Add position_est, an estimated 1-5 position for parsed matches - #2974
Merged
Conversation
Within each team, players are ranked by early farm priority (gold_t and lh_t averaged over minutes 10-12, early ward purchases breaking ties): top 3 cores, bottom 2 supports. Cores map to 1/2/3 through lane_role; safe lane supports read as 5 and the rest as 4, with farm order resolving conflicts. Computed at query time in buildMatch, only when the whole team has the parsed fields. Validated against reference position labels on 100 pro matches (930 players): 98.9% exact, 99.8% core/support. Details in odota#1590.
howardchung
approved these changes
Aug 9, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Implements #1590 with the design discussed there: computed at query time (in
buildMatchviacompute.ts, next to the existinglane_role/is_roaming), no storage changes. The field only appears when the whole team has the parsed data it needs (gold_t/lh_t/lane_role), so unparsed matches just don't get it.The heuristic is the one validated in the issue — 98.9% exact / 99.8% core-support against reference labels over 100 pro matches, holdout-checked: per team, rank by early farm priority (
gold_t/lh_taveraged over min 10-12, early ward purchases as tiebreak) -> top 3 cores / bottom 2 supports; cores map to 1/2/3 vialane_role; supports: safe lane -> 5, mid/off -> 4; leftovers by farm order. This TS implementation produces identical output to the study script on all 1000 players of the sample.Also effectively covers #1844, since core/support is just 1-3 vs 4-5.