The Avalanche have scored 41 goals in 10 games and allowed 30. Only Edmonton has a better goal differential in the Western Conference, and Colorado's comes with a 5-4-1 record. Colorado flew home from New Jersey on Tuesday with zero points from a two-game trip, after a 3-2 loss in Philadelphia and a 4-3 loss to the Devils that the Denver Post headlined with a line from the dressing room: "We looked slow, unenergetic."
A plus-11 team at .500 usually means one of two things. Either the team has been unlucky in close games, or the goal differential is flattering it. Colorado is the rare case where both are true at once, and the two pieces of luck point in opposite directions. Strip them out and what is left is still a very good team, with one problem that was its biggest strength a year ago.
Where did the plus-11 come from?
Four games. Colorado beat Los Angeles 8-4, St. Louis 6-1, Calgary 7-3 and Pittsburgh 5-2. Those four wins add up to 26 goals for and 10 against, a plus-16. In the other six games the Avalanche have been outscored 20-15.
| Game | Score after two | Final |
|---|---|---|
| Oct. 7 at Winnipeg | 1-2 | L 2-3 |
| Oct. 10 vs Toronto | 3-4 | L 3-6 |
| Oct. 14 at Dallas | 3-1 | W 3-2 |
| Oct. 15 vs Dallas | 1-1 | SOL 2-2 |
| Oct. 19 at Philadelphia | 1-2 | L 2-3 |
| Oct. 20 at New Jersey | 1-2 | L 3-4 |
Those six games are the season so far. Colorado is 1-3-1 when the margin is one goal, 0-4 when trailing after two periods, and has one comeback win all year. It has also blown three leads. The late pushes have been real, but they have arrived too late. Both of Artturi Lehkonen's goals in New Jersey came in the last three and a half minutes, after the Devils had gone up 4-1.
So on the face of it, this is a good team losing coin flips. The trouble is that the same team has been winning a different coin flip by a very wide margin.
How hot has the finishing been?
Hotter than anyone's. At five-on-five, Colorado has scored 34 goals from chances worth 23.1, according to MoneyPuck. That 10.9-goal gap is the biggest in the league, ahead of Seattle at 9.4 and Edmonton at 8.7. The Avalanche are converting 14.2 per cent of their five-on-five shots, second in the NHL, and scoring 4.12 goals per 60 minutes, first by a distance.
| Colorado at five-on-five | 2026-27 | League rank |
|---|---|---|
| Expected goals share | 54.3% | 5th |
| Expected goals for per 60 | 2.80 | 9th |
| Expected goals against per 60 | 2.35 | 10th |
| Shooting percentage | 14.2% | 2nd |
| Goals for per 60 | 4.12 | 1st |
| Save percentage | .910 | 12th |
The chance numbers in the top half of that table are good. They are not the 57 per cent Colorado posted last season, the best in hockey, but fifth in the league is a contender's share. The bottom half is where the scoring has run ahead of the play.
Most of the overshoot belongs to a handful of players. Lehkonen has eight goals on 16 shots. Those shots were worth 4.2 expected goals. Martin Necas has five from 2.7, and Gabriel Landeskog four from 2.5. Nobody keeps shooting 50 per cent, and Lehkonen will not either.
The one big exception runs the other way. Nazem Kadri has 27 shots and no goals, from chances worth 3.8. He has seven assists, and Colorado has 64 per cent of the expected goals with him on the ice in all situations, so he is playing well. He is simply the only regular whose luck has gone wrong, and his goals will come back as Lehkonen's cool off.
Is MacKinnon's line getting caved in?
Not over the season. The Denver Gazette noted that in New Jersey, the top line and top pair were out-attempted 12-3 in their minutes together, and it was a bad night. Across 10 games the picture is much healthier.
| Colorado regular | Five-on-five TOI per game | On-ice expected goals share | On-ice goals for-against |
|---|---|---|---|
| Cale Makar | 15.7 | 59% | 13-7 |
| Devon Toews | 16.7 | 57% | 12-5 |
| Martin Necas | 15.9 | 57% | 13-8 |
| Sam Malinski | 15.9 | 57% | 11-5 |
| Nathan MacKinnon | 15.9 | 55% | 14-9 |
| Nazem Kadri | 15.7 | 54% | 10-4 |
| Artturi Lehkonen | 15.1 | 51% | 15-7 |
| Brett Kulak | 16.5 | 51% | 8-6 |
Nathan MacKinnon has 16 points in 10 games, leads the team with 44 shots and is at 55 per cent at five-on-five. Cale Makar has 11 points in nine games and the best on-ice share of any regular. In August we wrote that Makar is so good the team barely shows up in his numbers, and that Colorado's real issue was conversion. This October the conversion has flipped completely, and Makar's share is still the best on the team.
The one line worth watching is Lehkonen's. He is at 51 per cent of the expected goals but has been on the ice for 15 goals for and seven against. That gap between chances and goals is the finishing streak in one player.
So why is Colorado losing close games?
Because the edge that won it 121 points last season has disappeared, and that edge was in net.
In August we argued that goaltending was not Colorado's problem. The numbers backed it. Scott Wedgewood saved 23.1 goals above expected over 45 games, fifth in the league, and Mackenzie Blackwood added 8.3 in 38. At five-on-five, Colorado allowed 128 goals on chances worth 154. That cushion of about 26 goals is how a team wins one-goal games.
Ten games in, it is gone. At five-on-five, Colorado has allowed 20 goals on chances worth 19.4, roughly what an average crease would concede. Blackwood has started seven of the 10 games, gone 4-3 and stopped 160 of 180 shots for an .889 save percentage. MoneyPuck has him 2.5 goals below expected. Wedgewood has made three starts and sits at .912, almost exactly average.
Our August thesis holds as a description of last season. It does not hold for this October, and it is fair to say so. Average goaltending is not a disaster, and nobody should expect Wedgewood to save 23 goals above expected again. But a team whose identity was winning tight games behind a hot crease is now playing tight games behind a neutral one, and the record shows it.
There is a smaller leak too. Colorado has scored four power-play goals from chances worth 6.4, and MoneyPuck counts three goals against while it was on the five-on-four. One of those was Arttu Hyry's shorthanded goal early in the third period of the shootout loss to Dallas, which turned a 1-1 game into a chase.
The case that Colorado is fine, stated fairly
The optimistic reading is simple. Colorado is fifth in the league in five-on-five expected goals share, has the best player in the world playing like it, and has a defenceman driving 59 per cent of the chances. A 1-3-1 record in one-goal games over 10 games is noise. Over a season, close games tend to even out for good teams, and the goaltending is far more likely to drift back toward last year than to stay below average.
The pessimistic reading is that the finishing has been carrying the record, and it is about to stop. When Lehkonen and Necas cool off, Colorado loses the blowout wins that built its goal differential. If the crease is only average when that happens, the close games will not suddenly start breaking its way. A team at 54 per cent with neutral goaltending is good, but it is not the 121-point team.
Our view sits closer to the optimists, with one change we would make now. Colorado is a top-five team that has been unlucky in close games and lucky with its finishing, and the two have roughly cancelled out. The fix is in net. Wedgewood was the better goaltender over a full season last year and has been the better one this October in a smaller sample. He should be starting closer to half of the games, not three in 10.
What to watch
The one-goal record, which should climb toward .500 with better saves, and Lehkonen's shooting percentage, which will come down. If the five-on-five save percentage climbs back above .920 while the chance share stays near 54 per cent, Colorado is a Cup contender again and this month is forgotten. If the finishing cools first, the next 10 games will be harder than the last 10. Kadri's first goal is coming too, and Necas has already shown why Colorado made the Rantanen trade. Every Avalanche number is on the team stats page, and the Central race is on the standings.