- Home
- How We Score
What Goes Into a LegitBoosters Trust Score
Every rating on Legit Boosters begins the same way: I pay for an order out of the site budget, run it from checkout to delivery, and write down what actually happened instead of what the marketing page promised. That log now stands at 248 paid orders across the services we track, and the whole set gets re-audited every month - the most recent pass closed July 13, 2026. Nobody pays for a rating, a position, or a friendlier second look. If a claim cannot be traced back to something I watched happen on an order I paid for, it does not turn into a score here.
Our Scoring Model
The Trust Score runs 0 to 100 and comes out of five weighted inputs, applied in identical proportions to every service on the site, so a two-man Dota 2 shop and a marketplace with thousands of listings get graded by the same arithmetic. Account safety and handling carries 35 percent because it is the one category where a bad outcome cannot be undone with money. A late order costs you a week. A wrecked account costs you the account, and no chargeback puts a decade of cosmetics and rank history back where they were. Deadlines, pricing honesty, buyer track record and support all matter, and all of them matter less than whether your login survives the transaction.
| Input | Weight | How we measure it |
|---|---|---|
| Account safety and handling | 35% | Whether the service keeps the session hidden while it works, connects through an endpoint in the player's own country, and stores credentials somewhere other than a shared spreadsheet. I read what the site admits in writing about ToS exposure, since a vendor that pretends the risk is zero has already told me something. After delivery I audit the account itself: open sessions, password state, and any trace the person who logged in decided not to clean up. |
| Finishing on time | 25% | Did the paid work get finished at all, did it start when the site said it would, and did delivery land inside the quoted window. Mid-order pauses and silent reassignments to a second booster count against this input even when the clock is eventually met. So does a delivered result that does not match the order that was placed. |
| What the price really is | 15% | I compare the figure on the product page against the figure the payment form actually charges, then sweep the calculator across several rank ranges to find where the curve jumps or the discount quietly expires. Fees that surface only after the email field get flagged by amount. Written refund terms have to be readable on the site before payment, not produced by support afterwards. |
| What buyers report | 15% | Reviews with a verifiable purchase behind them, weighed against the ones that arrive in clusters within the same hour and share a vocabulary. I read for the shape of the complaints and for how the company answers them in public, because a vendor with a short record of fixing real failures is worth more than a taller pile of praise nobody can source. |
| Quality of support | 10% | Reply speed on live tickets opened during genuine orders, and, weighted heavier, whether the reply is correct. Each service also gets one deliberately awkward request - a refund edge case, or a direct question about ban exposure with no comfortable answer - to see whether the desk improvises, deflects, or tells me the truth. |
Trust Score Bands
A finished number lands in one of four bands: 90 and above earns Excellent, 75 through 89 reads as Good, 60 through 74 settles at Fair, and anything below 60 is Poor, which in practice means I would not spend my own money there twice.
What We Check
- Refunds, Tested Live. Written refund terms are a claim until somebody tries to use them. On selected orders I deliberately walk the refund path - a start that never happened, a result that came back wrong - and file the request exactly as an ordinary buyer would, then log the reply time and what actually returned to the card. Two vendors in the current run have policies that read cleanly and pay out slower than their own text commits to. That gap goes into the review, not into a tidy policy summary.
- Checkout, Line By Line. I capture the calculator total, then the payment screen, then the receipt, and compare all three against each other. Currency conversion, express toggles that default to on, and processing additions that appear only after the email field are where a quoted figure stops being the figure you pay. Anything that moves between the first step and the charge gets recorded with its amount and reproduced on a second order before it goes into the writeup.
- After The Handover. Once an order closes I go back into the account and inspect it the way an investigator would: active session list, login locations and devices, anything linked or unlinked while the booster had access, and how much friction a password change runs into. Sessions still open days after delivery are a red flag by themselves, however good the rank result looked, and across 248 orders that detail has separated vendors more reliably than any promise on their homepage.
Our Independence
No service buys a position on this site. Every order in the log was paid for from the same budget that pays the hosting bill, and three rules hold that line:
- Identical procedure and identical weights for every vendor, from the largest marketplace to the one-game shop nobody has heard of. No bonus points for a partnership offer, no deduction for a company that argued with me over email.
- When a reader or a vendor proves something here is wrong, the fix goes on the page itself with the date it changed, and the corrected claim stays visible. Quiet overwriting is how a review site burns its own track record.
- Orders are placed anonymously, through ordinary buyer accounts with ordinary buyer names, using my money. No vendor learns it was being audited until the review is live.
Report an Error
I get things wrong. Prices drift, staff turn over, and a refund policy verified in one quarter may not survive to the next. If something on this site does not match what you saw, email factcheck@legitboosters.net with the page URL and what you believe is wrong. Every disputed claim gets checked back against the order logs, receipts and ticket transcripts it came from before anything is rewritten or left standing.
The Person Doing the Testing
Derek Lawson is Research Director at Legit Boosters. He works as a statistician and has been playing Dota 2 for fifteen years, which was long enough to get tired of review counts that nobody could source. So he started buying orders and checking the claims against what arrived. He pays for the 248 orders in the log personally, runs the monthly re-audit, and signed off on the most recent pass on July 13, 2026.