Impeachments
Would you like to react to this message? Create an account in a few clicks or log in to continue.

Go down
avatar
Posts : 5
Reputation : 6
Join date : 2019-11-20
View user profile

Regression of TPR Differential and PD Empty Regression of TPR Differential and PD

Wed Oct 28, 2020 12:34 pm
One of the most commonly utilized resources in the stacking process is individual scores. It's one of the only data sources that Mock Trial is able to provide. These scores can then be normalized over the average judge score for a round to adjust for judges that are considered too harsh (3s and 4s only) and those that are considered too kind (9s and 10s only, "I thought you were law students"). However, this method makes a key assumption that all competitor teams are equal in strength. Obviously, this is not the case.

The question then becomes how to adjust scores for the difference in the strength of opposing teams. The easiest way to determine a team's potential strength is through its TPR. Typically, TPR provides a fairly accurate depiction of the likelihood of a team to succeed at ORCS or Nationals in a given year. However, these tournaments are after stacking and the scores they provide would not be of use to a team that wants to use scores for their stacking process. The question then posed is if there is any correlation between a team's unstacked performance in a round at an Invitational and the TPR of the team. In order to test this, I analyzed the data from the Fordham Invitational and calculated the TPR score differential for each round's pairings and compared that data to the PD for each round's pairings. This is admittedly a small quantity of data; however, I'm in the middle of midterms and didn't want to spend time pulling more. Performing a regression analysis on the two lets me know that I wasted my time. There is no meaningful correlation between differential team TPR and the PD of the round.

If anyone has any other ideas on how to potentially adjust for competitor strength or if anyone wants to pull more data and prove me wrong, I'd appreciate any input.

TLDR: TPR gives no indication for the PD of the round
Back to top
Permissions in this forum:
You cannot reply to topics in this forum