, a big deal for an add at 26% rostered.
The underlying numbers say he’s earning it. Eberle ranks 25th among forwards in ixG per game, 12th in ixG/60, 11th in Grade A chances per game, and 6th in inner-slot shots per game. That’s the profile of a player who lives in front of the net, and he’s actually 1.58 goals below expected, with four goals on 5.58 ixG. Seattle is controlling 52% of the expected goals at 5v5 when he’s on the ice, so he’s not being carried, either.
He’s averaging under 17 minutes a night, so the floor isn’t huge, but the combination of PPP, quality chances and some goals still to come makes him a strong add in standard leagues.
Jake Neighbours, STL – LW/RW (13% Rostered)
The Stat: Neighbours ranks 14th among forwards in inner-slot shots per game
Neighbours was moved to the Blues’ top line on Thursday and is the banger-league pickup of the week. He has 44 hits in 13 games, 17th among all skaters in hits per game, and he’s producing too: seven points (4G / 3A), with four on the power play.
He doesn’t take many shots (1.8 SOG per game), but nearly all of them are dangerous. 81% of his shot attempts come from the slot, and he ranks 14th among forwards in inner-slot shots per game. He’s also been a sneaky PP weapon, ranking 37th among forwards in PP ixG per game in just over two minutes of PP time a night.
His goals are right in line with his expected goals (4 on 3.70 ixG), so this isn’t a heater. At 15 minutes a night, his ceiling in points-only leagues is limited. But if your league counts hits and PPP, he’s a multi-category add with real scoring touch.
Noah Östlund, BUF – C (4% Rostered)
The Stat: Östlund is averaging 18:55 TOI and 3:23 of PP time per game
Östlund is playing nearly 19 minutes a night for Buffalo, and his 3:23 of power-play time ranks 88th among forwards. He’s turned that usage into nine points (4G / 5A) in 13 games, including three on the power play, and he’s still available in 96% of Yahoo leagues.
The chances back it up. He ranks 47th among forwards in ixG per game, 59th in Grade A chances, and 31st in PP ixG per game, so he’s a real threat on the man advantage, not just along for the ride. He’s averaging 2.3 SOG and 4.4 shot attempts per game, and the Sabres control 54% of the expected goals at 5v5 with him on the ice.
He won’t help you in hits (two all season), but in standard formats he’s one of the best under-5% options on the wire. That kind of deployment rarely goes unclaimed for long.
Buy-Lows
Zach Werenski, CBJ – D (100% Rostered)
The Stat: Werenski has 0 goals on 2.56 ixG, the worst goals vs. expected among defensemen
Werenski has nine assists in 11 games, but zero goals, and nobody on the blue line has been unluckier. He ranks 6th among defensemen in ixG per game and 2nd in inner-slot shots per game, while playing nearly 29 minutes a night.
It’s been a sluggish start for the entire Blue Jackets team, and even an elite defenseman like Werenski isn’t immune to it. But Werenski ranked 3rd among defensemen in goals over the last two seasons, so it’s shocking to see him without a goal in his first 11 games. The bad news is that his shot volume is down, from 3.6 SOG and 7.9 shot attempts per game during those seasons to 2.7 SOG and 6.7 shot attempts per game. The good news is that he still ranks 3rd among defensemen in shot attempts per game and 8th in SOG per game. That’s enough volume to compete with the top five at the position, but the start has been disappointing enough that you might be able to get him cheap.
His IPP (56%) is right on his three-year norm (59%), so the assists are real. Add the goals and he’s back to elite.
Rasmus Dahlin, BUF – D (99% Rostered)
The Stat: Dahlin ranks 2nd among defensemen in SOG per game (3.2)
Dahlin missed a few games due to injury, and with seven points in 10 games, he’s been fine but not elite. The underlying numbers say elite is coming. He ranks 2nd among defensemen in shots per game and offence-generating plays per game, and 4th in ixG per game and ixG/60.
The issue is the finishing around him. Buffalo has scored 13 goals with Dahlin on the ice against 16.8 expected, the 8th-worst on-ice goals vs. expected among defensemen. His IPP (54%) matches his three-year norm (56%), so once the Sabres start converting, his points will follow. His expected points are 9.4.
He won’t come cheap, but this is likely the lowest his value will be all season, so it can’t hurt to send a couple of offers.
Adrian Kempe, LAK – RW (95% Rostered)
The Stat: Kempe has 3 goals on 5.19 ixG, 16th-worst goals vs. expected among forwards
Kempe has just eight points in 12 games, but he’s been one of the most active shooters in the league. He has 41 SOG (12th among forwards in SOG per game) and 79 shot attempts, and he ranks 23rd among forwards in ixG per game and 27th in Grade A chances. He’s shooting 7.3%, well below normal 13.0%.
His assists are lagging too. Kempe has been in on just 47% of the Kings’ goals with him on the ice, way below his 72% IPP over the last three seasons. Put it together and his expected points are 12.8, almost five more than he has.
He’s playing nearly 21 minutes a night with 3:33 of PP time, and LA is controlling 53% of the expected goals at 5v5 when he’s on the ice. The goals and assists are both due. This is a classic buy-low on a player who’s done everything right except get rewarded on the scoresheet.
Shea Theodore, VGK – D (90% Rostered)
The Stat: Theodore has 5 points against 12.7 expected points, the biggest gap in the NHL
No player in the league has a bigger gap between his points and his expected points than Theodore. Vegas has scored 24 goals with him on the ice, right in line with the 23.3 expected, so the team is finishing. Theodore just isn’t getting on the scoresheet: he’s been involved in 21% of those goals, against a 54% IPP over the last three seasons.
Besides the lack of a consistent PP1 role, everything else looks great. The Golden Knights are controlling 62% of the expected goals at 5v5 with him on the ice (14th among defensemen) and he’s a plus-12 (if your league still has that as a category). He ranks 16th among D in shots per game and 25th in ixG per game.
That IPP won’t stay this low. A player who’s typically involved in more than half of his team’s goals isn’t going to keep missing four out of every five. If his manager is frustrated by five points in 14 games, send an offer.
Luke Hughes, NJD – D (70% Rostered)
The Stat: Hughes has a 30% IPP this season, against 51% over the last three years
Hughes has six points in 13 games, and a two-point night on Tuesday suggests things are starting to turn. The Devils have scored 20 goals with him on the ice, almost exactly the 21.0 expected, but he’s only been involved in 30% of them.
His expected points come out to 10.7, so he’s about five points behind where he should be. He’s playing 24 minutes a night with 3:35 of PP time (15th among D), ranks 10th among defensemen in SOG per game and 11th in ixG per game, and all four of his goals have come at 5v5. New Jersey also controls 55% of the expected goals at 5v5 when he’s on the ice.
Even with Tuesday’s game, his manager may still see “six points in 13 games” and feel underwhelmed. This might be your last chance to buy Hughes at a discount.
Sell-Highs
Leo Carlsson, ANA – C (97% Rostered)
The Stat: Carlsson has a 90% IPP this season, against 67% over the last three years
Carlsson has been excellent in fantasy, with 18 points in 13 games (8th among forwards in points per game) and 12 power-play points. But he’s been involved in 90% of Anaheim’s goals with him on the ice, and that won’t last. His expected points are 11.7, six fewer than he actually has.
He’s also not shooting much. Carlsson has just 20 SOG in 13 games (193rd among forwards in SOG per game) and ranks 181st in ixG per game. On the power play, he’s been all playmaker, ranking 245th in PP ixG per game despite 12 PPP. At 5v5, Anaheim controls 50% of the expected goals with him on the ice.
In redraft leagues, he’s a sell-high because his value may never be higher than it is right now. He carries name value, he was a top-50 fantasy pick this September, and he has 85 points (33G / 52A) in his last 83 games. The problem is that the Ducks look far from the team they were last season, and a lot of this season’s production is unsustainable. If someone in your league is paying for a point-per-game center, cash in.
Elias Pettersson, VAN – C (78% Rostered)
The Stat: Vancouver has scored 9.1 goals above expected with Pettersson on the ice, 5th-highest among forwards
Pettersson was borderline undrafted in fantasy this September and now has 16 points in 15 games. For a former 100-plus point player, this might look like a return to form. But the underlying profile doesn’t support it. He ranks 240th among forwards in ixG per game, 286th in Grade A chances, and 295th in inner-slot shots. At 5v5, he has zero goals on just 0.35 ixG.
The production is coming from the players around him. Vancouver has scored 24 goals with him on the ice against 14.9 expected, and at 5v5 it’s 13 goals on 7.8 expected. The Canucks control just 46% of the expected goals at 5v5 when he’s out there, 262nd among forwards. His expected points are 9.4. Pettersson is being buoyed by an astronomically high 18.3% on-ice shooting percentage, which is simply not sustainable.
He’s still useful on the power play (six PPP), and the 33 hits help in banger leagues. But he’s a strong candidate to sell while the box score looks like 2023 Pettersson.
Jack Roslovic, TOR – C/RW (55% Rostered)
The Stat: Roslovic has an 83% IPP this season, against 65% over the last three years
Roslovic has been one of the waiver-wire success stories of the season, with 15 points in 15 games, and his rostered percentage has doubled from 27% to 55% in the last two weeks. But his production is running well ahead of the underlying numbers. He’s been in on 83% of Toronto’s goals with him on the ice, way above his three-year norm, and the Leafs have scored 3.2 goals more than expected with him out there. His expected points are 9.7, more than five fewer than he actually has.
For most of the season, he was getting just 1:25 of PP time per game (221st among forwards), so his production wasn’t tied to a secure role. His shooting process has been solid, though: he ranks 66th among forwards in ixG per game and 57th in SOG per game, and five of his six goals have come at 5v5.
Now the role has changed. With Kirill Marchenko out for the next few weeks, Roslovic has moved up to PP1. That should keep the points coming in the short term, making this the perfect time to sell. Other managers will see 15 points in 15 games, top-line minutes and a power-play role, but that role disappears when Marchenko returns, and his 83% IPP is due to come down regardless. If you can get real value for him now, take it. If not, ride the PP1 bump. Trading Roslovic 1-for-1 and getting an upgrade will likely be difficult, but he’s the perfect sweetener for a deal right now.
Panic Meter
Sidney Crosby, PIT – C (94% Rostered)
The Stat: Pittsburgh controls just 47% of the 5v5 expected goals with Crosby on the ice (244th among forwards)
Panic Meter: 🟠 Worry
Crosby has seven points in 13 games and is a minus-8. There’s some bad luck here: he’s been in on just 47% of Pittsburgh’s goals with him on the ice, against a 71% IPP over the last three seasons, and his expected points (11.5) say he should have four or five more.
But the bigger concern is that Pittsburgh isn’t driving play with him on the ice. The Penguins control just 47% of the expected goals at 5v5 when he’s out there, and he ranks 109th among forwards in ixG per game and 129th in SOG per game. Those are fine numbers for a middle-six forward, not for a player drafted around the top 50.
The points should tick up as his IPP normalizes, so this isn’t a drop. But he’s no longer a guaranteed point-per-game player, and he may not be the buy-low case he may seem like.
Gavin McKenna, TOR – LW (79% Rostered)
The Stat: Toronto controls just 37.5% of the 5v5 expected goals with McKenna on the ice (349th of 396 forwards)
Panic Meter: 🟠 Worry
On the surface, McKenna has been fine: 11 points in 15 games, five of them on the power play, plus a three-point night (2G / 1A) in his last game. Underneath, there are real concerns. He’s averaging just 13:45 of ice time per night, and at 5v5, the Leafs are getting badly outplayed with him on the ice. Toronto controls just 37.5% of the expected goals in his minutes, among the worst marks in the league.
His role should be growing. With Kirill Marchenko out for the next few weeks, McKenna has moved up to the top line and was already seeing time on PP1, but he still only played 15:33 on Wednesday. Even with the promotion, Toronto isn’t giving him top-line minutes, and that caps his ceiling.
He’s also benefiting from some luck. He’s shooting 19.2%, and Toronto has scored 3.3 more goals than expected with him on the ice. He ranks 155th among forwards in ixG per game and 252nd in inner-slot shots, and he has zero hits on the season.
The PP1 role and the pedigree keep him rosterable, and he’s in dynasty leagues for the long haul. But in redraft, the underlying numbers say his current pace is more likely to drop than rise, especially once Marchenko returns. If someone in your league is buying the three-point night and the top-line promotion, it’s a good time to listen to offers.
Matthew Knies, CBJ – LW/RW (65% Rostered)
The Stat: Knies ranks 203rd among forwards in ixG per game
Panic Meter: 🟠 Worry
Knies had two goals and one assist in his Blue Jackets debut, and four points in his first two games. Since then, he’s on a nine-game goalless drought with just one assist, 14 SOG and 16 hits. The underlying numbers suggest this is his real level. His expected points are 5.7, almost exactly what he has.
He’s getting a solid role, 18 minutes a night and 3:11 of PP time, but he isn’t turning it into chances. He ranks 203rd among forwards in ixG per game and 161st in SOG per game, and Columbus controls just 48% of the expected goals at 5v5 with him on the ice. The 17 hits help a little in banger leagues.
There’s no bad luck to point to here, so there’s no reason to expect a bounce-back. He’s a hold in deeper leagues for the role, but in 10 and 12-team standard formats, he’s on the bubble.
Porter Martone, PHI – RW (64% Rostered)
The Stat: Martone ranks 287th among forwards in ixG per game
Panic Meter: 🔴 Drop (Redraft)
Martone was one of the most hyped rookies of the summer (ADP 63), but six points in 15 games, with 14:02 of ice time per night, isn’t what fantasy managers paid for.
The underlying numbers don’t offer much hope. Martone ranks 287th among forwards in ixG per game, 286th in Grade A chances, and 340th in offence-generating plays per game. He’s averaging 1.3 SOG per game, and the Flyers control just 41% of the expected goals at 5v5 with him on the ice (321st among forwards).
In dynasty leagues, he’s an obvious hold. In redraft, at 64% rostered, he’s taking up a spot that could go to a player like Viktor Arvidsson. Until Philadelphia gives him more minutes, he’s droppable.