The official Snipers Vs Runners description says wins can be used to upgrade speed and damage. It does not publish the price per level, maximum level, increase size, affected role, refund option, reset rule, or persistence. Treat the live upgrade menu as the source for every numerical or role-specific decision.
A good upgrade choice connects a displayed effect to a problem you can reproduce. Do not buy speed because every Runner must want it, or damage because every Sniper must want it; those role assumptions are not stated publicly.
Read the live upgrade card
Before purchasing, record the upgrade name, current level, next level, cost, currency or win requirement, displayed effect, and any cap or warning. Take the wording literally. If the card says an effect applies to one role or action, preserve that scope. If it gives only a name and price, the effect still needs controlled observation.
Also record the current balance. After purchase, verify the exact reduction and the new level. If either value does not change as expected, stop before buying again.
Choose by an observed problem
An upgrade can improve a valid action but cannot replace knowing where to go or whom to defend against.
Test speed without guessing percentages
Choose one repeatable path between two visible landmarks under the same role and current map. Record the upgrade level and compare travel or exposure only after changing one level. Avoid mixing a different spawn, route, device, latency state, or movement action into the comparison.
Do not publish a percentage from a handheld timer alone unless the method is repeatable and the game does not alter the route. The live menu may already state the effect more accurately. If it does not, describe the observed outcome cautiously rather than inventing precision.
Test damage without guessing health
Choose the same role, current weapon or attack, and a comparable target state. Record visible hit feedback and the outcome before and after one upgrade. A different target health, hit location, range, or round rule can change the result. One elimination does not establish a universal damage number.
Do not assume the scoped rifle shown in media is the only weapon, that a headshot exists, or that every Runner has identical health. Those details need current in-game evidence.
Cost, cap, and persistence record
Cost curve
Record each displayed next-level cost separately. Do not extrapolate later prices from the first purchase.
Maximum level
Use an explicit cap or completed menu state. A temporarily unaffordable level is not proof of a cap.
Persistence
Rejoin another session and verify the level. The public description does not promise how data saves.
Reset or refund
Use only a visible control and warning. Do not tell players to spend experimentally when recovery is unknown.
Record the date because balance updates can change the value of an old comparison even when the category name remains the same.
Common upgrade mistakes
Buying both categories together hides which one changed your result. Copying another player’s order ignores different roles, route knowledge, and current levels. Treating the word “wins” as a fixed one-point currency may also be wrong; confirm the displayed value and transaction.
Avoid Robux products that imply progression by name without a verified live effect. The API exposes off-sale Pass records named x2 Coins and x2 Skulls, but their functions are not described. Neither name proves that it changes win-based upgrade costs or is active in the current Experience.
A low-risk upgrade routine
Accumulate enough visible progress for one purchase, read the complete card, and choose the category whose confirmed scope matches a repeatable problem. Capture the before state, buy once, confirm the transaction, and repeat the same situation. Rejoin later to verify persistence before committing the rest of the balance.
Use Coins for the separate cosmetic currency, Progression for the complete system index, and Gamepasses to distinguish Roblox products from in-game upgrades.