New York IslandersThe Rangers are 4-1 and have won four straight. Alexis Lafrenière scored a hat trick in Tuesday's 5-2 win over the Islanders, Igor Shesterkin has been excellent, and Dylan Garand made 22 saves for a shutout in Detroit. After a 77-point season that finished last in the Eastern Conference, the question everywhere this week is what has changed.
Less than the record suggests. The Rangers have one new strength, a power play generating chances at a higher rate than last season. The rest is the same team as last season with the same goaltender, and that formula has already failed once.
What is driving the start?
Mostly the crease. New York has allowed eight goals on 13.6 expected, so its goaltending has saved 5.6 goals more than average, the most in the league. Shesterkin has allowed seven goals on 11.1 expected in four starts, and Garand's shutout came on 1.7.
At five-on-five the team is close to neutral. The Rangers have 49.3 per cent of the expected goals, 20th in the league, and have scored eight while allowing five. The scoring is not running hot either. They have scored 1.4 fewer goals than their chances at five-on-five.
| Rangers, 2026-27 | Goals for | Goals against | Expected for | Expected against |
|---|---|---|---|---|
| All situations | 16 | 8 | 17.7 | 13.6 |
| 5-on-5 | 8 | 5 | 9.4 | 9.7 |
| 5-on-4 power play | 3 | 0 | 3.6 | 0.2 |
Is this different from last season?
Not by much. Last October the Rangers also started well defensively, with seven goals allowed in their first five games. They went 2-3 because they scored 10. This October they have allowed eight and scored 16.
| First five games | Record | Goals for | Goals against |
|---|---|---|---|
| 2025-26 | 2-3 | 10 | 7 |
| 2026-27 | 4-1 | 16 | 8 |
Across the whole of last season the pattern held. Shesterkin allowed 126 goals on 147.3 expected, more than 21 goals better than average, and the Rangers still finished with 77 points because at five-on-five they took only 48.5 per cent of the expected goals. This season's 49.3 is almost the same team.
What is genuinely better?
Two things. The power play is creating chances at 10.2 expected goals per 60 minutes, up from 8.8 last season, when it was already third in the league. And Lafrenière looks like the real thing.
His hat trick was not a fluke of finishing. He has three goals on 2.5 expected and 11 shots on goal in five games. Last season he scored 24 goals on 23.3 expected, with a 52.5 per cent on-ice share at five-on-five, one of the better marks on a team that was under water. At 24, he is turning into exactly the player the first overall pick in 2020 was supposed to become, and he is doing it on his chances, not on luck.
The forwards who have been unlucky so far make the same point from the other direction. Pavel Dorofeyev has one goal on 15 shots and 1.8 expected. Will Cuylle has none on 14 shots and 2.3 expected. The team's goal total is not being propped up by everyone shooting hot.
The case for the Rangers, stated fairly
The best version of the optimistic read goes like this. Shesterkin is one of the best goaltenders in the world and saving goals above expected is what he does every season, so leaning on him is a strategy, not luck. The power play is better, Lafrenière is a genuine top-line winger, and Adam Fox is healthy and on five points. Four wins in a row are worth having in the bank whatever the underlying numbers say.
The other side has a full season behind it. A team that takes 49 per cent of the five-on-five chances needs its goaltender to be extraordinary every night, and last season he was, and it still was not enough for the playoffs. Nothing in the first five games says the even-strength play has moved.
The Rangers have the same problem as last season with a better start. The 4-1 record is a cushion, not a turnaround.
What to watch
The five-on-five expected goals share is the number. If it climbs above 52 per cent by the end of October, something real has changed, and the cushion becomes a platform. If it sits near 49, this is last season with better bounces early.
We made a version of this argument in August, that the Rangers kept blaming the wrong things, and in September when they paid heavily for a power-play specialist. The power play part of that plan is working. Track it all on the standings and the team stats page.