How Many Games per Medal Star? Climb Math for Every Rank

How the Dota 2 Star System Works

Dota 2 ranks are divided into medals, and each medal has five stars. You move from one star to the next by accumulating MMR within that medal range. When you hit the upper MMR threshold for a medal, you advance — either to the next star or to the next medal entirely.

The core mechanic is straightforward: each match adds or subtracts approximately 25 MMR at steady-state calibration. Your “uncertainty” value can push this slightly higher when you have a returning-player flag or a fresh account, but 25 MMR per game is the reliable baseline for any account with a few hundred matches in the system.

Here is what most guides skip: the MMR range per star is not consistent across all medals. Herald and Guardian stars cover smaller MMR bands than Divine stars. That means the raw game count to earn one star is actually lower at Herald than at Divine, even though Herald is “lower ranked.” Once you understand the per-star MMR gap at your current medal, the climb math becomes completely predictable.

Understanding this math is the foundation of any serious rank improvement plan. Whether you are grinding solo or considering a professional MMR boost to accelerate a specific target, knowing exactly how many games each star requires lets you plan realistically instead of grinding blind.

MMR Per Star: Every Medal Explained

The exact MMR boundaries for Dota 2 medals have remained consistent through recent seasons. The star bands for each medal are as follows:

• Herald (0 — 769 MMR): Each star covers roughly 154 MMR. Herald 1 starts at 0, Herald 5 ends around 770.
• Guardian (770 — 1539 MMR): Same band structure as Herald — approximately 154 MMR per star.
• Crusader (1540 — 2309 MMR): Approximately 154 MMR per star.
• Archon (2310 — 3079 MMR): Approximately 154 MMR per star. This is the most populated bracket in Dota 2.
• Legend (3080 — 3849 MMR): Approximately 154 MMR per star.
• Ancient (3850 — 4619 MMR): Approximately 154 MMR per star.
• Divine (4620 — 5420 MMR): Approximately 160 MMR per star for Divine 1-4. Divine 5 transitions to Immortal at around 5420 MMR.

The good news: the per-star MMR gap is nearly identical across Herald through Ancient (about 154 MMR each). Divine is slightly wider per star. This means the game count per star — assuming the same winrate — is virtually the same across the six lower medals. Where rank does matter is in the external variables: corruption, smurfs, hero pools, and draft quality shift significantly between brackets.

The Table: Games Per Star at Every Rank

The table below shows the estimated number of games needed to earn one medal star, based on three winrates: 50% (dead even), 55% (modest edge), and 60% (strong performance). Calculations use 25 MMR per win and -25 MMR per loss.

At 50% winrate, your net MMR gain per game is zero — you never climb a star regardless of how many games you play. That row exists to illustrate why “just play more games” at a break-even winrate does nothing.

The math behind the table: At 55% winrate, you win 55 games and lose 45 out of every 100. Net gain: (55 x 25) – (45 x 25) = 1,375 – 1,125 = +250 MMR per 100 games, or +2.5 MMR per game on average. Divide 154 (MMR per star) by 2.5 = 61.6 games, rounded to 62.

At 60% winrate: Net per 100 games = (60 x 25) – (40 x 25) = 1,500 – 1,000 = +500 MMR, or +5 MMR per game. Divide 154 by 5 = 30.8 games, rounded to 31.

These numbers assume you sustain that winrate perfectly across every game. Real-world performance fluctuates — which is why the actual game count is almost always higher than the theoretical minimum. For the larger-scale version of this math — how game counts compound over 1,000 MMR — see our dedicated study on how many games it takes to climb 1,000 MMR.

Per-Rank Breakdown: What the Numbers Mean in Practice

Herald and Guardian (0 — 1539 MMR)

At 55% winrate, you need 62 games per star. Herald has 5 stars before Guardian, then Guardian has 5 more before Crusader — that is 10 stars total, or about 620 games at 55% to move from Herald 1 to Crusader 1. At 60%, that drops to roughly 310 games.

The challenge in this bracket is not the math — it is sustaining a genuine winrate edge when draft quality, communication, and hero knowledge are highly inconsistent. Many Herald players report feeling stuck because their real winrate, measured over 100+ games, sits closer to 51-52% than the 55% they believe they have. At 51% winrate (net +0.5 MMR per game), one star takes 308 games — a figure that feels like a broken system but is just math doing its job.

The most powerful thing a Herald or Guardian player can do is reduce their hero pool to 2 heroes maximum and track winrate by hero across 50+ games. The difference in real winrate between a player who plays 20 heroes and one who masters 2 is typically 6-8 percentage points in this bracket — which translates directly to cutting games-per-star roughly in half.

Crusader and Archon (1540 — 3079 MMR)

The math per star is identical: 62 games at 55%, 31 at 60%. What changes is the field. Crusader and Archon players have more game sense, smurfs are more common in the 2,000-2,500 MMR range, and draft impact starts to matter more. Your hero pool consistency is a bigger variable here than in Herald.

Archon is statistically the most populated medal in Dota 2 — the largest share of ranked players sit somewhere between Archon 1 and Archon 5. Getting stuck here is extremely common, not because the math is harder, but because the gap between “I play well in my good games” and “I consistently outperform my opponents” is hardest to close in the middle of the distribution.

If you are at Archon and want to climb to Legend, you are looking at 5 stars — 310 games at 55% winrate, or 155 games at 60%. That is a useful target: set a goal of playing 50 games to reassess your actual winrate before drawing conclusions about your pace.

At this bracket, the coaching data from TeamSmurf clients shows the single highest-leverage habit: watching your own replays specifically for the 3-minute window after each death, looking at where the team is on the map and what you would do differently if you had the information your opponents had. This habit alone has moved players from 52% to 56% winrate within 30 games of applying it consistently.

Legend and Ancient (3080 — 4619 MMR)

Same 154 MMR per star, same 62 games at 55%. By this point, however, your hero pool depth, understanding of vision, and itemization decisions are the primary determinant of whether your real winrate hits 55% or stays around 50-51%. A Legend player who spams 3-4 mastered heroes typically climbs faster than an Ancient player with 40 heroes but no depth in any of them.

Ancient is also where party queue dynamics shift. Playing in stacks with coordinated teammates can provide a genuine edge on winrate here in a way that is harder to find in lower brackets. It is still a solo-queue experience for most players, but voice coordination in a 2-3 stack makes a measurable difference in objective timing and rotation efficiency.

The boost from professional MMR boosting has the biggest absolute value at Legend-Ancient: the MMR gap to Divine represents hundreds of games at 55% for a solo queue player, but a contracted booster can compress that timeline dramatically because their real winrate in that bracket runs at 65-70%+.

If you are serious about climbing from Legend to Ancient or Ancient to Divine as a solo player, the safety and efficiency trade-off of a professional service versus months of ranked grinds is worth calculating with the numbers this guide provides.

Divine (4620 — 5420 MMR)

Divine stars are marginally wider at approximately 160 MMR each. At 55% winrate: 64 games per star. At 60%: 32 games. The numbers are barely different from lower brackets, but the reality is that hitting 55% winrate in Divine is genuinely harder — the caliber of opponents and allies is sharply higher than even Ancient 5.

Most Divine players average closer to 51-53% winrate in the long run, which means their actual games-per-star count is much higher than the table suggests. At 52% winrate, net gain per game is +1.0 MMR — one star takes 160 games at that rate. That is why Divine 5 is the longest plateau many players experience before hitting Immortal.

If you are in Divine and targeting Immortal, the math says you need approximately 160 games at 55% or 64 games at 60% to cross the Divine 5 threshold (approximately 5420 MMR). However, Immortal in many regions also has a leaderboard gate — you need to be in the top N players on your server, not just hit an MMR number. That gate is separate from the star-gain math discussed here.

What Winrate Do You Actually Need to Climb?

The minimum winrate to climb at all is anything above 50%. Even 50.1% means you are gaining +0.05 MMR per game on average — mathematically climbing, practically standing still.

For a noticeable climb that feels like progress within a reasonable number of games, the practical thresholds are:

• 50-51%: You are climbing but it will take 300+ games per star. This feels like stagnation and often gets misread as a broken system.
• 52-53%: Visible progress over a season. About 125-200 games per star. Improved hero pool focus or replay review will move you here.
• 55%: The comfortable climb speed. 62 games per star. A dedicated player can clear 2-3 stars per month at this rate with consistent play habits.
• 58-60%: Fast climb. 31-45 games per star. This is what experienced role specialists typically achieve when they have mastered their heroes and bracket meta.
• 65%+: Booster territory. 5-7 MMR net per game. One star in 20-22 games. This is what professional MMR boost orders run at in any given bracket.

The best benchmark for your own winrate is your last 100 games in Dota 2 match history or your Dotabuff recent-match statistics. If your 100-game winrate is at or below 51%, no amount of extra games will deliver meaningful star gains. The game count math is fixed — only the winrate input is in your control.

Hidden Variables That Change the Game Count

The table gives you the floor — the minimum games assuming a clean 25 MMR per match and a perfectly sustained winrate. Real climbs take longer for several reasons that are worth understanding so you can account for them in your planning.

1. MMR Uncertainty and Calibration Flags

Valve uses a hidden “uncertainty” value that makes MMR swings larger when you have been inactive, switched regions, or are freshly calibrated. When uncertainty is high, wins can give 30+ MMR and losses can take 20 — or the reverse. This noise typically resolves after 20-30 games back into active play, but it makes early-sample winrates unreliable for star-count projections. If you just came back to ranked after a break, run 30 games before drawing any conclusions about your pace.

2. Abandon Penalties

Abandoning a game — whether intentional or due to a disconnect — costs MMR with no compensating win. If you average even one abandon per 30 games, you are adding a non-trivial drag to your effective winrate. A 55% winrate with one abandon per 30 games is actually closer to 53% in terms of net MMR gain per game played. The safe and consistent play patterns used by boosters include zero abandons as a strict rule.

3. Smurf and Party Matchmaking Variance

High-uncertainty accounts (smurfs, returning accounts) are matched into your bracket and have dramatically different MMR relative to their actual skill level. Playing against a 4,000 MMR player calibrated at 2,000 does not change your 25 MMR loss — but it does artificially depress your winrate in games that were never fair. You cannot fully control this variable, but you can reduce its impact by playing during peak hours when the matchmaking pool is larger and smurf encounters are statistically rarer.

4. Tilt and Session Length

Analysis of boost order data at TeamSmurf consistently shows that winrate in the last 30 minutes of a long session (5+ hours) is measurably lower than in the first two hours. Fatigue affects micro decisions and draft reads before players realize it. For a solo-queue player targeting a specific star count, 2-3 game sessions with a firm stop after a loss streak are more efficient than marathon grinds — even if the total game count in a marathon session is higher.

5. Hero Pool Depth vs. Breadth

Players who concentrate on 2-3 heroes per role consistently outperform their theoretical winrate at any given MMR compared to players who play 15+ heroes. Role specialists at Archon with 200+ games on two heroes regularly hit 57-60% real winrate. Generalists at the same bracket rarely exceed 53%. This is the single highest-leverage variable you control directly — and it is not captured anywhere in the 62-games-per-star table. The table assumes your winrate is fixed; this variable is how you actually move that input number.

How to Speed Up Your Star Climb

Everything below is about moving your winrate input number upward. The math table is fixed. Your winrate is not.

Pool Down to Your 2 Best Heroes Per Role

Pull up Dotabuff or your match history and sort by winrate across heroes you have played 30+ games on. Find your top 2. Play only those heroes for 30 games and record the winrate. If it is above 57%, stay pooled. If it drops below 53%, reassess one hero and introduce a third option. This single habit is responsible for more star gains per month than any other mechanical improvement in TeamSmurf coaching sessions — including movement, warding, and farm routing.

Identify Your Current Real Winrate

Open Dotabuff or your in-game match history. Filter to the last 100 games and note your winrate. If it is below 53%, the star-count math is working correctly — you are just at a rate where progress is very slow. Address the underlying skill gap first: either through deliberate practice, replay review, or a coaching session with a high-MMR player who can identify specifically what is capping your winrate in your hero pool and role.

Use the Game-Count Table as a Time Planner

If your real winrate is 55% and you want to gain two stars (approximately 308 MMR), you need roughly 124 games. At 2 games per day, that is 62 days — about two months of consistent play. Knowing that in advance helps you evaluate honestly whether you want to grind solo for two months or whether a targeted MMR boost order covering that 308 MMR is a better use of your time and energy. TeamSmurf typically completes a 300 MMR boost in 2-4 days — the same result in less than 3% of the time.

Track Per-Session Stats

After every session of 3-5 games, note the net MMR and the winrate. Over four weeks, you will have a reliable sample of your real winrate across different days and times. Use that data to identify your best conditions — time of day, session length, hero matchup performance — and optimize your schedule around them. Most players discover they perform 3-5% better in their first 2-hour session of the day than in late-night extended sessions.

Set Explicit Star Goals and Track Progress

Rather than checking MMR after every game, set a target (for example, next 2 stars = +308 MMR) and track progress weekly. This perspective shift reduces tilt from single-session variance and keeps the long-term math visible. A player who loses 3 games in a row at 55% winrate has not broken their climb — they have just played out the expected variance within a 62-game sample. Knowing the full count keeps individual losses in proportion.

Frequently Asked Questions