How to Build a Game Progression Model
A game progression model links XP, match time, resource balances, and unlock rules to predict when players can use a new hero, upgrade a card, or field a viable alternate build.
A player just unlocked a second hero. The profile says progress. The player sees a choice: try someone new with a weaker build, or return to the hero they already know and have invested in. The economy determines whether the unlock feels like a new way to play or another item in a collection. To predict that moment, a model has to connect time, match results, rewards, hero investment, and card eligibility.
This article follows the structure of a progression model built for a competitive game with a roster of heroes and a shared card collection. The workbook connects a summary of the player journey to assumptions, calculations by profile level, and configuration tables for implementation. The numbers in the worked example below are illustrative.
Jump to the model-building guide
Design the journey in three units
XP tells the system when to open a gate. Matches tell the designer how many opportunities a player has had to learn. Minutes and active days tell the team how long the wait may feel. Track all three. In a game with five-minute matches, a milestone 12 matches away takes about an hour of combat. It may fit into one long visit or stretch across four visits for someone who plays three matches at a time. If the promise is “a new hero by the end of the first session,” calculate against first-session length, not the average length of a later visit.
Keep the boundary of a level explicit. The XP required to start level 4 is different from the XP required to finish it. Rewards paid when the player reaches level 4 enter the wallet at that boundary. An unlock that appears at the start of a level should be tested against the starting balance and starting roster, even if the summary also shows what the player can afford by the end.
Translate match results into progress and purchasing power
Use a reward schedule by finish position. In a five-player match, third place can serve as a baseline scenario. It is not automatically the expected payout. With a steep first-place reward, a player who alternates between first and fifth can earn a different amount from one who always finishes third, even though both average third place. To estimate XP and soft currency per match, multiply each rank’s reward by that rank’s probability and sum the results separately for each resource.
Fast progression can reinforce itself. Better finishes pay more XP and currency, currency buys upgrades, and stronger upgrades may affect later finishes. Matchmaking can offset some of that advantage as players face stronger opponents. Model that as a hypothesis: compare a case in which early success regresses toward the baseline with one in which it persists. Look at the spread in time to the next unlock and whether early losses leave players with a credible route back into competition.
Give profile, heroes, and cards distinct jobs
The profile controls when the roster and collection expand. Hero investment deepens a specific play style through attributes such as vitality, damage, haste, or resilience. Card progression expands or improves the actions available in a build. A useful model can tell you at any milestone which heroes the player owns, what each can do, which cards are unlocked, and which of those cards can actually be equipped on each hero.
Resist the temptation to balance distinct heroes to a single equal score and stop there. A model can use a weighted power score to translate attributes into upgrade tiers, but equal scores do not make a hunter, a front-line defender, and a support hero interchangeable. A survivability upgrade may buy a defender another turn to protect the team; a haste upgrade may let a fragile attacker act before a counter. Check who benefits from each attribute and each card in actual matchups. If one hero has a strong starting kit, it may require a different upgrade curve even if its eventual score is similar.
Cards introduce two gates. A profile threshold can unlock a card in the collection, while a minimum hero level can determine who can equip it. Keep those states separate. A late-game protection card that opens on the profile but can only be used by an upgraded secondary hero may encourage experimentation. If the secondary hero is realistically weeks away from meeting the requirement, the same reward becomes a preview of content the player cannot use. A third metric, affordable to upgrade, tells you whether the collection feels usable or merely broad.
Model what players choose to invest in
The source model assumes players concentrate hero investment on a main and a secondary hero, even after the roster expands. It treats cards as a broader investment because different cards serve different tactical purposes. Both are hypotheses. Match share and spend share can diverge: a player may use a familiar hero for most matches but save currency for a new one, or play several heroes while upgrading a small set of shared cards.
Represent those decisions with separate weights for hero use, hero investment, and card investment. Renormalize the weights across eligible choices each period. Run a main-heavy case, a rotation case, and a case where the player switches mains after an unlock. Compare the time needed for a new hero to reach a playable build, the number of usable cards per hero, and how much old investment still helps after the switch. If switching always leaves the player underpowered, a starter upgrade or portable card can reduce that cost. If all investments transfer too easily, test whether mastery of a specific hero still feels distinct.
A worked example of how the gates interact
Suppose a profile milestone requires 600 XP and a reference player earns 50 XP and 25 soft currency (SC) per five-minute match. That is 12 matches, 60 match minutes, and 300 SC from matches before other rewards. If the player starts with no balance and assigns 30% of the earnings to hero progression and 70% to cards, the budgets are 90 and 210 SC. Suppose the hero budget converts at one attribute point per 100 SC. If points are whole units, the player can carry the fractional entitlement forward or keep the unspent currency, depending on the game’s rules. The model has to match that rule.
A card with a cumulative upgrade threshold of 200 SC could hit it if the player concentrates the 210 SC on that card. If they split the budget equally across three cards, each receives 70 SC and none reaches 200 from a zero starting investment. Now add a hero-level requirement: even the upgraded card cannot help the new hero until that hero qualifies. An economy review should ask whether that gap creates a worthwhile near-term goal or discourages the first trial of the new hero.
How to build the model in a spreadsheet
- Specify the first playable milestones. Write down when players should see their second hero, first meaningful upgrade, and first viable alternate build. For each milestone, record the intended experience and whether the target is measured in matches, minutes, or active days.
- Build an Assumptions tab. Enter match duration, minutes by active day, rank probabilities, XP and SC by rank, daily rewards, and level-up grants. Add a hero table with stable IDs, base stats, unlock thresholds, conversion rules, and per-hero tiers. Add a card table with stable IDs, functions, profile and hero gates, and cumulative upgrade costs. Keep starting balances, reward events, and spend weights visible inputs.
- Build the XP and time ladder. Use one row per profile level with start XP, end XP, expected XP per match, cumulative matches, and cumulative minutes. For a constant expected XP rate, a milestone estimate is
ROUNDUP(CumulativeXP / ExpectedXPPerMatch, 0). Derive incremental matches by subtracting consecutive milestone estimates so XP left over from crossing a boundary is not lost. Convert minutes to active days using a cumulative play calendar. For variable rewards or exact boundary behavior, simulate matches and carry the actual XP forward. - Roll forward the wallets. Calculate
Closing SC = Opening SC + SC earned − SC spent. Carry closing balances to the next row. Keep XP, SC, premium currency, and attribute points in their own units. A common value can help compare grants for reporting. Only add converted units to a playable wallet when the game actually permits the conversion. Award each daily and milestone grant once. - Calculate hero and card states. Allocate the hero budget to available heroes using normalized investment weights and apply each hero’s specific upgrade rules. Carry cumulative card investment separately by card ID. Use the highest affordable cumulative threshold for the new card tier. If the implementation charges for each tier, subtract the incremental cost from the prior paid tier and reconcile it to the wallet. Calculate card usability for each hero after applying both the profile and hero gates.
- Build a journey summary. For each milestone, show start and end time, roster, hero levels, cards unlocked, cards usable per hero, and balances. Compare a low-activity case, the baseline, and a faster player, then stress-test main-switching and higher concentration. Check the results against your milestone targets before tuning a reward or moving a gate.
- Generate the Tech Center tables. Export the XP ladder, rank rewards, hero and card gates, costs, and any randomized reward weights using stable IDs. Validate that probabilities sum to 100%, thresholds increase, no balance goes below zero, and locked items cannot receive spend or enter a loadout. Test the first unlock, a match that crosses two levels, a hero switch, and a card that is unlocked but still ineligible.
- Calibrate with playtests and telemetry. Log finish rank, match duration, hero selection, unlocks, equips, and wallet movements. Compare the predicted and observed time to a first usable build, split by session length and hero preference. When a milestone misses, diagnose the path: too little XP, insufficient currency, concentration on one hero, or an eligibility gate. Change the input that explains the gap and run the journey again.



