Talk:2008 Democratic Party presidential primaries/Archive 3

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

Sources for candidate results

There has been discussion here and there on this article and also on Results of the 2008 Democratic Party presidential primaries about which sources are most appropriate in these articles for pledged and unpledged (super) delegate counts. Currently, the agreement is to use CNN for pledged delegate totals and the 2008 Democratic Convention Watch Blog for superdelegate totals. However, this agreement was somewhat informal and there now seems to be some significant interest in changing the sources. Rather than continue wrestling with this issue in a piecemeal fashion scattered across multiple places on multiple discussion pages, I propose we focus our discussion under this heading on this page. I'll post a comment on the Results article's discussion page attempting to direct discussion here.

The essential question I think we're trying to answer is:
Which sources should we use in Wikipedia for the numerical results of the 2008 Democratic Party presidential primaries?

I think our answers to this question should address such aspects as:

• using multiple sources vs. a single source
• which sources are available?
• accuracy and reliability vs. currency
• maintainability by Wikipedia's editors
• conformance with Wikipedia's policies and guidelines
• usefulness to Wikipedia's users
• other aspects I'm missing?

So, have at it! Let's hear your answer to the above question and the reasons which support it.

Bryan H Bell (talk) 20:27, 8 February 2008 (UTC)

In the CNN vs DemConWatch again section above, Northwesterner1 suggests using AP for the pledged delegate totals and continue using DemConWatch for the superdelegate totals. AP seems like a good choice since most of the other news-reporting orgs use it as their source. My concern with using the AP totals is where we'll obtain these numbers. Correct me if I'm wrong, but I don't think AP publishes these numbers directly. This means we'd have to use a secondary source that reports them and that begs the question of which one we should use. Bryan H Bell (talk) 21:12, 8 February 2008 (UTC)

Google News is the only outlet that publishes AP content in full and without advertisements, such as with this example. The AP also counts superdelegates. I think this is a sensible choice. 66.167.141.207 (talk) 18:32, 9 February 2008 (UTC)

For the February 5 voting results for withdrawn or low percentage Democratic candidates in Minnesota that I'm reporting on Minnesota Democratic caucuses, 2008#Results, I've found it necessary to go to the Minnesota Secretary of State's website, where the numbers are still being updated with ~90% of districts reporting. (Delegate counts aren't reported there, however.) I've copied those voting numbers into the major candidates table in Results of the 2008 Democratic Party presidential primaries#Minnesota caucuses, rather than use CNN numbers. This seems good practice to me, since I'm going as near as possible to the original source; I didn't find voting numbers on the Minnesota DFL party website. There's currently variation in this practice among editors for the states; Missouri does as I do, while Connecticut uses CNN and doesn't report minor candidates (if there were any). Wdfarmer (talk) 21:40, 8 February 2008 (UTC)

DemConWatch has updated their Super Tuesday count to say that they do not intend to keep it current. They are using AP numbers now and they link to this AP count at the Chicago Tribune. The AP totals at the Tribune include superdelegates. They're pretty easy to read all in one place, but you have to subtract the superdelegates out of each state total for each candidate. Do we have any AP source that cleanly reports pledged delegates only? It seems in any case that DemConWatch is no longer viable as a source for state numbers. Northwesterner1 (talk) 22:03, 8 February 2008 (UTC)

Here you go, AP numbers from CSPAN. And this site did a good job of having updated numbers on Super Tuesday evening. Simon12 (talk) —Preceding comment was added at 05:07, 9 February 2008 (UTC)

Thanks, Simon12. I see that C-SPAN has a nice list of political news resources at their site too. Wdfarmer (talk) 07:21, 9 February 2008 (UTC)

This is a good site, clean and easy to read in one place. One thing against it, though, is that they don't show how many delegates are yet to be awarded in each state. For example, they've got Alabama at 21-20, with no indication that more delegates are on the way, even though there are 52 pledged delegates up for grabs. I think there is something to be said for keeping the "still counting" column in our tables. (Of course, we could just do the subtraction ourselves if we decide to go with this source.) They also seem to be calling delegates more conservatively than the DemConWatch sources, which have it at 25-24 in Alabama. That's okay, I guess, as long as we keep the "still counting" column. Otherwise, I think these numbers are misleading, as they under-report the total delegates at stake in the Feb 5 states. Another consideration is that this page includes only Super Tuesday states, so we would need to find other pages on the C-Span site for the other states. That said, i think it is a good source. Northwesterner1 (talk) 09:19, 9 February 2008 (UTC)

I understand it also may be difficult to update numbers here and at the results article and at the individual state articles, and to have those numbers be consistent across the various articles, as well as internally consistent within the article. I added columns with state-by-state results to the January and February 5 tables because I thought it was strange to have the tables for states that have already voted without the results, and initially I did not see the link to the separate results article. One possible solution is that we could take those columns out and just have a column added to the January and February tables under the header "results," with a wikilink in each cell called "results" that points to the appropriate state results article. That allows for a state-by-state solution as Wdfarmer is using above. The final cell on the January and Feb 5 tables, "total," could have a "results" wikilink that points to the main results article. We could continue to report total pledged delegates and superdelegates in the summary table, using numbers from a source consistent with the main results article. Northwesterner1 (talk) 22:23, 8 February 2008 (UTC)

Yes, I'd like to see the delegate and superdelegate results removed from Democratic Party (United States) presidential primaries, 2008. I believe that article is really about the primary process, and that results belong in Results of the 2008 Democratic Party presidential primaries. The state sections within the Results article, and the state articles linked from those sections, should just have number of the pledged delegates resulting from the popular votes in that primary or caucus. The question then is whether those pledged delegate numbers should appear in the state rows of the Results of the 2008 Democratic Party presidential primaries#Overview of results table, or whether the (pledged delegate + superdelegate) sum should appear there. Wdfarmer (talk) 00:41, 9 February 2008 (UTC)

It would be easier for Wikipedia's editors if the results were listed in a single place and linked to from everywhere else. However, it might be more useful to Wikipedia's users if high-level results were immediately visible without having to click around to other articles. Bryan H Bell (talk) 02:19, 9 February 2008 (UTC)

That's a good point. Northwesterner1 (talk) 09:20, 9 February 2008 (UTC)

I can think of one method that would accomplish both; perhaps it has been applied elsewhere in WikiMedia. A template could be defined that, like a database, would contain the pledged delegate numbers and superdelegate numbers, within a party, for all of the states and candidates; e.g. [[template:dem2008]] and [[template:rep2008]]. Articles for a party would then invoke their party's template multiple times, each with the combination of parameters needed to select the one desired delegate number for invocation; e.g. {{dem2008|pledged|MN|Clinton}} would return the text string "24". Advantage: all the numbers are maintained and defined in a central place, but can be accessed and displayed anywhere. Disadvantages: the database templates might require expertise to maintain (maybe that's a good thing); template invocations would require expertise to maintain; articles would be larger; unscrupulous editors could ignore the templates and post their own numbers instead. Wdfarmer (talk) 10:24, 9 February 2008 (UTC)

I propose to switch to CBSNews data: http://election.cbsnews.com/campaign2008/d_delegateScorecard.shtml that are constantly updated and accurate, much better than CNN data (very obsolete), at least as a temporary solution. (only for the pledged delegates). Do you agree? --Subver (talk) 19:07, 9 February 2008 (UTC)

DemConWatch has switched to using AP's numbers (via the Chicago Tribune). Andareed (talk) 22:55, 9 February 2008 (UTC)

It's surely better than CNN's (they show only an old sum...the sigle numbers includes also superdelegates...very confusing) but I think the best and more accurate/updated are CBS News. http://election.cbsnews.com/campaign2008/d_delegateScorecard.shtml even now they are updating... Does someone agree to use CBS News site numbers? --Subver (talk) 23:11, 9 February 2008 (UTC)

You should just go ahead and make the change. It's obvious that there's consensus that the CNN numbers are bad, but I doubt we'll have consensus on which alternate source to use. Andareed (talk) 23:21, 9 February 2008 (UTC)

Done. I hope everybody agree. --Subver (talk) 23:47, 9 February 2008 (UTC)

I have no opinion about which source should be used, but it's confusing when now the total numbers are from CBS (pledged), but state-specific data is from DCW, so the total amounts don't match.--217.140.231.213 (talk) 00:10, 10 February 2008 (UTC)

Pledged from CBS is simply taken by the total minus the "CBS superdelegates". Moreover the previous total was different even with CNN data... —Preceding unsigned comment added by Subver (talk • contribs) 00:26, 10 February 2008 (UTC)

Can we use CBS for the state results to be consistent? Andareed (talk) 01:00, 10 February 2008 (UTC)

I think it's better to add single state delegates only when the results are definitive (all pledged delegates are assigned). For the list already written, if you want, you can change the numbers but it is not essential IMHO. --Subver (talk) 01:38, 10 February 2008 (UTC)

Check out this article in the New York Times: Media and Candidate Methods of Counting Delegates Vary and So Do Totals. The New York Times provides a much more conservative delegate count than other sources since they don't make any projected estimates. I wonder if perhaps that might make it a more "encyclopedic" source (i.e. taking the longer view and sticking closer to verifiable facts). The problem is that the count will always be less "current" than other sources. What do you think? -- Bryan H Bell (talk) 05:35, 10 February 2008 (UTC)

Also, the NY Times methodology matches more closely with our reasons for using DemConWatch for the superdelegates. -- Bryan H Bell (talk) 05:38, 10 February 2008 (UTC)

I think you are totally right saying that results are not definitive, but NYT numbers are totally obsoleted and they doesn't reflect real situation. I'm changing the "count" in "estimate" --Subver (talk) 11:47, 10 February 2008 (UTC)

I don't think I'd go so far as to say that the NYT numbers are obsolete or don't reflect the real situation. The NYT numbers reflect the actual committed delegates, rather than estimates (which, as we've seen in many examples lately, can be just plain wrong). However, I think I get the general gist of your thoughts. In this day of 24-hour news cycles and up-to-the-second Internet updates, the slower progress of the Democratic Party process (particularly the caucus process) just can't match the public's demand and expectations for immediate results. --Bryan H Bell (talk) 22:50, 10 February 2008 (UTC)

Looks like we still haven't reached consensus. In the meantime, we might as well try to keep the numbers as accurate and consistent as we can. I updated the calendar tables using numbers from the main results article. Northwesterner1 (talk) 09:43, 10 February 2008 (UTC)

I corrected the single state delegates estimates according CBS News. Now the total candidate delegates perfectly matches the sum of single state delegates estimates. --Subver (talk) 20:33, 10 February 2008 (UTC)

New York Times vs. CNN

The New York Times numbers are a lot different from the CNN numbers. I noticed this because the New York Times numbers are used on the Barack Obama presidential campaign page, while the CNN numbers are used on this page. Which one is more correct? —Preceding unsigned comment added by 24.186.232.27 (talk) 06:52, 9 February 2008 (UTC)

The New York Times and CNN use different methods to generate their delegate counts. The New York Times takes a fairly conservative approach by not making any estimates about the number of delegates until after they are officially chosen. For example, my home state of Washington just held it's initial caucuses today, but won't make a final decision until after a series of further caucuses culminating in its state convention in June. The New York Times will wait until June to report Washington's delegates, but CNN looks at the initial caucus results and projects what they think will happen in June. There are pros and cons to each approach. You could say that the New York Times numbers are more "correct", however CNN's numbers tell us what we really want to know right now: "how are the candidates doing so far?" The question of which source is the best to use for this article is a matter of current debate. See the discussion elsewhere on this talk page: Sources for candidate results. --Bryan H Bell (talk) 08:37, 10 February 2008 (UTC)

Also, I've noticed that the CNN numbers have superdelegates included in them. This inflates the totals and can cause some confusion if you don't know this (and proceed to add the superdelegates in again) Padillah (talk) 14:45, 11 February 2008 (UTC)

Not the only results page -- merge?

I just saw this Results of the 2008 Democratic Party presidential primaries in the Recent Changes log. Looks as though we have two articles showing primary results. -FeralDruid (talk) 23:19, 10 February 2008 (UTC)

Okay, I see the other article linked at the top of this one, but am not sure why there are two articles covering this information. -FeralDruid (talk) 23:28, 10 February 2008 (UTC)

There's been some discussion about this. This is primarily an article about the process behind the nomination, how delegates are selected, etc. The results article has deeper results (state-by-state vote counts, state percentages, etc.) So they cover different terrain, and I don't think they should be merged. The question then is whether this article should include any results. Some editors feel that it shouldn't. Some feel that certain high-level summary results are useful here to readers. I agree with the latter, but I see strong arguments for the other side also. Northwesterner1 (talk) 23:52, 10 February 2008 (UTC)

I have wrestled with this question myself and have sometimes thought that the articles should be merged. I notice that past presidential primary articles have the data from both these articles combined into a single article. However, those articles also include far less detail than the ones covering the 2008 primaries. As Wikipedia (and the Internet) grows older, the compexity of its newer articles will likely increase since recent data is more easily available on the Internet than past data. I'd hate to lose the increased detail. Therefore, at this point I agree with Northwesterner1 that we probably shouldn't merge these articles. --Bryan H Bell (talk) 00:17, 11 February 2008 (UTC)

I agree. Don't merge. —Preceding unsigned comment added by Subver (talk • contribs) 00:32, 11 February 2008 (UTC)

Using templates for calculated table cells

In the 'Percentage of total delegate estimate' column, I just substituted the text values with calculation templates. My intention is to make updating the calculated values in this table easier to maintain. What do you think? If I knew how to do it, I'd like to replace all the calculated values with templates so all we have to do is update the non-calculated entries, such as 'Pledged delegate estimate' and 'Superdelegate estimate'. I don't know how to substitute values from other table cells into a calculation. Anyone know how to use create/use templates that could help out? --Bryan H Bell (talk) 00:11, 11 February 2008 (UTC)

I don't know how to do that, but it would certinately elimate future math errors. (In fact there are many pages that could use such formulas; I think on the NFL 2007 season page I had to fix over 100 winning percentages over the course of the season from editors changing the Win Loss and not adjusting the percentage.) Jon (talk) 00:24, 11 February 2008 (UTC)

I think it's impossibile. And I think the current "formula" is useless. --Subver (talk) 00:40, 11 February 2008 (UTC)

Impossible? Useless? These pronouncements seem a bit extreme and harsh. Subver, perhaps you have a more complete understanding of Wikipedia templates than I do, but I'd be surprised to find out there was no possible way at all to use templates in Wikipedia for calculating table values. I can appreciate that you might find using the formulas difficult, but I would have appreciated you leaving them in place until a few more editors had weighed in on whether or not they were helpful. You've put in some excellent work on this article, but you aren't the only editor working on it. --Bryan H Bell (talk) 07:03, 11 February 2008 (UTC)

Sorry a misunderstanding...I think it was useless use a formula like 1100/(26+1100+1120) codified directly with numbers and not references. And I meant "impossible" without changing the structure of the table...I'm not an expert at all! I'm very sorry. --Subver (talk) 15:52, 11 February 2008 (UTC)

I did some experimenting. Although I don't think calculations in one table cell can reference other table cells, it is possible to build a template to do the computations for an entire table row at a time. Here's an example that takes two vote counts, vote1 and vote2, sums them, and gives the percentage of each vote count relative to the sum:

Template to generate a one-row table

Invocation of "calculate_row"

{{calculate_row
| name1=Bill
| vote1=500
| name2=Bob
| vote2=800}}

Definition of "calculate_row"

{| class="wikitable" style="text-align:center"
|+Vote table
|-
! State !! {{{name1}}} !! % !! {{{name2}}} !! % !! Total
|-
! California
| {{{vote1|0}}}
|| {{decimals|{{#expr: {{{vote1|0}}}*100/{{sum|{{{vote1|1}}}|{{{vote2|0}}} }} }}|2}}%
|| {{{vote2|0}}}
|| {{decimals|{{#expr: {{{vote2|0}}}*100/{{sum|{{{vote1|1}}}|{{{vote2|0}}} }} }}|2}}%
|| {{sum|{{{vote1|0}}}|{{{vote2|0}}} }}
|}

Result

Vote table

State	Bill	%	Bob	%	Total
California	500	38.46%	800	61.54%	1300

Wdfarmer (talk) 10:28, 11 February 2008 (UTC)

Continuing the discussion

Nice work, Wdfarmer! Would it be possible to make a template that generates four contiguous cells in a table row? If so, couldn't we use a template similar to the one above that generates these four cells in each row:

• Pledged delegate estimate
• Superdelegate estimate
• Total delegate estimate
• Percent of total delegate estimate

The idea would be to have table row source code that looked like this:

|style="text-align:center;"| [[Image:ObamaBarack.jpg|100px|Barack Obama]]<br />'''<span style="display:none">Obama, Barack</span>[[Barack Obama]]''' || [[U.S. Senator]]
|| [[Illinois]] || {{ calculate_delegates | pledged=997 | unpledged=127 }} || Current || [[Barack Obama presidential campaign, 2008|campaign article]]<br>
[http://www.barackobama.com/ campaign website]
|-

which generated a row just as we see it currently in the article. --Bryan H Bell (talk) 10:56, 11 February 2008 (UTC)

That wouldn't quite do it, because:

• Pledged delegate estimate = pledged for candidate
• Superdelegate estimate = unpledged for candidate
• Total delegate estimate = pledged + unpledged for candidate
• Percent of total delegate estimate = (pledged + unpledged for candidate) / (pledged + unpledged for all candidates)

So the "calculate_delegates" template would need a third parameter that was the (pledged + unpledged for all candidates). That appears to be a changing number that increases as the primary process continues. True?

Wdfarmer (talk) 20:36, 11 February 2008 (UTC)

Ah, yes, you are quite correct. Hmm. Just brainstorming here, but what if we created two templates that merely returned the sum of its two inputs:

• {{ obama_delegates | pledged=997 | unpledged=127 }}
• {{ clinton_delegates | pledged=920 | unpledged=223 }}

and then created a third template that accepted the first two templates as inputs and generated the four cells we require:

• {{ percent_total_delegates | candidate1={{obama_delegates}} | candidate2={{clinton_delegates}} | resultsfor={{obama_delegates}} }}

Could that work? --Bryan H Bell (talk) 20:57, 11 February 2008 (UTC)

As I think you just realized, you also need to account for Edwards' delegates. But even then, editors would have to change their practice from editing the table to editing the templates, which would be in the global template space instead of the article. That's a pretty big change, and I don't think it would pass consensus. I think the best we could do would be to pass the same (pledged + unpledged for all candidates) number as a parameter to each row_calculation, but that means keeping those numbers in sync: a difficult thing to do. I suppose JavaScript and XML are possible ways out of this mess, but that's probably not currently feasible. Wdfarmer (talk) 21:44, 11 February 2008 (UTC)

Another completely different approach is to let an externally maintained Excel file be the master table, and then use an Excel-to-Wikipedia tool to cut and paste the entire table's text and numbers into the article, replacing the entire table each time. Nice idea, but clearly not suited for a collaborative effort. It also would lose the reference tags that are currently within the table.

In a Microsoft Office business environment, normally all this would be done by having Excel spreadsheets embedded as objects within Word documents, which were then collaboratively edited on a server. Quite a different world here.

Wdfarmer (talk) 22:14, 11 February 2008 (UTC)

You're right about Edwards. We'd need four templates in that case, not three. As far as ease of editing, maybe I misunderstand how templates work, but wouldn't the editors simply need to edit the template parameters (i.e. pledged=997) right on the article itself? They wouldn't need to visit the template namespace unless they needed to change the calculation, which I doubt would happen very much. --Bryan H Bell (talk) 00:56, 12 February 2008 (UTC)

Correct, if the templates didn't have any default parameter values. But then, your third template would have to look like:

• {{ percent_total_delegates | candidate1={{obama_delegates | pledged=997 | unpledged=127 }} | candidate2={{clinton_delegates | pledged=920 | unpledged=223}} | resultsfor={{obama_delegates | pledged=997 | unpledged=127 }} }}

and again we have the nightmare of updating the same number in several different places within the same article. To avoid that, you'd need to drive the whole process by a higher-level template that would be only invoked once, and would generate all the non-header rows for the whole table. A simplified example of invoking such a template, for just three candidates with pictures, allowing sorting by last name, would be:

{{delegate_table_rows
| name1="Clinton, Hillary" | display1="Hillary Clinton" | image1=image:hillary.jpg | pledged1=997 | unpledged1=127
| name2="Obama, Barack" | display2="Barack Obama" | image2=image:obama.jpg | pledged2=920 | unpledged2=223
| name3="Edwards, John" | display3="John Edwards" | image3=image:edwards.jpg | pledged3=26 | unpledged3=0
}}

That's a feasible approach, although you'd lose the ability to tag the generated rows with custom reference tags, etc. afterward. All non-computable information to appear in the table would have to be passed as parameters. The template might even be able to automatically bold the largest entry in a column.
Templates, by my best guess, work like macros in programming languages, or the #define in C language. They just substitute parameter values into the template text, and then that text-with-substitutions is reprocessed, possibly invoking new templates (even recursively). This continues until all substitutions have been made.
Wdfarmer (talk) 08:22, 12 February 2008 (UTC)

You are, of course, correct again. I can't believe it didn't occur to me that one would have to include the parameters for each candidate in the "percent_total_delegates" template I suggested. That would indeed be tedious. As to your suggestion of generating entire non-header rows with a template, it does, as you say, place limitations on our ability to custom-modify the table. I'm starting to think that templates might not be an adequate solution to all the manual math we currently do. Maybe I'm giving up to easily, though. --Bryan H Bell (talk) 06:27, 13 February 2008 (UTC)

Hey, I make lots of mistakes, too. See this latest example, which resulted from this research.  :) Well, at least you know what could be done with a template now. And I've learned something about templates. Good deal. Wdfarmer (talk) 08:12, 13 February 2008 (UTC)

I propose to deny editing to unlogged/unregistered users

I'm wasting much time undoing many modifications that turn the article incoherent. Also when notes are present, unregistered users doesn't respect it. Is it possible to block unregistered? Do you agree? —Preceding unsigned comment added by Subver (talk • contribs) 00:40, 11 February 2008 (UTC)

I saw the work you did today on that front. I would be in favorite of temporary semi-protection Northwesterner1 (talk) 01:54, 11 February 2008 (UTC)

It does get tiring constantly reverting or correcting unilateral changes, particularly to the delegate counts and sources for these. Certainly, some cases are simple vandalism, but in many cases I think the changes are being made by careless editors who are failing to notice the editor's notes and genuinely think they're correcting "errors". I also think many editors are unknowingly or forgetfully applying non-NPOV edits to make their favorite candidate look better. As much as I'd like to just shut down the careless/POV IP users who are making it difficult to maintain this article, these kinds of edits are not vandalism per se and I'm not yet certain that temporary semi-protection is warranted. That said, it might not take long before I'm ready to change my mind on this! --Bryan H Bell (talk) 04:56, 11 February 2008 (UTC)

"still counting" = ?

I'm glad that this column in the super Tuesday table was added, but I don't understand what it means. Does anyone? I noticed this myself; the number of delegates awarded doesn't seem to match the number that states like Colorado are supposed to have. Are they actually still counting, and will they have the results in the next few days or so? I'm really confused about this; thanks if you're able to explain it. --Hermitage (talk) 08:26, 11 February 2008 (UTC)

I second this request. It is a long time from Super Tuesday for the results for so many delegates to be still counting. The article about Colorado itself doesn't seem to mention the 27 "still counting" delegates... Rachel Pearce (talk) 10:09, 11 February 2008 (UTC)

"Still counting" is not the best term. It means that these number of delegates hasn't been assigned yet to one of the candidate. Some caucuses (es Colorado) don't end in a day. In other cases, the final results are not yet completely available, so the complete delegate count is still impossible. If you have a better term to substitute "still counting" please propose it. --Subver (talk) 15:43, 11 February 2008 (UTC)

Perhaps "not projected yet" would be a better term? Jon (talk) 18:40, 11 February 2008 (UTC)

In any case, I think Solver is right about Colorado as well as the other cacus state with a lot of non projected delegates. As for the 1/2/3s in various primary states what's probably going on is that the estimated margin in a few Congressional Districts (or possibly state wide) was very close to a cut off point that determine :number of delegates so they are waiting for the counties and states to certify the results. Jon (talk) 18:45, 11 February 2008 (UTC)

"Undetermined"? Northwesterner1 (talk) 19:17, 11 February 2008 (UTC)

I'm not arguing for the column to be deleted (I find it very helpful) or even necessarily renamed; I was just hoping that someone could provide a link to an article describing why some of those delegates haven't shown up yet, and add it to the column header as a reference. I checked the CBS site and didn't find anything like that. Thanks for whatever you can do! --Hermitage (talk) 19:30, 11 February 2008 (UTC)

http://www.nytimes.com/2008/02/09/us/politics/09delegates.html?hp . I'm changing in "Unassigned yet" --Subver (talk) 21:27, 11 February 2008 (UTC)

Thanks for that link, but I think the article could do with a snappy decription of why some delegates remain unassigned (I am not competent to provide such a description as I have barely understood the NYTimes article). Also, surely ALL the delegates from states which have yet to hold their Primaries/Caucuses are Unassigned? So shouldn't these columns be full of numbers? Rachel Pearce (talk) 11:20, 12 February 2008 (UTC)

Obviously they are unassigned. But I think it is useless to fill the entire column, because now the "not yet assigned" well highlights the big or small incompleteness of the already hold primaries. --87.13.55.17 (talk) 13:48, 12 February 2008 (UTC)

Agreed 100% with 87.13.55.17. Jon (talk) 18:35, 12 February 2008 (UTC)

I took a stab at clarifying this in a new section on "reporting delegate totals." Even better would be a small table giving the total count from each of the major news organizations, for a side-by-side comparison, but of course that would require that we maintain it. Northwesterner1 (talk) 20:29, 12 February 2008 (UTC)

I think it's too difficult to maintain coherent, always updated, different sources. Not to complicate our work, I added a reference to other sources table by DCW. Do you agree? Please correct my English that is very bad :D --Subver (talk) 21:54, 12 February 2008 (UTC)

I like the link you added. Good solution. Northwesterner1 (talk) 06:07, 13 February 2008 (UTC)

Delegate bonuses

The eventual delegate bonuses are listed here assigned to each state (including Puerto Rico) , and they are already counted in our tables. References here: http://www.thegreenpapers.com/P08/D-Alloc.phtml --Subver (talk) 21:27, 11 February 2008 (UTC)

Consensus Sought: Change Map Colors

Appearance with various types of color blindness or a monochrome screen

File:Results by County of the 2008 Dem Presidential Primaries.png

Proposal: Change the color scheme of maps used in this article.

To quote Mikelove on 7 January 2008 from Archive 1 of this talk page:

Does anyone think it might be a good idea to change the colors on the map so that no candidate is assigned blue? I'm actually an Obama supporter, but I worry that giving him the color generally associated with the Democratic party might introduce some small amount of bias, suggesting he's already the Democratic candidate or somesuch. Likewise on the Republican map it might be best to avoid making anyone red (as Giuliani is now).

• I too am an Obama supporter, but the current colors introduce BIAS by making Obama seem more liberal or Democratic (big D). -Infoporfin
• The colors are also difficult to distinguish for people with color blindness (probably 5% of wikipedia readers) and people viewing wikipedia on monochrome monitors (A larger percentage than one might think, remember: Wikipedia is global.) I have used this online [color blindness simulator] to create this animation showing how the colors might appear to various readers. Notice that the red blends almost completely into the black or the green (depending on the type deficiency) and that for a person with Monochromacy (0.0001% of the population) and on a Monochrome machine (much more prevalent) the red/green/blue contrast completely disappears.

-Infoporfin (talk) 23:44, 11 February 2008 (UTC)

• I actually don't find this map very useful, whatever the colors are. It's hard to read -- it's a complicated way of showing that the candidates are essentially tied in the delegate count. But that information is easier to digest from the tables. And it's hard to tell where the candidates' support is coming from in terms of states or regions of the country, which is the value of getting this information in the form of a map. I recognize that a winner-take-all state map is misleading. Of the three maps on the main results article, I find the county-by-county map the most useful. That map also has the benefit of being color-coded in a more neutral way. So I would propose deleting this map altogether and replacing it with the county map. Northwesterner1 (talk) 00:21, 12 February 2008 (UTC)

• I'm a web developer by trade, so in regards to whatever color scheme is used, here's what I sugest. First print that graphic on a black & white printer; if you can't tell the difference between some of the colors on the printout, it's a poor color selection. Next Red-Green color blindness if the most common of then, so don't use both Red & Green. Now as to the design of the pictograph itself, it's extrely poor, if it's desirable to include every candidate's count in every state on the graph, listing the states in alphabetical order would be a much less confusing visualization. Jon (talk) 18:44, 12 February 2008 (UTC)

• We never reached consensus on this, but I deleted the cartogram and replaced it with a new graphic I designed. I hope it is easier to read. I used purple/green for color neutrality. Northwesterner1 (talk) 18:18, 6 March 2008 (UTC)

Support

---

• Support as the proposer. -- Infoporfin (talk) 23:44, 11 February 2008 (UTC)

---

---

Delegate Count Source

Why are different sources used for superdelegates versus pledged? if the data were all taken from CBS, Obama would be in the lead, if it were all from DCW, then clinton would have a much larger lead in the delegate race..

should this not be streamlined?

The Talking Sock talk contribs 00:22, 12 February 2008 (UTC)

Check out the discussion at Candidates: Using CNN Election Center 2008's numbers to see why different sources are being used for pledged vs. superdelegates. --Bryan H Bell (talk) 01:00, 12 February 2008 (UTC)

I strongly suspect that Obama will be in the lead here after today's delegates from the Potomic[sp?] Primaries are allocated. Jon (talk) 23:52, 12 February 2008 (UTC)

Campaign Narrative

It seems odd to me that this article is missing a narrative of the campaign. Reporting primary rules and delegate totals is a useful service, but the article in its current form fails to provide a meaningful narrative of the race that would be useful now to someone just "tuning in" or that will be useful in the future as a historical account. Such a section would tell the story of the campaign and its key moments, the major turning points, the events that generated news stories and discussion. Some of this is covered in the campaign articles for the individual candidates, but as far as I can see, we have no article that gives an overall narrative of the race. That seems to me a striking omission. 2004 has something like that. It's not very detailed, and we could do much better this year, but it might provide a useful model. Northwesterner1 (talk) 05:05, 12 February 2008 (UTC)

The main problem with trying to write such a section now is that it would run into Original Research and Unverifable Sources issues. After the primary is over, then there should be plenty of sources to cite, but not now. Jon (talk) 18:54, 13 February 2008 (UTC)

I disagree. There are an enormous number of magazine and newspaper articles that can be used as sources. Indeed, Hillary Clinton presidential campaign, 2008 and Barack Obama presidential campaign, 2008 both have extensive narratives of key moments in the race. There will be original research and NPOV issues, as there are with most articles about current events, but these can be addressed as they would be in any other article. I think it would be useful to readers to have this narrative now, rather than after the fact. Northwesterner1 (talk) 22:27, 13 February 2008 (UTC)

I have reorganized the sections. I think that a small chronicle can be added after each month/superTuesday race. --Subver (talk) 22:30, 13 February 2008 (UTC)

Superdelegate source

The table now shows superdelegates at 230/142 using this[1] source. That source has five news sources with different numbers - none of which match ours'. What source is actually being used here?--Appraiser (talk) 19:58, 13 February 2008 (UTC)

I see it's fixed now, at 232-143. For future reference, those numbers should always match the "DCW" column at the source. We have a lot of editors selectively changing results to numbers that don't match the consensus sources. (Not bad faith edits, I don't think, just unaware of the consensus.) So the totals get off sometimes... Northwesterner1 (talk) 20:39, 13 February 2008 (UTC)

And to summarize that consensus, the reason we are using DCW is that it is the only fully transparent and sourced superdelegate count that is available. The major news organizations just give totals of their estimates. In contrast, DCW has compiled an endorsement list that names the superdelegates and links to news articles about their endorsements. Northwesterner1 (talk) 20:49, 13 February 2008 (UTC)

Now that I've looked at DCW, I'm glad we're using it. It is really useful to know exactly who has pledged to whom. And it looks like the bloggers there keep track of any conflicting statements, or changes in allegiances pretty promptly. Thanks for cluing me in.--Appraiser (talk) 20:52, 14 February 2008 (UTC)

Puerto Rico take-all or not?

I hided the Puerto Rico special rules section. We need to verify if Puerto Rico has or not a take-all system. By The last 2004 Democratic Primaries it seems yes... --Subver (talk) 22:30, 13 February 2008 (UTC)

This article seems to contradict David Brooks and Michael Barone, but I think the possibility of different rules or application should at least be broached in the entry. It is an issue being debated in the media and should be presented as such. http://blog.washingtonpost.com/fact-checker/2008/02/will_puerto_rico_decide_everyt.html —Preceding unsigned comment added by 76.88.53.100 (talk) 23:48, 13 February 2008 (UTC)

This entire Puerto Rico section should be deleted. The only source for this notion are Brooks an Barone, and both admit to being wrong in the above mentioned article: "Both Brooks and Barone backed off the "winner takes all" talk when I caught up with them today. Barone said he had based his analysis on past elections, when Puerto Ricans traditionally plumped for one candidate. Brooks said he got his information from Barone." —Preceding unsigned comment added by 24.82.186.87 (talk) 08:18, 20 February 2008 (UTC)

2004 is a bad example; the campaign had effectively ended about 3 months before Puerto Rico in 2004 (shortly after Super Tuesday), and there are several states in May and June which saw their entire delegations go to Kerry. '96 and 2000 would have similar issues; perhaps '84, '88, or '92 would be better to check for precedence. Jon (talk) 18:50, 14 February 2008 (UTC)

1984 appears to be the most recent close democratic primary in recent history (post enlargement of super delegates), so if any previous article can tell us anything about PR that would be most likely. Jon (talk) 05:39, 16 February 2008 (UTC)

Check this New York Times 1988 report: http://query.nytimes.com/gst/fullpage.html?res=940DE4D61230F93BA25750C0A96E948260 —Preceding unsigned comment added by Seablade (talk • contribs) 17:03, 16 February 2008 (UTC)

Numbers don't match

The delegate numbers in the candidate table of the article are not matching the delegate numbers in the table at the far botom of the table so I've placed a contradict tag on the article until this is resolved. I don't know which one is correct. Jon (talk) 05:36, 16 February 2008 (UTC)

Solved. That table was not using DCW superdelegates numbers. And it was useless and repetitive. --Subver (talk) 05:47, 16 February 2008 (UTC)

Half Super-Delegates???

How is it possible for candiates to even have a half a superdelegate in the delegate counts? (Obama showing 160.5 super delegates, hillary with 230.5 super delegates). Jon (talk) —Preceding comment was added at 21:48, 17 February 2008 (UTC)

Because superdelegates from Democrats Abroad get 1/2 vote each. Simon12 (talk) 22:33, 17 February 2008 (UTC)

One related issue, this page is stating you need 1/2 a delegate less to win than the results page. I think these two ought to be consistent. Jon (talk) 00:58, 20 February 2008 (UTC)

2,024.5 vs 2,025 Majority

2,024.5 keeps on getting changed to 2,025 (likely out of confusion/misunderstanding). I changed "simple majority of the 4,048 delegate votes, or at least 2,024.5 votes" to "simple majority of the 4,048 delegate votes, i.e. more than 2,024 votes". Both are correct, but the latter avoids confusion. Andareed (talk) 07:02, 20 February 2008 (UTC)

I think part of it though is that CBS & CNN both list 2025 on their web sites as the number needed. Another is unless the top section of the results page has been changed since last night, it is listing 2025 as well. (In fact I'm not aware of a source that states 2024.5 is needed.) Jon (talk) 14:26, 20 February 2008 (UTC)

If the votes are 4048 and half votes are admitted (and they are), the majority is 2024,5. In fact if at the convention votes are 2024,5 for A and 2023,5 for B (2024,5+2023,5=4048), then A has majority. So the "magic number" is 2024.5. Also DCW reports 2024.5 and also the Result page (because I had edited :-) ). So, if you agree I would change back to 2024.5 .--Subver (talk) 15:15, 20 February 2008 (UTC)

I understand why majority is 2024.5. My edit was intended to avoid confusion (and subsequent reverts) by stating that majority was any number of delegates > 2024; either 2024.5 or 2025 works. Andareed [edit: "> 2025" -> "> 2024"] (talk) 18:58, 20 February 2008 (UTC)

In that case, either 2024.5 or 2024 1/2 is okay here (and there, but not 2024,5) , but add a cite to DCW for it on both places as well. Jon (talk) 19:00, 20 February 2008 (UTC)

DCW is a blog with the same autorithy of us...moreover there is no specific reference. This is math. --Subver (talk) 19:34, 20 February 2008 (UTC)

I don't get it. I like Andareed's solution, and no one here really argues against it on the merits. Andareed's solution is accurate, as "more than 2024 votes" means 2024.5, and it has the added benefit of preventing edit conflicts. Let's just leave it at "simple majority of the 4,048 delegate votes, i.e. more than 2,024 votes." Northwesterner1 (talk) 21:56, 20 February 2008 (UTC)

Removed "States won" column

I have removed the "States won" column in the candidates table. I feel that this information is misleading in describing a candidate's standing in the contest for nomination. Democratic Party candidates cannot win states, they can only win delegates. It is true that a candidate can win the most delegates in a state, but because states have significantly different numbers of delegates, the number of states in which a candidate has won the most delegates does not mean they are closer to securing the Party's nomination than another candidate. The column thus introduces POV into this article by portraying the contest results in a way that favors one candidate. --Bryan H Bell (talk) 15:46, 29 February 2008 (UTC)

States Won is not POV, it is a fact. Voters vote by State. Delegations are assembled by State. States matter. This is an important statistic used by the press and is of interest to our readers.

I agree with you that the number of states won should not be editorialized or emphasized unfairly in the article copy. But to present the data does not reflect bias. (Otherwise, this page would have to remove the maps, state-by-state results, etc. so as not to "mislead" readers -- who actuall are interested in this statistic.) --Scantron2 (talk) 16:01, 29 February 2008 (UTC)

I'm against keeping the column, simply because "states won" is not relevant to the selection process. However, if we do end up keeping it, we should add another note to the notes section, along the lines of: "The states won column indicates the winner by popular vote. Primaries in the Democratic presidential nomination process do not use a winner-take-all system." This would avoid confusion that the number of states won is somehow directly relevant to the outcome of the election. Andareed (talk) 17:07, 29 February 2008 (UTC)

I also note that if we leave the "states won", what's to stop other columns from being added? Perhaps someone will want to add a "popular vote" column next... Andareed (talk) 17:21, 29 February 2008 (UTC)

Thanks for the Comments, Andareed. Perhaps I should clarify my perspective.

• First, I'm not married to the States column. If the consensus is to pull it, then OK.
• The more important thing is that state-by-state elections are the process by which this plays out. The process plays for out delegates in the end, but it's really a series of state contests. That's why I see this as useful. It's in the public imagination. There's intensive media as to who's won which state, who's expected to win each state, who's won how many states, or how many in a row. For this reason, I think it's a disservice to our readers to willfully conceal the tabulation.

Maybe the problem is that it sits next to the delegate number? One resolution would be to create a supplemental table that tabulates the # of states and the popular vote -- by recognized contests, recognized + florida, and recognized + florida + michigan. This could come with narrative that draws exactly the distinction you mention -- that only the delegate count is considered, but that other metrics you can look at. This would be consistent with reporting in the general where they do talk about electoral votes, states won, and popular vote. Thoughts??? Thx! --Scantron2 (talk) 19:06, 29 February 2008 (UTC)

Scantron2, I think you may be confusing the process of delegate selection with the results of that process. While it is true that the process of nominating a Democratic presidential candidate takes place in a series of state contests, the Democratic Party does not add up the first place wins in each state to determine the winner. It's easily possible for a candidtate to "win" in more states and still lose the nomination. This is exactly the situation we face in the current race for Democratic nomination. Obama has racked up more first place state wins than Clinton, but could still lose the nomination if Clinton racks up first place wins in states with lots of delegates. When the media speaks of a candidate "winning" a state, it is usually speaking in shorthand for winning a significant number of delegates. This is why you see more intense interest in populous states with lots of delegates (such as California or Texas) and less interest in states with few delegates (such as Vermont or Wyoming). While speaking of "winning a state" is conveniently simpler and shorter than speaking of proportional representation, doing so introduces confusion and can be misleading about a candidate's actual standing. I think we should be careful not to introduce such confusion into this article. --Bryan H Bell (talk) 00:56, 1 March 2008 (UTC)

What is the source for this column? Is it nothing more than a count of the highlighted cells in the January and February tables later on in the article. Strictly speaking, wouldn't this qualify as OR? I just made a correction to the January table, which invalidates the 26-12 numbers currently appearing in the States Won column, anyway. New Hampshire was highlighted for Clinton, even though she and Obama both have 9 delegates. Nevada was also highlighted for Clinton, even though Obama has 13 delegates to her 12. This changes the States Won numbers to 27 for Obama and 10 for Clinton. I haven't updated the States Won column because... I don't know its source. I don't like the idea of tallying up entries from elsewhere in the article to arrive at these numbers. -FeralDruid (talk) 19:11, 29 February 2008 (UTC)

Hmm, looking at the article more closely, it already appears as though large portions of the tables are added from other numbers anyway. So I guess there's already precedent within the article for tallying up numerical results from elsewhere within the article. I'm not convinced this number is useful, though, because as others have pointed out, state tallies aren't what determines the candidate. If the point is to draw attention to popular vs. delegate vote, state tallies really aren't the number to look at anyway, IMHO, but overall voter number (i.e. 23 million for one candidate, 18 million for another). -FeralDruid (talk) 19:26, 29 February 2008 (UTC)

You're convincing me that maybe the column is a bad idea. I hadn't thought about the source, and that is a strike. Maybe the answer is to either:

• drop the column and the subject altogether, or
• do a side-table on the basis of popular vote. I'd suggest realclearpolitics as a source. The emphasis would be on total popular vote (comparing incl/excl of MI/FL). In this contest, it would be possible to include the states won tabulation - but solely on the basis of popular vote. Whatever, any material should be solid, cited, free of editorializing, and have an explained methodology. IMO. --Scantron2 (talk) 19:34, 29 February 2008 (UTC)

Popular vote tallies might be useful, but present a host of problems in a nomination process that uses both primaries and caucuses. We've been struggling with this on Talk:Results of the 2008 Democratic Party presidential primaries#Popular vote. Also see the note above that article's popular vote section. --Bryan H Bell (talk) 01:06, 1 March 2008 (UTC)

I'm opposed to the "states won" column. Reasons: (1) I think it is useful to keep this article contained to the most relevant statistics only. We have a more detailed results article, and if readers are interested in popular vote totals, etc., they can head over there. (2) This article already shows how many states have been won by each candidate. Wins by number of delegates are visually represented by colored cells on the state-by-state delegate columns. And wins by popular vote are represented in the second map. Those methods are better because they put the number of states won in the necessary context, i.e. when the states were won, how large they are, whether they are caucus or primary states, where they are geographically located, whether the wins were close or blow-outs, etc. It's actually easier to get a sense of what's going on with the race in the state-by-state columns, and we don't need to duplicate this information in the top table. (3) We have a potentially very large and complex problem coming up with the Texas primary. If the primary and caucus votes end up differently, we're not going to be able to call a winner based on "popular vote." In the interest of avoiding a mess of footnotes and captions explaining the table -- and the accompanying edit conflicts on election night -- I think it's best to just leave this column out. I don't think this is a NPOV issue, as Bryan suggests, but I do think it's an issue of keeping the article as clean as possible so that it can be of greatest use.Northwesterner1 (talk) 03:45, 1 March 2008 (UTC)

Upon further reflection, I think perhaps I was wrong about the "States won" column introducing POV. The totals of states "won" could shift towards either candidate and do not in themselves unfairly favor one candidate over another. I think that the introduction of this column at this point in the race, when the data clearly favors one candidate, threw me off. However, I still think the column should be removed because it misleads Wikipedia's users about the true standing of the candidates in winning the nomination. It is a bit like the joke about the loser in a two-person footrace proclaiming that he placed 2nd while his opponent only came in "next to last." --Bryan H Bell (talk) 08:32, 1 March 2008 (UTC)

The consensus seems to be favoring removal of the column, so I'm going to remove it. --Bryan H Bell (talk) 08:35, 1 March 2008 (UTC)

Good Job You Guys

Just want to say to anyone working on this article: it looks so much better than it did a month ago. Good job. Pullarius1 (talk) 23:10, 4 March 2008 (UTC)

popular vote?

Is there any way to get a count of actual votes cast by actual human beings in primaries and caucuses, for the various candidates? I don't mean delegate numbers, but voters, as in "one man one vote" kind of voters. --24.85.68.231 (talk) 07:43, 5 March 2008 (UTC)

The individual state primary/caucus pages have this information. Andareed (talk) 07:46, 5 March 2008 (UTC)

The main results article, Results of the 2008 Democratic Party presidential primaries, also has the nationwide popular vote total. Note, however, that this number is somewhat problematic; many fewer voters turn out for caucuses, so it's hard to measure candidate support by comparing the popular vote in a primary state with the popular vote in a caucus state. Northwesterner1 (talk) 08:55, 5 March 2008 (UTC)

It's worth noting that despite worries to the contrary, the pledged delegate count actually seems to be a reasonable proxy for the popular vote. As I write this, the pledged delegate count stands at 1341 Obama to 1191 Clinton (52.44% to 47.31%), while the popular vote count looks like 10,300,410 Obama to 9,375,213 Clinton (52.35% to 47.65%). Northwesterner1 (talk) 09:06, 5 March 2008 (UTC)

Northwesterner1, I'm curious as to which source and what specific methodology you're using? Does that include yesterday's totals? (so far). (Are you counting all caucus states? Does that include extrapolations where the state democratic primary didn't actually release that count and so it's being extrapolated by percent of "delegate equivalents" times number of those attending? Does that include unsanctioned contests (MI/FL/ [and WA straw poll])? Is it potentialy double counting states with both a primary & a caucus (TX & the WA straw poll])? Earlier this season, Sabsito[sp?]'s Crystal Ball made the point in one of their articles that it's in the eye of the beholder. Jon (talk) 14:17, 5 March 2008 (UTC)

My source? Wikipedia, of course! I was using the pledged delegate count from this article and the popular vote count from the main results article. I see now, however, that the editors on that page have not updated the count to match the stated source. The source now says 12,957,086 Obama (51.15%) to 12,375,168 Clinton (48.85%). The pledged delegate count per CBS is currently 1354 Obama (52.82%) to 1209 Clinton (47.17%), so it looks like my point about the percentages matching up nicely no longer holds. (I've adjusted the pledged delegate percentages to exclude Edwards, as the source doesn't have popular vote totals for him.) The popular vote total excludes Fl and MI; it excludes the caucus states of IA, NV, WA, and ME; and it doesn't extrapolate from caucus attendance in those states that do report numbers. Obviously, the popular vote count looks more favorable for Clinton if you include FL and MI and more favorable for Obama if you extrapolate from the caucuses to project what primary turnout might have been. Both of those adjustments would be extremely problematic, though. I assume the editors over at the main results article have looked at various popular vote estimates and found this one to be the best for Wikipedia. Northwesterner1 (talk) 20:13, 5 March 2008 (UTC)

Texas combined Primary / Cacus and highlighting

In Texas, Clinton is projected to win the primary, but Obama currently has the lead in the Cacus (too close to call). If Obama later becomes projected to win this cacus, should we highlight both cells, unhighlight both of them, are come up with some tiebreaker of who gets the most combined delegates? (As a comment, this would be much easier to handle if CBS had split TX delegates into primary & cacus determined ones on that result summary instead of lumping together.) Jon (talk) 14:08, 5 March 2008 (UTC)

I don't think we should split Texas caucus and primary results. The main results article and the Texas results article will have detailed numbers, but the focus of this article is reporting the pledged delegate count in each state. Each state has its own system for allocating delegates. Texas has decided to have a two-part system for figuring out its delegation, but it's not as if it gets a second delegation to the convention or should count as a 51st state. That said... we clearly do have a problem with how to determine the popular vote result here. My feeling is that either (a) we go with a strict total "popular vote" count, which will likely go to Clinton, as caucus attendance will be much lower, or (b) we change the highlighted cell to reflect the winner of pledged delegates. If we take option (b), we will also need to change the highlighted cell in New Hampshire and Nevada. Northwesterner1 (talk) 20:27, 5 March 2008 (UTC)

A third option might be to use a lighter shade of the same color to highlight cells in IA, NV, NH, and TX, highlight both cells in each state, and then put footnotes, i.e. "Edwards placed second in the statewide caucus vote in Iowa, but Clinton received 15 delegates to Edwards' 14"; "Clinton won the popular caucus vote in NV, but Obama received an extra pledged delegate by winning in rural areas throughout the state"; "Clinton won the popular vote in NH, but Obama and Clinton received the same number of pledged delegates"; "Clinton won the popular vote in the Texas primary, which accounted for 2/3 of the pledged delegates selected, while Obama won the caucus vote, accounting for 1/3 of the pledged delegates." Northwesterner1 (talk) 20:36, 5 March 2008 (UTC)

I don't recommend using a lighter shade of the same color to highlight certain cells. As we've discovered on the main results article, using color to convey information is quite problematic, and doesn't conform with Wikipedia's WP:COLOR guidelines. In fact, this article's use of color already doesn't conform to that guideline. Instead, I think it best to keep things simple by sticking to the methodology used in the source we're using. That means combining the delegates from both the Texas primary and caucus as CBS does and highlighting the cells according to CBS's listing of the projected "winner" in each state. Also, we should keep in mind that the primary and caucus "wins" are not equal. There are 126 delegates determined by the primary and only 67 determined by the caucuses. Further, the caucus delegates aren't actually determined until June 7 so their results at this point are mere estimates. --Bryan H Bell (talk) 21:35, 5 March 2008 (UTC)

Northwesterner1, you forgot one from here in Missouri (Obama won the popular vote but tied with Clinton in delegates.) It was basically a landslide for Clinton in MissorUH that was canceled out by Obama's landslide in MissouRI. Jon (talk) 02:05, 6 March 2008 (UTC)

Where a popular, significant delegate-getting vote is available, that should be what's portrayed in the chart. Consider that in Texas those who caucused had to have voted in the primary first, meaning their vote was literally counted twice. That says nothing of who won the state, but who has the most fervent supporters. Nevada too; Clinton won the popular vote, but Obama got more delegates. That anamoly would never happen in a general; I think the chart should be consistent with that idea in mind. --Kallahan (talk) 02:15, 6 March 2008 (UTC)

I agree 100%. We should not be in the business of highlighting election quirks. I think our Nevada, New Hampshire, and Missouri highlighting is spot-on: we're indicating who gathered the most votes in each contest. --Scantron2 (talk) 04:40, 6 March 2008 (UTC)

Texas is a different situation than Nevada. We don't have to answer the question of what "winning" means -- winning the popular vote vs. the resulting delegates. In Nevada's case, the contest could be argued both ways. Here, there are two answers because there are two separate contests for delegate selection.

In Texas, the primary ran 51-47 for Clinton and she also has a majority of delegates through this process (65-61 per Texas Secretary of State). No ambiguity. The current caucus results are 56-44 Obama per the Texas Democratic Party website. This is a partial result and delegates are TBD, but FWIW, analysis on demconwatch projects 37-30 Obama. DCW's "placeholder" (per NPR) is 30-27 for now.

The problem here is that we have two separate, equally legitimate processes in play at the same time.

If the consensus is to combine the two contests into a single row and single indication of "winning", my vote would be to go with the final delegate count. Otherwise, we're favoring one process over the other. And again, this isn't a "Nevada" situation where the vote and delegates differed. This is two, equally valid and legitimate election systems, each one of which yielded appropriate and consistent results to the process. If we're going with a combined line, highlight should go to the overall delegate winner as that's the only "objective" criteria and avoids our bias in favoring one contest over the other.

• My preference, however, is to present Texas more accurately. It is a special case comprised of two separate contests, each one of which is a substantial election in its own right. IMO. Regards, --Scantron2 (talk) 04:20, 6 March 2008 (UTC)

• I understand the sentiment, but Texas favors one over the other by virtue of putting 2/3rd of the delegates into the primary AND with the caucuses being contingent on primary participation. Primary is favored, and it's more representative. --Kallahan (talk) 04:45, 6 March 2008 (UTC)

Thanks, Kallahan. Quick response, then I'll be quiet (and go with the consensus). Though TX "favors" (weights) the primary at 2/3, the balance of the two a weight of its own. A 4pt victory in 2/3rds and a 12pt loss in the remaining 1/3 is a net loss. That's not me talking, that's the Texas Democratic Party. I don't like their system -- wish they went all primary -- but I think we need to have some respect for their process -- or two processes. I think we're missing the big picture: Clinton won the primary and Obama won the caucus. This is why I favor splitting this out into two lines to more accurately reflect the facts and provide more clarity for our readers into the true result. Again, thanks, and I'll go along with whatever. --Scantron2 (talk) 05:05, 6 March 2008 (UTC)

In response to Subver's table layout example below, I tried making this:

Details			Delegate Votes						Pledged Delegate Votes Estimate by CBS
Date	State	Type	Pledged Delegate Votes				Super- delegate Votes	Total	Obama	Clinton	Not Yet Assigned
			District-level	At-large	PLEO	Total
March 4, 2008	Ohio	primary	92	31	18	141	20	161	64	74	3
	Texas	primary	126	0	0	126	35	228	61	65
		caucus	0	42	25	67			20	16	31

Would this be clearer than Subver's example, or more confusing? (P.S. I placed this posting up here above the "Moving toward consensus" section in order to keep this discussion from getting in the way of the voting). --Bryan H Bell (talk) 11:07, 6 March 2008 (UTC)

I like this. Northwesterner1 (talk) 11:15, 6 March 2008 (UTC)

I think it's better to keep together the delegate totals, in order to maintain comparability among states. And if we divide these totals we have to divide a lot of others. It you want...OK for primary/caucuses in two lines. --Subver (talk) 12:00, 6 March 2008 (UTC)

Subver, since you applied the split using your format below to the article, I applied a variation of the above that might work. I split the District, At-Large, and PLEO numbers, but left the totals unsplit. --Bryan H Bell (talk) 18:19, 6 March 2008 (UTC)

Hi, wiki-novice here...sorry I'm not sure how to comment properly...Just want to note that I believe the green/purple map of 'popular vote margins in the Democratic primaries' at the top of the article should have Texas changed to purple for Obama since he's won the most delegates total for the state. You can argue that the description in the picture only mentions primaries, but in that event, why are the other caucus-only states colored? 72.48.196.196 (talk) 08:37, 9 March 2008 (UTC)

Welcome to Wikipedia! Thanks for your comment. The map you mention is a map of the "popular vote," which Clinton won in Texas (when you add caucus turnout + primary turnout). I'm working on another map that focuses on pledged delegates. Northwesterner1 (talk) 09:33, 9 March 2008 (UTC)

Good lord, I'm a dolt. And a little buzzed on a Saturday night. That didn't need to be bolded. My apologies! Carry on, you're all doing excellent work. 72.48.196.196 (talk) 10:00, 9 March 2008 (UTC)

Hey, no worries. You're probably not alone in that confusion -- anything we can do to clarify the results helps, I think. Northwesterner1 (talk) 10:04, 9 March 2008 (UTC)

Moving Toward Consensus

Lots of good points here, but the discussion is getting hard to follow. I'm listing all potential options here. Please vote below. Add more options if you think I'm missing anything. Keep your rationale short (if you need to make a longer point, please put it above with the rest of the discussion).

For the record, the main election results page at CBS has a single row for Texas, with the pledged delegates from the caucus and the primary combined. They don't call a "winner" on that page for any state. On the Texas page, they have the caucus and primary results separated, and they are calling a winner for each contest separately. They call winners for other states based on vote totals, not delegates won (e.g. Nevada). Note that for the moment Obama seems to have caught up with Clinton in the Texas primary delegate count (81-81), so we may have a Nevada situation here with the primary alone, no matter what happens with the caucus. Northwesterner1 (talk) 08:23, 6 March 2008 (UTC) UPDATE: CBS must be wrong here. It can't be 81-81 in the primary alone, as there are only 126 primary delegates at stake. They must be erroneously including caucus delegates in the primary table. Northwesterner1 (talk) 08:59, 6 March 2008 (UTC)

I believe we have consensus:

• Do not include popular vote totals on this page. Leave those numbers to individual state articles and the main results article.
• Stick with CBS as our source for pledged delegates and the Democratic Convention Watch blog for superdelegates.
• Leave the top-level summary table as is. Do not add a "states won" column.
• I added an infobox above at #Talk Page Consensus Am I missing anything?

I believe we still need consensus on Texas:

• (A) Leave Texas as one row. Calculate the "winner" by adding the primary vote count and the caucus vote count from the CBS Texas page and highlight the resulting cell.
• (B) Split Texas into two rows. Highlight the winner for each contest, as called by CBS on the Texas page.
• (C-1) Leave Texas as one row. Highlight both cells or highlight no cells. Add a footnote with explanation. Leave other states as is.
• (C-2) Same as above, but also use the both cells/no cells strategy for other states where the "winner" is ambiguous (NH, NV, MO).
• (D-1) Leave Texas as one row. Highlight the winner of the pledged delegate count, according to the CBS main page. Leave other states as is.
• (D-2) Same as above, but also use the pledged delegate strategy for all states (requiring a change in NH, NV, MO).
• (E) Split Texas into two rows. Highlight the winner of the pledged delegate count for each contest separately and for all other states (requiring a change in NH, NV, MO).
• (F) Leave Texas as one row. Highlight the winner according to the "starred" projected winner on the Texas page. If CBS stars two different winners, revert to B or C-1. Leave all other states as is. (added by Bryan and modified by Northwesterner)
• (G) Am I missing anything?

VOTE HERE

• Changing my vote to B, convinced by Andareed's rationale and by the look of the Subver/Bryan table. I also like D-2 and C-2 as solutions. Northwesterner1 (talk) 11:14, 6 March 2008 (UTC)
• B first choice, C-1 second choice. In NH, NV, and MO, the media consistently named one candidate to have "won" the state. Assuming Obama wins the caucuses, the media will likely label Obama as the winner of the caucuses (with Clinton winning the primary). Assuming this to be the case, Texas will have two "winners" and this should be indicated by highlighting the winners independently. I prefer B because it makes the primary vs. caucus issue quite clear. It also conveys more information, since it lists the delegates won in each of the primary and caucuses separately. Of course if Clinton wins the caucuses, this is a non-issue. Andareed (talk) 08:47, 6 March 2008 (UTC)
• F first choice. D-2, second choice. Choice F seems to be the way we've been doing it so far. Since we're citing CBS as the source for these columns, we should use it consistently for all data points in those columns (including the highlighting). --Bryan H Bell (talk) 09:01, 6 March 2008 (UTC)
• B. As below:

Details			Delegate Votes						Pledged Delegate Votes Estimate by CBS
Date	State	Type	Pledged Delegate Votes				Super- delegate Votes	Total	Obama	Clinton	Not Yet Assigned
			District-level	At-large	PLEO	Total
March 4, 2008	Ohio	primary	92	31	18	141	20	161	64	74	3
	Texas	primary/caucus	126	42	25	193	35	228	61	65
									20	16	31

The sum is 81-81. The 61-65 is the official result. The 16-20 is the CBS estimate for caucuses (81 minus 65 and 81 minus 61). The popular vote winner is highlighted. There is no need to modify other primaries. (e.g. Washington also have both primary and caucuses, but the primary doesn't add delegates). --Subver (talk) 10:29, 6 March 2008 (UTC)

• B For reasons above. But also perhaps this would be a good time to also become wiki color compliant by adding Bold text to the highlights for the winners. The Bold text currently representing totals of pledged delegates and total delegates for the state should also change to Italtics as part of this. Jon (talk) 14:17, 6 March 2008 (UTC)

• B. Prefer upper layout, but both are nice and clear. Otherwise, I have a preference for popular vote (or pop+cauc) and not to alter the NV/MO/NH rows. --Scantron2 (talk) 16:32, 6 March 2008 (UTC)

I think there is sufficient consensum, and I'm applying the modification. I split primary/caucuses cell, but I keep joined the totals, as above. --Subver (talk) 16:01, 6 March 2008 (UTC)

Kudos to Bryan and Subvertigo for the table formats. Simple, elegant, easy-to-understand. --Scantron2 (talk) 16:37, 6 March 2008 (UTC)

• B. Coming in too late for the poll, I'm glad to see you agreed on the choice I would have preferred. ;) Separate meetings that produce delegate allocations should indeed have separate rows. In keeping with that, there's something else that I haven't seen mentioned: isn't Washington state a similar case? See my separate section below. Wdfarmer (talk) 19:18, 6 March 2008 (UTC)