Even as everyone repeatedly talks about how tomorrow will be "about delegates," the pre-game spin is still about winning states. In fact, virtually none of the spin is about how the margins in different states affect the delegate counts. This is markedly different from the analysis before a Presidential election, when the discussion is all about the "electoral math."
To try and remedy that, I have put together some numbers that give us a sense of what the playing field looks like tomorrow, in terms of the only numbers that actual count: the delegate count.
To spoil the surprise, we're dealing with a best case scenario for Clinton of: Clinton: 779, Obama: 584. Obama's best case scenario is Clinton: 604, Obama: 759. The middle of those numbers has Clinton slightly ahead, 691 to 671.
Methodology and the full numbers on the flip.
Methodology
For each state that's voting tomorrow, I took the average of the polls taken since John Edwards dropped out. If there were two or more polls, I used an MoE of 3%, otherwise I used the actual MoE of the poll in question. Where there were no polls since Edwards dropped out but a single relatively recent poll, I used that one. Otherwise, I assumed a +- 100% for each candidate.
I also assumed that delegates will be apportioned precisely based upon the candidates proportion of the vote in a state, and did not attempt to reapportion fractional delegates where both candidates rounded down. These assumptions are all wrong, of course, but they should be close enough to get meaningful results.
I only used polls included in pollster.com's averages, and got the delegate counts from the DNC's website.
The numbers
Alabama
44-42 Clinton
Best case Clinton: 22-17 Clinton
Best case Obama: 21-18 Obama
Arizona
44-40 Clinton
Best case Clinton: 24-18 Clinton
Best case Obama: 22-20 Obama
Arkansas
Unknown
30 Delegates
California
44-40 Clinton
Best case Clinton: 151-119 Clinton
Best case Obama: 138-132 Obama
Colorado
34-32 Obama
Best case Clinton: 17-15 Clinton
Best case Obama: 18-14 Obama
Connecticut
46-44 Clinton
Best case Clinton: 21-17 Clinton
Best case Obama: 20-18 Obama
Delaware
44-42 Clinton
Best case Clinton: 6-5 Clinton
Best case Obama: 6-5 Obama
Georgia
50-35 Obama
Best case Clinton: 36-29 Obama
Best case Obama: 40-24 Obama
Idaho
Unknown
16 Delegates
Illinois
58-30 Obama
Best case Clinton: 102-61 Obama
Best case Obama: 113-50 Obama
Kansas
Unknown
28 Delegates
Massachusetts
52-39 Clinton
Best case Clinton: 45-29 Clinton
Best case Obama: 40-34 Clinton
Missouri
47-42 Obama
Best case Clinton: 29-27 Clinton
Best case Obama: 32-24 Obama
Minnesota
40-33 Clinton
Best case Clinton: 28-18 Clinton
Best case Obama: 24-22 Obama
New Jersey
48-40 Clinton
Best case Clinton: 47-34 Clinton
Best case Obama: 42-40 Clinton
New Mexico
48-42 Obama
Best case Clinton: 11-9 Clinton
Best case Obama: 13-8 Obama
New York
53-36 Clinton
Best case Clinton: 114-65 Clinton
Best case Obama: 100-80 Clinton
Oklahoma
54-27 Clinton
Best case Clinton: 19-8 Clinton
Best case Obama: 17-10 Clinton
Tennessee
51-32 Clinton
Best case Clinton: 32-17 Clinton
Best case Obama: 28-21 Clinton
Utah
53-29 Obama
Best case Clinton: 10-7 Obama
Best case Obama: 12-5 Obama
Summary
Basically, Clinton's expected big win in NY is blunted by Obama's expected big win in IL. Most of the other states are either toss-ups or have equivalent states by their opponent. In the end, Clinton has a slight advantage, but it's an extremely close race that's going to come down to the delegates.
Yes, I know you've heard it before, but I was sick of hearing the contest described in terms of who wins what state ("if Clinton loses CA, that would be huge!!!!!!"), because all the analysis along those lines is straight up wrong or an advance strike-team toward some spin tomorrow night.
Note that the "best case" and "worst case" numbers are based on the MoE in either direction, and we can expect that some of these polls are dead wrong (and at least one or two is wrong simply because of the confidence interval), but the overall picture should be about right.