How we test and score Cronus Zen scripts
Every Cronus Bench score comes from the same eight tests, weighted the same way for every script. This page sets out what each test measures, how we run it and how the result becomes a number, so you can check any score we publish against the rules that produced it.
The weights
A script's overall score is a weighted average of eight dimension scores. The weights below were fixed before we published any ranking, and they apply to every script on every board, whoever sells it.
| Dimension | What it measures | Weight |
|---|---|---|
| Shot timing repeatability | How often the script lands on-time releases, and whether it holds up on a second day | 25% |
| Feature depth | How much the script does, counting only features that work as advertised | 15% |
| Menu clarity | Whether a new owner can find and change the main settings without the manual | 15% |
| Setup friction | Time and steps from purchase to a working script on the Zen | 15% |
| Input passthrough and ban-safety posture | Whether the controller behaves as stock with the script off, and how honestly the seller describes risk | 10% |
| Update cadence | How quickly the script is updated after game patches | 10% |
| Support | How quickly and usefully the seller answers a standard question | 5% |
| Price to value | Twelve-month cost against what the script delivers on the other seven dimensions | 5% |
| Total | 100% |
Shot timing carries a quarter of the score because it is the reason most people buy a basketball script. Feature depth, menu clarity and setup friction share the next 45%: they decide whether you can get the script working and keep it set the way you want. Support and price carry the least. They matter, but a cheap script with quick replies that misses its timing is still a poor buy.
How the overall score is worked out
Each dimension gets a score from 1 to 10, in half points. We multiply each score by its weight, add the results and round to one decimal place, with halves rounding up. Here is the arithmetic for an example script. The scores are made up to show the method. They are not a test result.
| Dimension | Score | Weight | Points |
|---|---|---|---|
| Shot timing repeatability | 8 | 25% | 2.00 |
| Feature depth | 7 | 15% | 1.05 |
| Menu clarity | 6 | 15% | 0.90 |
| Setup friction | 9 | 15% | 1.35 |
| Input passthrough and ban-safety posture | 7 | 10% | 0.70 |
| Update cadence | 8 | 10% | 0.80 |
| Support | 6 | 5% | 0.30 |
| Price to value | 5 | 5% | 0.25 |
| Overall | 7.35, shown as 7.4 |
Every score on a review links back to the section of this page that explains it.
What the numbers mean
| Score | Meaning |
|---|---|
| 9 to 10 | Among the best we have tested on this dimension, with nothing to warn you about. |
| 7 to 8.5 | Good. Small faults most owners will live with. |
| 5 to 6.5 | Usable, with a weakness you will notice. |
| 3 to 4.5 | Poor. It works, but we would not pick it for this. |
| 1 to 2.5 | Broken, missing or misleading. |
What every test records
So that anyone with the same setup can repeat a result, every review states:
- the script version tested
- the date of the test
- the platform, meaning the console
- the controller
- the firmware on the Cronus Zen we tested with
- the game version at the time
If any of these change, the old result no longer describes the script you would buy, so we test again.
Shot timing repeatability 25%
How often the script produces a release the game grades as on time, across a fixed set of attempts, and whether that rate holds up on a second day.
How we test
We test in the game's offline practice modes only. A session is 200 attempts with the same player and the same jump shot: 100 open catch-and-shoot jumpers and 100 off the dribble. We log the game's own timing grade for every attempt. Then we run the same 200 attempts again on a different day.
Every script on a board is tested on the same console, display and controller, with the display in its low-latency game mode and the controller wired, so the setup does not favour one script over another.
How it is scored
The score uses the weaker of the two sessions. A script that is perfect on Monday and patchy on Tuesday is not repeatable, and the weaker day is the one you will remember. This dimension is scored in whole points.
| On-time attempts in the weaker session | Score |
|---|---|
| 97% or more | 10 |
| 93% to 96.5% | 9 |
| 88% to 92.5% | 8 |
| 82% to 87.5% | 7 |
| 75% to 81.5% | 6 |
| 65% to 74.5% | 5 |
| 55% to 64.5% | 4 |
| 45% to 54.5% | 3 |
| 35% to 44.5% | 2 |
| Below 35% | 1 |
We do not test in online modes. Results there depend on connection and server conditions we cannot hold constant, and using a script online puts your account at risk. See ban-safety posture below.
Feature depth 15%
How much the script does, counting only features that work.
How we test
We list every feature the seller advertises for the game and check each one during the offline sessions. A feature counts if it does what the product page says. Each one that does not is listed on the review as a failure.
How it is scored
Feature depth is scored against the board. The script with the most working features sets the 10, and the others score in proportion, rounded to the nearest half point. Each advertised feature that does not work takes off a further point: a missing feature is honest, and a broken one is not. Because the scale is relative, a script's feature score can fall when a deeper script joins its board. When that happens, the review says so.
Menu clarity 15%
Whether someone new to the script can find and change the main settings without reading the manual.
How we test
Before opening the manual, the tester does three tasks: turn the script off and on, change the timing adjustment by one step, and switch to a different saved setting or profile. We time each task and note every wrong turn. Then we make three checks:
- the current state is shown, on the Zen's screen or in the game
- labels use plain words rather than codes or slot numbers alone
- settings survive switching the Zen off and on again
How it is scored
Start at 10. Take off a point for each task that takes longer than a minute, a point for each task that cannot be done without the manual, and a point for each check that fails. The lowest score is 1.
Setup friction 15%
How long it takes, and how much can go wrong, between paying for the script and taking a first shot with it.
How we test
We start from a Cronus Zen on current firmware and a Windows PC with nothing script-related installed, then follow the seller's own instructions. We time the whole process, count the steps and log anything that failed or needed a second try. A step is any separate thing the buyer has to do: make an account, install software, enter a licence key, copy a file, choose a slot.
Many sellers ship a file that you load with Zen Studio, the Cronus Zen's own PC software, which theofficial Cronus Zen guide covers. Others let you program the Zen from a web page, with nothing to install.
How it is scored
| Result | Score |
|---|---|
| Under 5 minutes and 5 steps or fewer | 10 |
| Under 10 minutes and 8 steps or fewer | 8 |
| Under 20 minutes and 12 steps or fewer | 6 |
| Under 30 minutes | 4 |
| 30 minutes or more, or a step we could not finish without help from the seller | 2 |
A result that meets the time limit of one row but only the step count of the row below gets the odd score in between. Each step that failed and had to be retried takes off another point, down to 1.
This dimension favours scripts that load from a browser, because there is no software to install. Cronus Zens, which shares owners with Cronus Bench, sells its scripts that way. We say so here so the effect is visible: in a close ranking, setup friction is often where a Cronus Zens script gains ground. Browser loading is not unique to Cronus Zens, and other sellers who offer it get exactly the same credit.
Input passthrough and ban-safety posture 10%
Two halves, each marked out of 5 and added together: whether the controller behaves like a stock controller when the script is off, and how honestly the seller describes the risks of using it.
How we test
Passthrough. With the script loaded but switched off, we run the controller through the Zen into a PC controller test and compare every button, both sticks and both triggers with the same controller plugged in directly. Five marks if nothing differs, less one for each input that does not match, down to zero.
Posture. We read the seller's product page, FAQ and terms. Five marks to start. We take off two for any claim that the script is undetectable, ban-proof or safe to use online, because no seller can promise that. We take off one if the script cannot be switched off from the controller, and one if the seller gives no risk information at all.
How it is scored
The two halves are added to give a score out of 10. A total of zero is scored as 1.
We do not test whether a script can be detected, and we do not give advice on avoiding detection. No script is ban-proof. Using a Cronus Zen in online play can breach a game's terms of service and lead to action against your account. NBA 2K27 is published by 2K, part of Take-Two, and its rules are inTake-Two's terms of service.
Update cadence 10%
How quickly the seller updates the script after the game changes.
How we test
For the 90 days before the test date, we list every game patch, using the publisher's patch notes and theofficial NBA 2K27 site, and match each one against the seller's changelog, release notes or announcement channel. For each patch we count the days until the seller's next update, or until the seller says publicly that no update was needed. The measure is the median of those counts. If the game had no patches in that window, we use the 90 days before it.
How it is scored
| Median days from patch to update | Score |
|---|---|
| 2 days or fewer | 10 |
| 3 to 5 days | 8 |
| 6 to 10 days | 6 |
| 11 to 20 days | 4 |
| More than 20 days, or no public record of updates | 2 |
A seller with no public changelog loses a further point, down to 1, because buyers cannot check the record either.
Support 5%
How quickly and usefully the seller answers a buyer.
How we test
We send the same two questions to every seller, through the support channel they publish, as an ordinary buyer: one about changing a setting in the script, and one about using it on a second console. We do not say we are reviewing. We record the time to the first reply that actually answers each question.
How it is scored
| Result | Score |
|---|---|
| Both questions answered within 12 hours | 10 |
| Both answered within 24 hours | 8 |
| Both answered within 72 hours | 6 |
| One answered within 72 hours | 4 |
| Neither answered within 72 hours | 2 |
An answer that was wrong, or did not address the question, takes off a further point, down to 1.
Price to value 5%
What the script costs over a year, against what it delivers.
How we test
We work out the twelve-month cost at the listed price on the test date: the one-off price, or twelve months of a subscription, plus anything else you have to buy to use it. The Cronus Zen itself is left out, because every script needs one. We then compare that cost with the median for the scripts on the same board, and the script's score on the other seven dimensions with the board's median.
How it is scored
Start at 6. Add 2 if the script costs less than the board median, or take off 2 if it costs more. Add 2 if its score on the other seven dimensions is above the board median, or take off 2 if it is below. A script that costs more has to earn it elsewhere to keep a good mark here.
Re-testing and dates
Every review is re-tested and re-dated once a month. If a game patch changes shooting, we run the shot timing test again before the next monthly date and note it on the review. The date at the top of a review is the date of the last test, not the last time we edited the words. When a score changes, the review shows the old score, the new one and the reason.
What this method does not cover
- Online play. Every result comes from offline modes. We do not claim a script performs the same way online.
- Other kinds of script. Version 1.0 is written for basketball scripts. Shooter scripts need a different repeatability test, and they will get their own method version before we publish a board for them.
- Your setup. Your display, connection and controller are not ours. Treat the timing numbers as a fair comparison between scripts on one setup, not a promise of what you will see at home.
Who sets the method
Cronus Bench is owned by the team behind Cronus Zens, which sells scripts that appear on our boards. The weights on this page were set and published before any board, and they apply to Cronus Zens scripts exactly as they apply to everyone else's. The about page explains how we handle that conflict.
Method changelog
| Version | In effect from | Change |
|---|---|---|
| 1.0 | First published. Eight dimensions, weights fixed as listed. Covers basketball scripts only. |