DOTA 2 · Lineups & drafts
Both lineups are in. What should you check first?
Compare Dota 2 roles, three lanes, counters and synergy in Bladeform. Learn to check lane assignments, samples and player records, and distinguish lineup scores from probability estimates.
Which of two Dota 2 lineups is stronger?
First define whether stronger means better lanes, a more cohesive draft, or a better chance of winning the map. Check the patch, all ten heroes, their actual roles and the players. Bladeform helps compare lanes, counters and synergy in a known lineup; its draft simulator helps compare the next choice during BP. State what supports the lean, the conditions it depends on, and what would change it.
Prediction-market research: what to check at each stage
| Stage | Question to check | Using Bladeform |
|---|---|---|
| Before the draft | Are the roster, stand-ins, format and target map confirmed? | Check available team and player hero-pool records. Keep unpicked heroes unknown. |
| During the draft | Which picks, bans, pick order and open roles are confirmed? | Enter the actual sequence and compare candidates and opposing replies under the same assumptions. |
| After the draft | Where will flex heroes play? Which lane advantages have sample support? | Check roles and lane evidence. Separate composite scores from explicitly labelled probability estimates. |
| After the match | What was recorded at the time, and what became known only later? | Retain the original inputs for comparison with the result. Label retrospective simulations as reviews. |
How can a Polymarket Dota 2 user use it?
Use Bladeform for the lineup and draft-research part of a match assessment: organize confirmed heroes and candidate replies. It currently has no live Polymarket price integration and is not an official Polymarket product. Check map versus series scope, deadlines and resolution terms on the original market. A single-map draft assessment is not a series win probability.
Does a model lean prove the market is wrong?
No. A composite draft score is not a market probability, so subtracting a market price from it does not measure an edge. Even a separately labelled probability estimate requires calibration on comparable events, information available at prediction time and a complete historical record. This guide does not establish profitability.
What makes an assessment reviewable?
Record the match and map, analysis time, patch, heroes and roles, data snapshot, supporting evidence, opposing evidence and unknowns. State a conditional lean and the roster or lane changes that would require a new assessment, rather than recording only a predicted winner.
Read the worked example: Yandex–Spirit map 2, full draft, five candidate comparisons and data record
When is it useful?
Use Lineup overview when both teams have heroes selected and you want to analyze a matchup, review a game or compare compositions. Use Draft simulator when picks and bans are still in progress and you need to evaluate the next move.
Check the input before the conclusion
Enter five picked heroes per side and confirm Radiant and Dire. Do not include banned heroes. Check actual positions 1–5, especially for flexible heroes: changing a role also changes the meaning of lane comparisons.
Read each lane separately
Review mid and both side lanes separately, then check whether the evidence is from professional or high-rank games, which patch it covers and its sample size. Insufficient lane evidence should remain unknown rather than be interpreted as an even matchup.
Place counters and synergy in the full lineup
Historical hero relationships provide partial evidence. The full lineup also involves roles, lanes and player records. One counter relationship alone does not determine the outcome of a match.
Worked comparison: what if a role changes?
Suppose the same hero plays a different role in the target match. Adjust the role first, then compare lane and lineup results. This illustrates the review process, not the strength of any hero in the current patch.
Record a conclusion others can review
Record both lineups, actual roles, the data scope and samples supporting your conclusion. Separate estimates shown by the tool from your own judgement; do not rewrite an overall score as a win rate.