Ask anyone who has played both and they will tell you that shooting in NBA 2K27 is far harder than it was in NBA 2K26. We had the same reaction. So we did the one thing we can do about it: we ran the same green window test in both games, one millisecond at a time, on a station that takes the human hand out of the release.
The window did shrink. It did not shrink by anywhere near as much as it feels like it did, and the distance between those two things is what this page is about. If you only read one section, read the one on why it feels harder than it measures.
Mike Wang, the gameplay director on NBA 2K, has said more about 2K27 shooting in its first two weeks than in most full seasons. Four of his posts frame everything below. The first settles what kind of system this is.
@SSJerdo … can you confirm if it's green or miss again or is it back RNG with a pure green window attached
Mike Wang @Beluba “ That clip was captured with a hack to make all shots go in. Shooting in 2K27 is green or miss on all difficulties with a pure green window just like last year. ”
That is the same model as last year: there is a pure green window, hit it and the shot goes in, miss it and the shot does not. It is what makes the test on this page possible, and it is also why the second post matters so much.
Mike Wang @Beluba “ There's no RNG factor that will make you randomly miss shots. If you release your shot perfectly every time, you will make every shot. Your shot speed can vary based on a number of different factors (stamina, defensive pressure, the catch, pulling up off dribble vs. triple threat, sig shot grades… ”
Read the second half of that carefully, because it is the list of everything that changes your shot from one attempt to the next: stamina, defensive pressure, whether you caught the ball or pulled up, and the grades on your jumper. Each of those can move the window on the clock, and several of them, contests and stamina in particular, also shrink it. Our test holds every one of them still, which is the difference between the window you can measure and the window you feel in a game. The third post is his read on where percentages will land.
@TV_Bullet If your having trouble shooting on NBA 2K27 try changing your jumpshot speed and mix it up with your badge selection. Combining different things together might be what your missing
Mike Wang @Beluba “ this is good advice. you also might need to reset your expectations. if you're shooting over 40% from 3, you're already doing well. very few are going to be able to get close to last year's shooting percentages. ”
And the fourth is about the one shooting setting you choose yourself, which comes back up further down.
Mike Wang @Beluba “ If you're using a custom jumper, the release speed setting will have a small impact on your green window. Very Quick will reduce your green window size by 10% and Quick by 5%. Slow and Very Slow will actually give you a tiny boost. So choose what's more important for your… ”
Then, on 30 August 2026, he posted the community's actual shooting numbers for both games side by side. That table is reproduced in full in its own section below, because it is the best evidence anyone has for how much harder 2K27 has been in practice, and it lines up with what we measured in a way that is worth walking through.
Mike Wang @Beluba “ 2K26 vs. 2K27 community shooting stats to date in case you're interested. ”
Here is the table from his data. These are every shot taken by everyone, with nothing filtered out, four days into early access. Comp 5v5 is ProAm and Proving Grounds in 2K26 and Crew HQ in 2K27. The last column is the three point drop as a share of the 2K26 figure.
He also broke out players at 85 overall and above in 2K27, which is the closest thing to a like for like with a built up 2K26 player:
Mike Wang @Beluba “ Casual Corner numbers: FG%: 73.8% | 3PT%: 66.9% ”
Shots were taken on an automated testing station that holds the shot button for a set number of milliseconds and then releases it. The release timing is fixed by the device rather than judged by a human hand, and the result of every shot is read off the timing feedback on screen. Nothing in the chain has a reaction time.
Each game was tested in 1 millisecond increments from a 541ms hold to a 590ms hold, 100 attempts at every step, 50 steps per game, 10,000 shots in all. Both tests used the same build, the same jumpshot and the same spot on the floor, standing catch and shoot, wide open, with no badges, hot zones, takeover or boosts in play, on a full stamina bar, offline. The only thing that changed between the two tests was the game.
This year the controls are not a footnote, they are the point. Go back to Wang's list of what changes your shot speed. Every item on it changes the shot in a way the station has to hold still: some move the window on the clock, some shrink it, and some do both. A comparison that lets any of them vary is measuring the mix rather than the base window:
There is a built in check that the setup really was the same in both games. Two tests run on the same jumper should put the centre of the window in the same place. The centres landed at 563.1ms and 562.3ms, 0.8ms apart, which is inside a single sampling step. If the jumper, the spot or the setting had differed between the runs, that gap would have been tens of milliseconds and this page would not exist.
One more thing about reading the results, since it is not obvious. Because there is no RNG, a cell that made 62 out of 100 is not a 62% shot. It is the share of attempts on which the window happened to cover that millisecond, because the release animation does not land in exactly the same place every rep. That means the whole curve, not just its peak, is the window, and its width is what the two numbers at the top of the next section are measuring.
The pure green window for this shot is ~35.3ms wide in 2K27 against ~39.2ms in 2K26 , measured as the full width at half maximum of each curve. The area under each curve, which is the same width read a second way, gives 31ms and 34.2ms, a gap of 3.2ms, and it runs a little low on both because neither test was taken all the way out to zero. Either way the window lost about 10% of itself between the two games.
Where the window lost that width is more interesting than how much it lost. Split the hold times into six stretches and compare the two games in each:
Both games on one pair of axes, in real hold time. If 2K27 were dramatically harder to shoot in you would expect one curve to sit far inside the other. What you see instead is the 2K27 curve tucked a few milliseconds inside the 2K26 curve on each flank, meeting it at the top, with a shorter stretch along the ceiling.
Move across the chart, or use the slider, to compare the two games at one hold time.
So the window is a tenth smaller and everyone feels like they are shooting twice as bad. Both of those are true, and here is how a 3.9ms cut turns into the season everyone is having.
The window alone is worth about 10% of your makes. Contests and drift each take a share on top, and shares multiply. Carry all of them and you are closer to a fifth of your makes gone, which is what the REC and comp columns show, from a window that moved 3.9ms.
2K27 shipped an overhauled shot contest system. 2K's own gameplay report describes a Defensive Impact Indicator under the shooter that grades a contest by how hard it was and which direction it came from, from a small white arc up to a large crimson one, and Wang's early access briefing put eight tiers on it:
Our read after two weeks in the game is that contests are a little stronger than they were, in two ways. A contest of the same apparent quality knocks more off the shot than it did in 2K26, and the middle tiers, Bothered through Solid Contest, cover a lot of shots that would have registered as open last year. If you feel like you are getting contested more, you probably are, because the grading got finer and the arc now tells you about hands that used to go unmentioned.
The evidence for it that does not rely on feel is in Wang's own table. The base window is the same in every mode, so if the window were the whole story every mode would have dropped by about the same 10%. They did not. Park, the mode with the least defence, dropped 10% from three. REC dropped 18%, Comp 3v3 dropped 21% and Comp 5v5 dropped 15%. The extra loss shows up exactly where the defenders are, which is what you would expect if contests are taking a bigger bite this year and if the shot speed changes they cause are landing on a window with less room in it.
We have not yet tested contests so we cannot tell you how many milliseconds each tier removes. That test is on the list. Until it is done, treat this section as what it is: what the game looks like in play, and what the mode by mode numbers point at.
Every cell we recorded, unsmoothed, so nothing above has to be taken on trust. 100 attempts each.
Reset the target first. Wang put it at 40% from three being good, and the table backs him up: the whole population is shooting in the 30s from deep in every mode with defenders in it. If you are over 40% you are already ahead of the game, and if you are in the mid 30s you are average, not broken.
If you were a 70%+ shooter last year, that is still achievable, but not likely. Make sure you find a jumpshot that fits you and practice.
How to Shoot · Shooting Ratings Test · Mike Wang Dev Updates
Subscribe to our YouTube channel to see our badge tests the instant they are complete.