BLOCK SMASH PERMANENT UPGRADE GUIDE
Written and tested by Diligesker Lab · Upgrade catalog reviewed July 30, 2026 · Report a correction
Block Smash has two different layers of power. Rewards selected during a run shape that attempt. Permanent upgrades are stored in meta progression and applied to later starts. Coins earned at game over are added to the persistent balance, where they can be spent on these upgrades.
There is no universal first purchase because repeated failures can have different causes. A player who reaches few level-ups has a different problem from a player who sees enough rewards but dies to descent damage. This guide treats spending as a diagnosis: identify the repeated limit, buy a measurable answer, then compare the next set of runs.
Current permanent-upgrade catalog
| Upgrade | Current effect per level | First cost | Max level |
|---|---|---|---|
| Ledger Engine | Coin gain +3% | 560 | 12 |
| Combat Journal | XP gain +3% | 560 | 10 |
| Lucky Thread | Reward luck +0.02 | 600 | 10 |
| Tactical Draft | Level-up choices +1 | 13,500 | 1 |
| Future Sight | Starting rerolls +1 | 4,500 | 2 |
| Guardian Sigil | Starting Turn Guards +1 | 4,800 | 2 |
| Boss Cache | Boss reward choices +1 | 14,500 | 1 |
| Steel Hammer | Damage +2.5% | 900 | 8 |
| Keen Sight | Crit chance +0.6% | 980 | 8 |
| Execution Scope | Crit damage +4% | 1,000 | 5 |
| Acceleration Coil | Ball speed +2% | 880 | 5 |
| Spare Sling | Starting balls +1 | 3,200 | 1 |
| Guard Armor | Starting HP +1 | 3,000 | 2 |
| Counter Intel | Ghost, shield, and regen counter stats +10% | 2,400 | 4 |
The table contains all 14 permanent upgrades in the current catalog. It is a catalog, not a ranking. Ledger Engine and Combat Journal change resource flow. Tactical Draft, Future Sight, and Boss Cache change how many decisions can be seen or replaced. Steel Hammer, Keen Sight, Execution Scope, Acceleration Coil, and Spare Sling change the starting combat package. Guard Armor and Guardian Sigil address different kinds of survival pressure.
Why later levels cost more
The displayed next-level price is calculated as the base cost multiplied by the upgrade’s cost multiplier raised to the current level, rounded down. The first purchase therefore equals the listed base cost, while repeated levels become progressively more expensive. Before buying a high level in one line, compare its next price with the first level of an unowned upgrade that addresses the same failure.
A large price does not make Tactical Draft or Boss Cache automatically better than a smaller stat purchase. Their value depends on whether access to more choices is the actual constraint. If launches are inaccurate or a build has no boss-capable damage, another visible card does not repair that execution or planning problem by itself.
Path 1: runs level too slowly
Start the comparison with Combat Journal, which adds 3% XP gain per level. Its purpose is narrow and measurable: it changes XP flow, not damage or survival directly. After buying a level, compare the level reached at similar turn counts across several runs. If the level gap improves but the board still collapses at the same time, the original limit was not only reward frequency.
Tactical Draft and Future Sight belong to a related but separate problem. Draft adds one level-up choice, while Future Sight adds a starting reroll per level. They improve selection control after a reward screen appears; they do not generate XP themselves.
Path 2: runs end to descent damage
Guard Armor adds starting HP. Guardian Sigil adds starting Turn Guard charges, and each charge can cap one instance of turn-end damage at exactly one. These are not interchangeable. HP expands the damage budget, while a guard changes a particular turn-end damage event. Record which resource was actually consumed before deciding which line helped.
Counter Intel is the specialized option. Each level adds 10% to ghost, shield, and regen counter stats, up to four levels. Choose it when those block behaviors repeatedly disrupt otherwise workable runs, not merely because it has a broad-sounding name.
Path 3: contact is reliable but damage plateaus
Steel Hammer raises damage by 2.5% per level. Keen Sight adds 0.6 percentage points of crit chance, and Execution Scope adds 4% crit damage. Acceleration Coil raises ball speed by 2% per level, while Spare Sling adds one starting ball. These stats influence different parts of a volley, so the run summary should decide the experiment.
If the ball already makes repeated contact but individual hits are weak, test damage. If a crit-centered package appears often enough to evaluate, compare crit chance and crit damage. If one extra projectile would create another full path through the board, Spare Sling tests a different hypothesis from a percentage increase.
Path 4: boss rewards are the bottleneck
Boss Cache adds one boss reward choice to the starting meta bonuses. It is a single-level, high-base-cost purchase. Its value begins only after a boss is defeated and the reward screen is reached, so it should not be expected to help a run that consistently ends before the first boss. For preparation before that checkpoint, use the separate boss guide.
Refunds make controlled experiments possible
A single-level refund returns the calculated purchase cost of the latest level. Resetting all permanent upgrades returns the sum spent on all purchased levels and clears the upgrade allocation. That means a spending test does not have to become a permanent commitment under the current system.
Use that flexibility deliberately. Change one category, not the entire shop, then play five comparable runs. Record character, best turn, level, coins, boss reached, and the reason the board failed. Keep the allocation if the targeted measure improves. Refund and test another explanation if it does not.
A five-run worksheet
| Field | What to record |
|---|---|
| Purchase hypothesis | The exact repeated problem the upgrade should reduce. |
| Comparable checkpoint | Turn, level, or first boss reached. |
| Observed result | Damage, HP loss, choices used, rerolls used, or coins earned. |
| Failure cause | Board pressure, weak direct damage, poor reward control, or execution. |
| Decision | Keep, add one level, or refund and test another category. |
Permanent progression is most useful when it answers evidence from completed runs. The shop can strengthen a plan, widen its decisions, or protect a mistake, but it cannot replace reading the board and launching a shot the build can use.
Last reviewed July 30, 2026. Found an outdated rule? Send a correction.