Hali

What to improve

1. 01

   Find the cause of punish events

   Got Punish Countered: profile 0.1 · reference 0.06

   Review five Punish Counters received and five landed. Separate whiffed attacks from unsafe actions on block before choosing what to practice.

   Try this

   Recreate one repeated situation in training. Check the spacing and frame meter, then practice a safer choice or a punish that actually connects.

   Supporting stats

   • Got Punish Countered: profile 0.1 · reference 0.06
   • Punish Counters landed: profile 0.1 · reference 0.09
2. 02

   Review defensive decisions

   Drive Reversal: profile 0 · reference 0.2

   Compare five defensive situations: what did the opponent do, what did you choose, and what happened? Low parry or throw-tech counts alone do not establish a weakness.

   Try this

   Record two different options from one troublesome situation. Practice identifying them and choosing a response, then review whether the same decision improves in your next set.

   Supporting stats

   • Drive Reversal: profile 0 · reference 0.2
   • Drive Parry: profile 1.6 · reference 1.47
3. 03

   Study a recurring matchup situation

   vs Jamie: 15 wins in 28 games

   Use the cited matchup record to choose footage to review. Look for a repeated decision across several losses, and compare it with a win before drawing a conclusion.

   Try this

   Pick one recurring situation against that character. Recreate it in training, verify an alternative response, and track that decision across the next five matches.

   Supporting stats

   • vs Jamie: 15 wins in 28 games
   • vs M. Bison: 22 wins in 31 games
   • vs Elena: 29 wins in 53 games

AI selects and prioritizes reviewed practice plans from the cited stats. Habits cover the whole profile; character records and recent samples have separate scopes. These are ideas to test, not confirmed weaknesses or video analysis. Latest reported trends across the last 100 matches, not lifetime habit averages.

Data saved Sep 15, 2026 · AI: nex-agi/nex-n2.5-pro:free

How to beat this player

1. 01

   Look for a repeatable punish opportunity

   Got Punish Countered: profile 0.1 · reference 0.06

   Use the stats as a scouting lead. Watch for an action you have already verified as punishable at that spacing; do not assume every apparent opening is unsafe.

   Try this

   Test one confirmed response when that action appears. If it stops appearing, return to your normal spacing instead of forcing the punish.

   Supporting stats

   • Got Punish Countered: profile 0.1 · reference 0.06
   • Punish Counters landed: profile 0.1 · reference 0.09
2. 02

   Probe defense before committing

   Perfect Parry: profile 0.4 · reference 0.55

   Test one familiar pressure situation and observe the actual response. Use that observation to vary your next choice; do not infer a success rate from event counts.

   Try this

   Try your usual strike/throw decision in a situation you understand. Repeat only if the response is consistent; if they adapt, vary your timing or reset to neutral.

   Supporting stats

   • Perfect Parry: profile 0.4 · reference 0.55
   • Drive Parry: profile 1.6 · reference 1.47
3. 03

   Test the Drive Impact tendency

   Drive Impact: profile 0.2 · reference 0.15

   Observe whether Drive Impact repeats in a particular situation. Prepare a response you have verified in training, while respecting range, recovery and available gauge.

   Try this

   Revisit that situation with a less committal approach. If the Impact appears, use your practiced response; if it does not, resume your usual game plan.

   Supporting stats

   • Drive Impact: profile 0.2 · reference 0.15
   • Drive Impact gauge: profile 1.2% · reference 1.3%

AI selects and prioritizes reviewed practice plans from the cited stats. Habits cover the whole profile; character records and recent samples have separate scopes. These are ideas to test, not confirmed weaknesses or video analysis. Latest reported trends across the last 100 matches, not lifetime habit averages.

Data saved Sep 15, 2026 · AI: nex-agi/nex-n2.5-pro:free