Audamir

part 38

How many hours to level 30? The rewards of Kastgard and Rokgard in one table

Level 30 costs 22,500 XP in both realms, but the hours range from 38 to 151, and up to 326 at a slower pace. A table of pace, five causes of the difference and five balance options. It is an analysis: nothing has changed in the game.

· 16 min read

Audamir has two realms with their own level, XP and currency, but with one curve: level 30 starts at 22,500 XP in Kastgard and in Rokgard alike. A note from playing on 1 November 2026 says that Golden Dice grow unusually fast compared with the tokens. In Kastgard it is, according to the note, a matter of luck: most games go through Principalities, a bad roll of the dice can be restarted, and the campaign with bonuses can be farmed faster. In Rokgard a bad start cannot be reversed, and a win is not certain even against a weak opponent. This analysis puts numbers on it from the documents and the code: the hours to level 30 range from 38 to 151, and up to 326 at a slower pace. Nothing has changed: neither the rewards nor the curve move in the game, and the proposals at the end are options, not decisions.

One curve for two worlds

A level n (from the second on) starts at 25 × n² XP (js/data/hero-level.js: A = 25, P = 2): level 2 at 100, level 10 at 2,500, level 20 at 10,000, level 30 at 22,500. Rokgard has its own XP but the same curve; the package of the realm registers it in the registry Economy. Both manifests write level: { curve: { A: HeroLevel.A, P: HeroLevel.P } }, and tools/core/realms.js checks that they agree. XP is paid only for won games (in Kastgard Principalities, free play, the daily challenge, a campaign run, the first stars of minigames and puzzles; in Rokgard Colony and Lines), not for coins and achievements. The hours to level 30 are therefore simple arithmetic:

hours = 22,500 ÷ pace (XP per hour)

At 300 XP an hour, both realms need the same hours (level 10 about 8, level 30 about 75), so the sum of the levels, the reputation in Asgard, compares like with like. 75 hours is the design target, not a measurement.

Table: hours to level 30

The paces come from xp.js --check, from pay.js --profile and from the documents TUNING and PAY-BY-MOVES. **The hours column is the arithmetic 22,500 ÷ pace**, calculated from the pace in the table when this was written, not a number from a document.

Kastgard (Golden Dice in the arena are paid the same as XP):

RowXP per hourHours to level 30Kind
Principalities, as I play them (19 timed games, 200 minutes, losses included)59437.9measurement on one save
Iron Run, 3★, threat 0 (with a chance to win)569 – 60037.5 – 39.5model
Campaign, run on 3★ / 1★560 / 40140.2 / 56.1model, 74.1 min per run
the same with a 93% chance to win the missions521 / 37343.2 / 60.3model
Free play, default setting41054.9model
Principalities, the player wins every game: island 2 players / field 2 players / field 8 players316 / 278 / 24571.2 / 80.9 / 91.8model
Average player (AI of level 2) today: duel / 4 players / 8 players431 / 211 / 12152.2 / 106.6 / 186.0model

Rokgard (tokens = XP, 1 for 1). All of it is a model: 30 s a game, 5 s a move for the human, the players are AI opponents.

Game and rowXP per hourHours to level 30
Colony, win every game against Weak / Normal / Strong AI150 / 308 / 592150.0 / 73.1 / 38.0
Colony, level 2 player against Normal (win and half of the draws, 0.485)149151.0
Colony, level 3 player against Strong AI (0.46)27482.1
Colony, average player against Weak AI (about 60%, the tool counts 0.63)94239.4
Colony, win every game against Normal at 3 / 8 / 12 s a move465 / 205 / 14148.4 / 109.8 / 159.6
Colony, level 2 player against Normal at 3 / 8 / 12 s a move225 / 99 / 69100.0 / 227.3 / 326.1
Lines, Drop five against Normal: best player / player of own level293 / 14376.8 / 157.3
Lines, Drop five against Strong AI: best player / player of own level561 / 28040.1 / 80.4
Lines, boards tic, 9×9 / 4, 7×7 / 4, Hex four, 15×15 / 3, 5×5 / 40never

What follows from this

Hours to level 30 for selected table rowsHorizontal bars: hours to level 30 for thirteen rows of the tables of the article; Kastgard in one colour and Rokgard in the other; the dashed line is the design target, 75 hours.Principalities, as I play them37.9Campaign, run on 3★40.2Free play, default54.9Principalities, winner, island 2 players71.2Principalities, winner, field 8 players91.8Average player, 8 players186.0Colony, win against Strong AI38.0Colony, win against Normal AI73.1Colony, level 3 player, Strong AI82.1Colony, level 2 player, Normal AI151.0Colony, average player, Weak AI239.4Colony, level 2 player, 12 s a move326.1Lines, player of own level157.3KastgardRokgardDashed line: design target, 75 hours (level 30 at 300 XP an hour)
Hours to level 30 for selected table rows. The dashed line is the design target, 75 hours; the numbers are from the tables of the article
  • The range is 38 to 326 hours. The target of 75 hours is hit by the model of Principalities (71 – 92) and by a Colony player who wins everything against Normal (73). At the pace with which I play Principalities, half the target comes out (38), for a typical Colony player double (151).
  • The ratio of the pace in Principalities, as I play them, to the pace of a level 2 player in Colony is 4.0 (594 ÷ 149 = 3.99). For the same type of player the ratio is smaller: the average Principalities player with four players (211) and a level 2 player in Colony (149) are in the ratio 1.4. Kastgard, however, pays by the number of players in the ratio 3.6 (duel 431, eight players 121).
  • Lines pay nothing on six boards (TUNING §43.4) and on the others pay 0.74 – 0.95 times the rate of Colony (§43.5).

The table does not count one-time content: minigames (18) and puzzles (28) pay once per star, 3,075 XP for three stars everywhere and 7,975 XP in all, that is level 17, when everything is at the maximum (xp.js --check). That is 35.4% of level 30. The rest comes out at 24.5 hours at the pace of 594, and at 52.2 hours with the model (÷ 278). The time to collect the stars was not measured. The daily challenge pays XP only for the first win of the day (score × 1.5, about 150 XP at a score of about 100).

Coins: Golden Dice and tokens

The observation about the coins does not follow from the curve of the level but from who pays what (PAY-BY-MOVES 1.2 and 1.5, Profile.coinsFor, Xp.board):

SourceXPCoinsCoins per hour
Principalities and free play, winscoreGolden Dice = XP594 (me), 245 – 316 (model)
Daily challengescore × 1.5, first win of the day onlyscore × 1.5 for every win–
Campaign, run on 1★ / 2★ / 3★495 / 594 / 6923,020 / 3,778 / 4,531 Golden Dice (6.1 – 6.5 per XP)2,270 – 3,410
Iron Run483 – 692none (Iron Dice)–
Minigames and puzzlesonce per starabout 0.5 – 1 per XP, once–
Kastgard achievements0291 steps, 51,922 Golden Dice–
Colony and Linesby movestokens 1 : 194 – 592
Rokgard achievements0none (proposal: 7 chains, 2,643 tokens)–
Rokgard daily challenge–none (challenges.daily is false)–

An hour of Principalities at the pace with which I play gives 594 Golden Dice, and an hour of Colony against Normal gives 149 to 308 tokens, that is 1.9 to 4.0 times as many. An hour of the campaign gives 2,270 to 3,410 Golden Dice, 7.4 to 22.9 times as many. Six Golden Dice per XP stays: a decision of 30 October 2026 (ROADMAP ch. 27, answer 12). In the save the ratio of Golden Dice earned to XP is 3.18 (48,063 ÷ 15,098), in Rokgard from the rule 1.00.

Achievements are the second source and do not concern the level. The list before the chains had 138 achievements for 18,020 Golden Dice (the old list); today's chains pay 51,922, that is 2.9 times as much, which at the pace of 594 is about 87 hours of play. The proposal of Rokgard achievements (R3g4) has 2,643 tokens, 1/19.6 of Kastgard. The numbers are not equivalent (most of the Kastgard total sits in long tails), but they show the direction. Golden Dice have somewhere to go (the box of 150, the steps of the Golden world 454 to 65,446), tokens do not yet: the wardrobe of Rokgard (R3e) is not built.

Why the difference

1. I win faster than the model. The model of time counts Principalities as 3 + 0.5 × rounds in minutes at the par number of rounds (24 – 38), which is 316 XP an hour. Wins with three players last on average 14.9 rounds (16 wins, TUNING §42.8). Colony was set to 300 XP an hour as the pace of Principalities (30 October 2026), but PAY-BY-MOVES 1.4 found that 300 is the pace at par rounds, while Principalities, as I play them, pay about twice as much. Colony also rests on an estimate of 5 seconds a move.

2. I win almost everything, the Colony player half. In the save 92% of the wins are in Principalities. The AI levels of Colony are a ladder without overlap: the AI beats a lower level in 95 – 100% of games and a higher one in 0 – 2.5% (TUNING §40.1). A player on his own level therefore wins about half: level 2 against level 2 won 45.8%, drew 3.7% and lost 50.5%; an average player against level 1 won 55.5%, drew 9.0% and lost 35.5%. A loss pays nothing. Doubling the reward for each AI level (0.5 / 1 / 2) is the rule against farming the weakest AI: a level 2 player earns 152 XP an hour against Weak AI and 149 against Normal (TUNING §40.2), so the pace of a player on his own level is about 150.

3. A bad start in Colony is not undone by a restart. A restart of Colony begins the same game from the setup and the seed: the same board, the same first player, and the AI answers the first move the same way (DESIGN §77.5). A discarded game of Kastgard is not recorded even as a loss (DESIGN §11), but Kastgard has dice. The difference between a bad choice that is restarted and a bad start that is not is reading the documents, not measurement.

4. Kastgard has more sources. Seven kinds of paid games against two, content worth 7,975 XP for one-time use, and the daily challenge. The tasks of R4 would add XP to Rokgard too, but their cap (28 × 30 = 840 XP a week, 120 a day) would need 187.5 days to reach level 30 (22,500 ÷ 120), and they are not built.

5. The sum of the levels assumes the same hours. HOME-FINAL D3 counts the sum of the levels R as taking R²/12 hours (25 R² ÷ 300). At the pace of 594 it is R²/24, at 149 about R²/6: the reputation point, which grows the helmet of the hero (DESIGN §118), costs four times as many hours in Rokgard.

Balance options (not decisions)

None of this has been decided or is being built. Each option decides what the "typical" player is, which the documents do not define: an expert with 92% wins, an AI of level 2, or a player with half the wins.

A. An own curve for Rokgard. It is data: level.curve in the manifest of Rokgard (tools/core/realms.js --check checks it); the analysis R3 (§4.2) says that a realm with its own curve gets it as data, and tests/core/economy.test.js has a toy realm with a steeper curve. If a player with the pace R in Rokgard is to reach level 30 in the same hours as a player with the pace K in Kastgard, then A = 25 × R ÷ K and level 30 starts at 900 × A:

Target (Rokgard ÷ Kastgard)ALevel 30 at XP
level 2 player (149) ÷ me in Principalities (594)6.275,644
level 2 player (149) ÷ model of Principalities on the island (316)11.7910,609
win every game against Normal (308) ÷ me (594)12.9611,667
level 2 player (149) ÷ average player after the introduction of rounds (196; step X3 is not built)19.0117,105
Level 30 in XP by option AHorizontal bars: experience needed for level 30. The current curve is the first row; each other row is a curve of Rokgard under one option A of the article table.Current curve (both realms)22,500Level 2 player ÷ me in Principalities5,644Level 2 player ÷ model on the island10,609Win against Normal ÷ me11,667Level 2 player ÷ average player17,105current curveoption AKastgard stays at 22,500 XP; the options change only the curve of Rokgard
Option A: XP for level 30. The current curve is 22,500 XP; each option is its own curve of Rokgard; the numbers are from the table

The analysis R3 (question 4) names the risk: a change of the curve after the players have XP moves their levels, the display of the levels and the achievements that read the level; before a release it is cheap, after it expensive. The game is not public yet (ROADMAP ch. 26). The hours in the proposal of the Rokgard achievements would change too (level 5 / 10 / 20 / 30 / 50 after 4.2 / 17 / 67 / 150 / 417 hours at 150 tokens an hour).

B. A share for losses. Xp.board pays a win in full, a draw half, and a loss nothing. A share f of the reward for the moves played in a loss would move the pace of a player with half the wins to 308 × (0.485 + f × 0.515): at f = 0.25 it is 189 XP an hour (119.0 hours), at f = 0.34 it is 203 (110.7). The limit 0.34 follows from the ladder: so that a loss against Strong AI pays no more than a game on one's own level, 592 f < 308 × (0.485 + 0.515 f) must hold, that is f below 0.34. It is arithmetic for games of equal length and zero wins against Strong AI, not a simulation.

C. The rate of Colony. ROADMAP ch. 27 (answer 2) kept Colony at 0.5 XP a move until it is played; PAY-BY-MOVES 4.3 speaks of the single constant ColonyData.PAY.rate, about 1.0. Doubling it doubles the XP and the tokens of Colony: a win in every game against Normal 616 XP an hour (36.5 hours), a level 2 player about 298 (75.5 hours). A wardrobe proposed at 4,650 tokens (about 15 hours at a pace of 300) would be collected in 7.5 hours (4,650 ÷ 616). Lines have their own rate (ChainsData.PAY.rate, also 0.5), so they would be changed separately.

D. Achievements and challenges of Rokgard paid in tokens. The proposal R3g4 is 2,643 tokens (57% of the proposal of the wardrobe, a band of 2,000 – 3,500) against 51,922 Golden Dice. A wider band, more families or a daily challenge of Rokgard (the manifest switches it off) would bring the tokens closer to the Golden Dice. Achievements and challenges do not pay XP, so level 30 would not move until a decision is made that a challenge pays XP as well as the coin (the decision on tasks of 30 October 2026: a task pays both XP and the coin).

E. Measure first. Between the 5 seconds a move of the model and the time with which I play, the hours of a level 2 player range from 100 to 326. A measurement needs a new save with lines m: 'colony' and m: 'chains', and a script that subtracts the gaps between the lines from history[].t (DESIGN §82.10). The campaign is not touched by this article; PAY-BY-MOVES 4.4 explains that a cut would take away what the players already got.

What could not be verified

  • Players. The chances (0.485, 0.63, 45.8%) are the chances of the AI; the only data of a human is the save of one expert. The pace of 594 is from 19 timed games (18 wins), the documents state 18 games.
  • Levels 30 and 2. HOME-FINAL D3 counts them as 30 + 2; whether these are measured levels or values from the mockup, the document does not say.
  • Stars, restart, the share for losses. The time to collect 7,975 XP was not measured; new rolls after a restart in Kastgard were not tried; the behaviour of players under a share for losses was not measured.
  • Documents and status. Step X3 (Principalities paid by rounds) is not built, so the paces of Principalities are by score.

Sources and notes

  • Curve and decisions: js/data/hero-level.js (A, P), js/realms/kastgard/data/realm.js (line 51) and js/realms/rokgard/data/realm.js (line 30) with level.curve; tools/core/realms.js (curveSame); docs/R3-REALM-ECONOMY.md §4.2 and §8; docs/DESIGN.md §118; docs/HOME-FINAL.md D3.
  • Paces and the model of time: node tools/core/xp.js --check (1 Nov 2026: one-time content 7,975 XP, that is level 17); node tools/core/pay.js --profile on the save of 23 Oct 2026 (594 and 549 XP an hour, 19 timed games); docs/TUNING.md §40 – §43; docs/PAY-BY-MOVES.md 1.2 – 1.5, 4.3, 4.4.
  • Coins and achievements: node tools/core/achievements.js --check (291 steps, 51,922 Golden Dice); the old list of 138 achievements in tests/core/achievement-chains-migration.test.js; docs/R3-REALM-ECONOMY.md §4.4 (wardrobe); docs/ROADMAP.md ch. 26 and 27.
  • Own calculations: all ratios and hours are arithmetic on the source numbers (hours = 22,500 ÷ pace); the formulas are in the text.
  • Related articles: How much a win costs when the AI played it, Every world its own currency, From one game, more games, A game without dice in a game about dice: Colony, A campaign with a budget.

What could not be verified: the pace in Rokgard, the time of a move and players outside the AI, the time for the stars, and the behaviour of players under the new rule. Only light reading tools were run, and node -e over the curve and the save.