Map veto analysis: the single biggest edge in CS2 betting

Map veto is the boring part of CS2 that drives the outcome. Most bettors look at team win rates, pick the favorite, and bet. Map veto modelers know that the same two teams produce different win rates depending on the three maps that actually get played.

This article is a practical guide to reading and modeling veto order for betting purposes.

What veto is, briefly

In a BO3, each team bans a map. Then each team picks a map. Then each team bans again. The remaining map is the decider. The veto order is set by the tournament organizer but commonly looks like: Team A bans, Team B bans, Team A picks, Team B picks, Team A bans, Team B bans, remaining map is the decider.

Small changes in veto preference produce large changes in outcome probability. A team that goes 60/40 on their comfort map and 40/60 on their worst map has a 24 percent spread just from map selection.

Why the market prices veto poorly

Two reasons.

First, the betting public prices on team identity. “NAVI is better than G2, I am picking NAVI.” They do not consider which maps will get played.

Second, map-specific stats are noisier and require more work. Bookmakers price them because they employ traders. The public bets without doing that work, which keeps the market soft in veto-dependent markets.

The three maps that matter

In a BO3, the decider (third map) matters most on paper. In practice, the first pick and the decider are both high-leverage.

First pick: the pickng team’s strongest map that survived the first ban round.
Second pick: the other team’s strongest map that survived their opponent’s first ban.
Decider: whichever map both teams tolerate best among the remaining pool.

If you can identify the three likely maps before the match starts, you can compute a per-map win probability and roll it up to a series probability. That is the whole game.

Predicting veto order

Teams have patterns. Most teams ban their two worst maps first. Pick their two best. The decider falls out from what is left.

To predict:

List each team’s active duty map win rate over the last 30 maps played.
Mark each team’s two worst maps. Those are likely first bans.
Mark each team’s two best maps that the opponent has not banned. Those are likely picks.
The decider is the remaining map that both teams have 45 to 55 percent win rate on.

This is a model that fits on a napkin. It correctly predicts the veto on the first two maps about 70 percent of the time. The decider is less predictable but has less variance in its expected win rate.

Counter-picking

Some teams are known counter-pickers. Vitality, for example, often picks against the opponent’s weakest map rather than their own strongest. This changes the veto model materially.

If a team has a pattern of counter-picking, note it. Their “pick” should be chosen against opponent weakness, not team strength. Track pick selections over a season and tag teams as pickers or counter-pickers.

Map-specific data sources

HLTV map stats page. Per-map win rate over selectable time windows.
csgostats.gg for pro team breakdowns. Less complete than HLTV but sometimes has per-map round differential.
Liquipedia for historical veto orders at specific events, which helps with pattern recognition.

A worked example

Match: Vitality vs Team Spirit, BO3, IEM Dallas.

Current active duty pool

Mirage, Inferno, Nuke, Anubis, Ancient, Vertigo, Train.

Vitality, last 30 maps

Mirage: 62% win rate
Inferno: 71%
Nuke: 58%
Anubis: 44%
Ancient: 67%
Vertigo: 53%
Train: 39%

Spirit, last 30 maps

Mirage: 55%
Inferno: 48%
Nuke: 69%
Anubis: 62%
Ancient: 45%
Vertigo: 41%
Train: 71%

Predicted veto

Vitality first ban: Train (their worst).
Spirit first ban: Vertigo (their worst). Ancient is close.
Vitality pick: Inferno (Spirit weak there).
Spirit pick: Train is banned. Next best is Nuke.
Vitality second ban: Anubis (Spirit strong, Vitality weak).
Spirit second ban: Ancient (Spirit weak).
Decider: Mirage (both teams in the 55-62% range).

Series win probability

Inferno (Vitality pick): Vitality ~71% vs Spirit ~52% on this map. Vitality ~58% to win.
Nuke (Spirit pick): Vitality ~58% vs Spirit ~69%. Spirit ~54% to win.
Mirage (decider): Vitality ~62% vs Spirit ~55%. Vitality ~55% to win.

BO3 math: Vitality needs to win 2 of 3. Probability: 58% * 46% * 55% + 58% * 54% * 55% + 42% * 46% * 55% + (other scenarios) = approximately 56% series win probability for Vitality.

Market price: Vitality 1.80 (no-vig 54.3%), Spirit 2.05 (no-vig 45.7%).

Our 56% vs market 54.3% = 1.7% edge. Below our 3% threshold. No bet.

That is the whole process. It takes 10 minutes with a spreadsheet. It is also 80 percent of what our full model does.

Markets that reward veto modeling

Map handicap

“Vitality +1.5 maps”. Requires Spirit to win 2-0. Given the map-specific win rates above, a 2-0 sweep for Spirit requires winning on Nuke and the decider. Probability: 54% * 45% = 24%. So Vitality +1.5 should be priced around 76%. If the market offers 72%, take it.

Correct score

2-0 Vitality, 2-1 Vitality, 2-1 Spirit, 2-0 Spirit. Four outcomes. Most casual bettors ignore this market, which keeps it soft.

First map winner

Bet on the team you think wins the first map specifically, not the series. If you are confident about the likely first pick, this market is more accurate than moneyline.

Common mistakes

Treating all maps as equal. They are not. A team’s map pool is never uniform.
Using old data. CS2 maps have been rotating. Pre-2024 map win rates are irrelevant.
Ignoring recent form on specific maps. A team can be “good on Mirage historically” and on a 5-map losing streak on Mirage. Recent matters.
Forgetting about coaching changes. A new coach often shifts veto strategy within a month.

The bigger point

Map veto modeling is the single largest source of edge still available to a smart recreational bettor in CS2. It is not easy, exactly, but it is tractable. Most people skip it. That is why it keeps working.

If you want to invest one skill into CS2 betting, this is the one.