How to Actually Climb in SMITE 2 Ranked

The arithmetic first

Before any strategy advice, the mechanical reality: rating accumulates from net wins. A 55% win rate nets one game of progress every ten matches, which is slow but reliable. A 48% win rate nets negative progress indefinitely, and playing a hundred more games at 48% moves you further from your target rather than closer.

This sounds obvious and it contradicts how most players behave. The instinctive response to being stuck is to queue more, and that only works if the win rate is above the break-even point. If it is not, volume amplifies the problem. Our climb calculator will do the arithmetic on your own numbers.

The three habits that move a win rate

Narrow your god pool

Two or three gods per role, maximum, and ideally fewer. A narrow pool means every game is played in familiar situations, which converts your attention from remembering what your abilities do into deciding what to do with them. It also means your item and matchup knowledge compounds rather than resetting each game. The most common objection is that a narrow pool gets countered — in practice, comfort beats the counter far more often than the counter beats comfort.

Play the map, not the lane

Most ranked games below the top brackets are lost by teams that win lanes and lose the map. Winning a lane is temporary and expires when the laning phase ends; the question is what permanent thing you exchanged it for. Tower damage, jungle control, vision and objectives are permanent. A kill count is not. Players who consistently convert lane advantages into objectives climb, and players who accumulate kills without converting them do not.

Reduce deaths rather than increasing kills

A death costs your gold, your experience, your wave, and gives the enemy all three. Kills are worth less than deaths cost, which means the fastest available improvement for most players is not playing better in fights but choosing fewer bad ones. Categorising your deaths — caught out with no vision, lost a chosen fight, died contesting something worth contesting — tells you which kind of restraint you are missing.

What to stop doing immediately

• Queuing after two consecutive losses. Tilt is not a character flaw, it is a measurable decline in decision quality, and the only reliable fix is stopping.
• Playing new gods in ranked. Learn in casual queues where a losing learning curve costs nothing.
• Continuing to fight a lane you are losing. Farm under tower, concede the ground, stay level. It feels passive and it is almost always correct.
• Calling for ganks from a losing position. Ganking into a fed opponent frequently feeds a second kill and costs your jungler tempo.
• Blaming the team in chat. Beyond the obvious, it demonstrably reduces the chance of coordination for the rest of the match, which is a mechanical cost rather than a moral one.
• Checking your rating after every game. It measures outcomes over a large sample, and reading it after a single match tells you nothing except how to feel worse.

Structuring a ranked session

How you organise a session affects your results more than most people assume, because decision quality is a resource that depletes.

1. Decide the session length before you start, and stop when you reach it regardless of how it is going. Open-ended sessions end on a loss by definition.
2. Warm up in a casual game first. The first ranked game of a session is statistically the worst one most players play.
3. Pick one thing to focus on for the session — wave position, vision, death categories — and only that. A list of five is a list you will not follow.
4. Stop after two consecutive losses. Not three, not 'one more to break even'. Two.
5. After the session, spend ten minutes on review: count deaths by cause, find the decision that lost the worst game, and write one sentence about what to change.
6. Read that sentence before the next session begins.

Our match review guide covers the review step in detail, including why the moment a game was lost is almost never the moment it looked lost.

What not to worry about

A lot of climbing advice focuses on things that matter far less than the effort spent on them.

• Tier lists. A B-tier god you have two hundred games on beats an S-tier god you have never played, in essentially every case.
• Optimal build orders. The difference between a good build and an optimal one is small compared with the difference between adapting your fourth item and not adapting it.
• Your MMR versus your SR. Both move with results. Knowing the internal number would not change a single decision you make.
• Duo queue versus solo queue. Play whichever produces better decisions for you; the effect on rating is much smaller than the effect on your enjoyment, which affects decisions.
• Whether the matchmaker is against you. It is not, and the belief itself measurably degrades performance by shifting attention away from things you control.

Frequently asked questions