Because I am the sort of person / poster that I am, I just constructed a spreadsheet that estimates the ARG tracking results for Dean, Clark, Kerry and Lieberman every day since December 26th.
This has allowed me to produce a pretty accurate estimate of the numbers the four candidates will be starting with in today's round of ARG polling. To the best of my ability, here are the numbers:
Dean: 34
Thursday: 34
Wednesday: 34
Clark: 21
Thursday: 24
Wednesday: 18
Kerry: 10
Thursday: 10
Wednesday: 10
Lieberman: 9
Thursday: 9
Wednesday: 9
Take these numbers with a grain of salt. In addition the increasing the MoE already in the ARG poll by reducing the sample size, I could be off an additional one point in either direction for every candidate.
If Clark goes up to 22 on Saturday, that will probably mean that he replicated his 24 spot.