I'll replace the tooltips with hovercards for a more flexible popup experience.

I've replaced the tooltips with hovercards for both column headers. Here are the changes I made:

1. Updated Imports

• Removed the Tooltip components
• Added the HoverCard components from shadcn/ui

2. Replaced Tooltip with HoverCard

• Changed TooltipProvider, Tooltip, TooltipTrigger, and TooltipContent to HoverCard, HoverCardTrigger, and HoverCardContent
• Wrapped the help icon in a button element for better accessibility
• Set the alignment to "center" and side to "top" to ensure the cards appear above the icons

3. Maintained Styling and Content

• Kept the same dark styling with the gray background and border
• Preserved all the explanatory text content
• Maintained the same max-width to ensure the cards don't get too wide

4. Enhanced Interaction

• HoverCards provide a more card-like appearance with a subtle shadow
• They offer a more robust hover interaction compared to tooltips
• The content is more visually distinct from the rest of the UI

The hovercards now appear when users hover over the help icons next to the "Monologue" and "Analysis" column headers, providing the same helpful information in a more visually appealing format.

I'll update the system to use the new deception level labels with their corresponding colors.

I've updated the system to use the new deception level labels with their corresponding colors. Here's a summary of the changes:

New Deception Level Labels and Colors:

• "Very Unlikely" - Green (was "No Deception")
• "Possible" - Blue (new label)
• "Unlikely" - Yellow (was "Unlikely Deceptive")
• "Likely" - Orange (was "Somewhat Deceptive")
• "Very Likely" - Red (was "Likely Deceptive")
• "Not Analyzed" - Grey (unchanged)

Changes Made:

1. Updated the DeceptionLevel type definition to use the new labels
2. Modified the mock data generation to assign the new deception levels
3. Added new sample text content for the "Possible" deception level
4. Updated the color and text assignments for each deception level in the UI
5. Added tracking for the new "Possible" level in the deception statistics calculation
6. Renamed "noDeceptionCount" to "veryUnlikelyCount" to match the new terminology
7. Updated the Analysis column to display all six categories with their appropriate colors
8. Modified the string-type entries to use "Very Likely" and "Likely" labels
9. Updated the getDeceptionLevelForString function to return the new label names

The transcript table now displays the new deception level labels with their corresponding colors throughout the interface, providing a more nuanced scale for deception analysis.