October 14, 2004 - Jill

1. Partisanships were off by a couple percentage points in almost every case. Data was added into columns CV through DI to determine partisanship. Columns

L and M were then made identical to column DH, with the new partisanships using data from the Almanac of American Politics 2004. This was NOT the

case with Texas. The altered district partisanships used to take into account the 2002 redistricting remain as they were.

September 16, 2004 - Jill

1. Updated with all primaries through September 14.

2. R. Alexander changed parties (Dem. to Rep). Handled this by treating him as if he were elected as a Republican in 2002.

3. Ralph Hall changed parties. Because the spreadsheet wasn't set up to handle party changing, we handled this by treating him as if he was newly elected in

2002 as a Republican rather than as a longtime Democratic incumbent.

4. In columns BE and BF, All formulae going to column L were changed to go to column M, to take partisanships from 2004 rather than 2002 and thus adjust

for Texas' redistricting for open seats.

August 31, 2004 - Jill

1. Updated with Aug. 24 primaries.

2. Added open seat for VA-2.

3. Uploaded all to web.

June 28th - stephanie

1. Updated all of the uncontested seats as of June 28th

2. Uploaded this version as the zip file to web.

3. Added information on special elections and SD-1.

4. Fixed the definitions of landslide, competitive and tight on the easy viewing sheet.

May 25 - ck

1. Problem with conditional formatting on Easy Viewing: deleted headers, re-copied formulas and conditional formats, re-inserted headers as rows.

2. Modifed formulas to the side of Easy Viewing to avoid coloring "New" p-ships.

May 17 - ck

1. Formula in easy viewing for 2000 result set to be blank if no data reported in the projections sheet.

2. Deleted cells in Open seat column that had spaces but no other entries in them.

3. Minor edits to Special Elections and notes. Deleted note about color shading showing incumbents whose district #s changed.

May 14th

1. Added open seat for Jack Quinn

2. Revised Algorithm document and User's Guide to include 5th projection model (twice-uncontested) and 2004 terminology, adjustments of performance based on p-ships, etc

April 29 - changes

1. Re-copied formulas for Codes in Easy Viewing (so it's blank if no code present, instead of zero, and made formulas deal with headers)

2. Re-did formulas for projections totals (totals are from Projections column P, Rep totals are from column S, and Dem/Ind totals are the difference) See Yellow box on Easy Viewing.

April 27- changes

1. Clean up blank cells in code column that had spaces or other invisible characters.

2. Add code of "2" for twice uncontested.

3. Modify 3 columns for twice-uncontested model

Democratic Projection: increased range of index function to include 5 projection (number 4 for twice-uncontested)

Model selection: included option 4

Inserted column for including twice-uncontested projection

Checks if code column contains the character "2"

Projection for twice uncontested is pship-3% for Reps, pship+3% for Dems

4. Delete Monopoly Politics sheet

5. Inserted column headers into the middle of "Easy Viewing" so they show up on the web page as you scroll down.

6. Adjust formula so only show codes if they exist, otherwise, leave cell blank

April 23 - changes

1. Caleb preparing to input data on 2 uncontested races and create separate projection model for that group.

2. Minor cosmetic changes to Easy Viewing sheet for webposting

March 17 - Apr 23

1. Stephanie working on Easy Viewing sheet (no other changes documented)

March 17 - changes to incorporate TX re-redistricting.

These changes only affect TX, since that's the only state where the 2004 p-ship is different than the 2002 p-ship.

1. Started with mp2004_march_3.xls (downloaded from website approx Mar 3)

2. Copy formulas from row 1 all the way to make sure totals don’t change (this ensures that range names are used consistently)

3. Redefine pship to be 2004 pship

4. Define pship1 and pship2 to be the pship for 2000 and 2002.

5. Define adj1 and adj2 in columns CF and CG to adjustment to 2000 and 2002 performances based on changes in p-ship.

6. Use adj1 and adj2 to adjust past performances in columns AR and AS

7. I added notes about TX, including the party shift (Ralph Hall, TX-4).

To do:

1. Check my work, and confirm that the algorithm is working correctly.

2. Enter data about uncontested elections (U for uncontested by major party, UU is completely uncontested).

3. Track down data about uncontested primaries (for Dubious Democracy)

March 01 '04-changes

1. Corrected the districts whos 2002 partisanship was based on 2000 partisanship, they were previously denoted by an *.

Feb 25th '04-changes

1. The open seats that are confirmed as of the 24th have been fixed.

2. The special elections that have already taken place have been updated. Ones that are upcomming have been noted in the speacial elections notes on the bottom.

3. An adustment of the 2002 Texas partisanship was made to reflect the recent redistricting.

These partisanship changes were based on 2 party percentages and might vary from the national percentages.

4. The correct winning 2000 percentages have been installed from CA-40 through WY-1.

Dec 11 - change

1. Winning party formula had switched "No proj" with "Most vulnerable". Corrected

November 14 - significant reformating, usability changes

1. For winner who defeated other incumbents in the general election, I listed 2000 data for the winner (% and p-ship). This affected CT-5, IL-19 and PA-17.

2. Order of variables and descriptions changed.

3. 2004 results and accuracy columns moved to left.

Nov 6 - changes

1. Preliminary winning party and % entered in mp2002 showing accuracy of projections for 2002 and using that data for 2004 projections.

October 30 - changes

1. The adjustment to worst performance was modified to only affect 2000 performance.

2. Patsy Mink's seat is listed an open seat elected in 2003.

Oct 25 - changes

1. This works off the Oct 18 version and splices in NCEC data for 2002 districts for AL, NE and NH.

2. This changes the codes by eliminating * in those seats.

NOTES - Monopoly Politics 2002 Projections

1) Please see the "Users Guide - Monopoly Politics," "Raw data.xls" and "Algorithm.doc" for additional description of how to use the spreadsheet,

how the model works and the source of data. These files are available at www.fairvote.org/2002

2) The spreadsheet allows the user to modify parameters in cells shaded blue to examine their effect on projections.

3) If you enter in cells N15 to O449 the winning party and percentage after the election, the spreadsheet will calculate the accuracy of projections.

if you copy the formulas in cells V15 and W15 into the cells V15 to W449.

4) In cells highlighted yellow, the spreadsheet keeps track of the number of projections by category, their accuracy, the number of vulnerable

candidates who lost, and so on. The number of projections by category for seats held by Dems and Republicans is in cells AA1 to AE8.

5) Cells AH1 to AL7 shows the number of races affected by each parameter (overachievers, underachievers, better 2nd election, uncontested).

The Center for Voting and Democracy

6930 Carroll Ave Suite 610

Takoma Park MD 20903

301-270-4616, 301-270-4133 (fax)

[email protected]