User:A Red Cherry/sandbox: Difference between revisions
A Red Cherry (talk | contribs) ←Replaced content with 'disruption takes place? newly formed alliances and composite position to compromise for you is amount to essentially not a dirty world well working word right now so lol! haha ... what? man, well eh fr eh soaking now then like!' Tags: Replaced Reverted |
A Red Cherry (talk | contribs) No edit summary |
||
Line 1: | Line 1: | ||
[[File:2020 Cook PVI.svg|thumb|360px|Map of 2016-2020 Cook PVI for all voting entities in [[United States presidential election]]s (states, federal district, congressional districts of Maine and Nebraska)<br><br>Each [[United States of America|United States]] state and federal district, as well as the congressional districts of Maine and Nebraska, is displayed herein.<br><br>'''Legend:'''<br>{{legend|#1a3165|State or district has a Cook PVI of D+10 or greater}}{{legend|#2750b0|State or district has a Cook PVI between D+5 and D+10}}{{legend|#668be1|State or district has a Cook PVI between D+2 and D+5}}{{legend|#a8c1fa|State or district has a Cook PVI between EVEN and D+2}}{{legend|#fbb7bb|State or district has a Cook PVI between EVEN and R+2}}{{legend|#f47178|State or district has a Cook PVI between R+2 and R+5}}{{legend|#e4252f|State or district has a Cook PVI between R+5 and R+10}}{{legend|#9e151e|State or district has a Cook PVI of R+10 or greater}}]] |
|||
disruption takes place? |
|||
<br> |
|||
newly formed alliances and composite position to compromise for you is amount to essentially not a dirty world well working word right now so lol! haha ... what? |
|||
man, well eh fr eh soaking now then like! |
|||
!!! GO DIRECTLY TO VISIT GERMAN SPANISH TAB AND WINDOW FIRST!!! |
|||
MAKE SURE ALL PAGES SAVED, INCLUDING MAIN, BOTH EDIT OPTIONS, HISTORY PAGE, ANY THING ELSE TO OTHER LINKS |
|||
MAKE SURE ALL BOOKMARKED |
|||
ARCHIVE ALL |
|||
GO TO HISTORY OF EACH |
|||
OVERWRITE File:US CPVI after 2016.svg |
|||
I have never been the most active of users, but I anticipate that I will contribute at a high rate in the future. |
|||
I have never been the most active of users on here, but I anticipate that I will contribute to the Commons at a high rate in the future. |
|||
Prior to that time, I edited Wikipedia on-and-off using an account that has now been lost. |
|||
For instance, I would like to fix the map and legend for [[:File:2016 battleground states.svg]]. It lists NV in the wrong category, despite being coloured correctly on the map itself. ME is coloured on the map, but nowhere to be found in the legend. The same can be said for AZ. Furthermore, CO is listed (and coloured) as a <2% state, but it should be in the <5% bucket instead. In addition, there is no reason for any of GA, OH, or IA to be included in this map if neither TX nor UT are. |
|||
I also feel like I made a fair number of useful and helpful improvements to existing images in advance of the permissions change last year (be it improving their visual effects, correcting outdated information, standardizing shading among various maps of the same category, or anything else), and would like the opportunity to continue further doing so freely. |
|||
clarity and concision: renaming to reduce confusion and make the title consistent with the other dozens of swing state maps within the same category |
|||
clarity and concision: renaming to make the title consistent with the other dozens of swing state maps within the same category; reducing confusion, as "map" is redundant (any image of swing states will necessarily have to be a map) |
|||
:as a near-exact (but somewhat outdated) duplicate of __.svg, by which it has now been replaced and superseded. |
|||
:_ contains a number of errors. For instance, New York, Michigan, Nevada, Maine, and others (???) are all coloured incorrectly on this map. It is also poorly cropped, with wonky margins. |
|||
:No pages on any of the existing wikis currently link to this map. In addition, none of the relevant pages would require (or even prefer) using a PNG file over an SVG, rendering it entirely redundant and therefore irrelevant as well. Even if they were to eventually need one, a higher-quality version could easily be created from the existing SVG file—one that is more up-to-date rather than being riddled with mistakes, and that has a transparent background, a superior colour scheme, greater dimensions, and nicer-looking borders. |
|||
Right, but is there any argument for keeping it? |
|||
Is it actually superior to _ in any way whatsoever? |
|||
In my personal view, |
|||
the existence of this file merely serves to cause greater confusion for editors, surrounding which version to use |
|||
and whether _ is |
|||
superior in any way (e.g. if it were to show newer figures than the SVG file does) at all |
|||
For the record, I am indeed of the opinion that Biden will outperform his eventual polling margin among younger voters, and underperform among the elderly once again—similar to my expectation that the racial voting gap will remain larger than they suggest, since surveys generally tend to get slightly ahead of themselves when it comes to demographic trends like these. |
|||
Of course, it's slightly likelier than not that Biden should be able close the gap a little bit among the core Democratic constituencies where he's currently struggling to maintain enthusiasm. As he consolidates the party's historical base back behind him after the GE campaign truly gets underway, at least a few undecided younger and non-white voters should come around once all is said and done. |
|||
However, the swings are indeed real, and I anticipate that [b]most[/b] of the movement currently reflected by the polls will materialize in the end. |
|||
https://en.wikipedia.org/enwiki/w/index.php?title=User:A_Red_Cherry/sandbox&oldid=1190971722 |
|||
https://en.wikipedia.org/enwiki/w/index.php?title=User:A_Red_Cherry/sandbox&oldid=1193718380 |
|||
https://en.wikipedia.org/enwiki/w/index.php?title=User:A_Red_Cherry/sandbox&oldid=1190857380 |
|||
[quote author=Virginiá link=topic=479962.msg9336095#msg9336095 date=1704077424 uid=13620] |
|||
If you hover over a map, you should see a little "edit this map" text block appear beneath the map interface. You can click on it, and it will be loaded into the editor. |
|||
[/quote] |
|||
[size=10pt][font=Segoe UI]Thanks. I was wondering, though, if this could be done with maps that have not been the subject of a forum post, but that we nonetheless have the code for. In those cases, will we have to publish it on a thread somewhere, in order to be able to access the "edit this map" button?[/font][/size] |
|||
[quote author=cherry mandarin link=topic=557933.msg9137807#msg9137807 date=1689796401 uid=37311] |
|||
The thing is, the PVI figures are (apparently) now "wrong" or "outdated", because (last year) the Cook Political Report supposedly changed the weighting that they gave to certain factors' consideration, shifting the balance to 75% in favour of the most recent presidential election and only 25% for the one before that (as opposed to the older, more even split of 50/50), for whatever reason. |
|||
I don't know if Virginia wishes to keep her current method of calculating things, or the new one that Cook uses nowadays—in fact, I've personally gone by the statewide margin relative to the national margin in terms of *total* votes, not just the 2-party share, which I think more accurately reflects where any given state stands, once you take third-party supporters into consideration too. |
|||
Either way, the PVI for all 5 CDs is, once again, still incorrect. You can let me know how you'd like them to be calculated (old or new boundaries), and I'll gladly do that for you. :) |
|||
[/quote] |
|||
[size=10pt][font=Segoe UI]For reference, here's the overview and run-down that I wrote for the PVI issue at the time, as well as the minor glitch regarding the boldface highlighting of party names using the selection menu:[/font][/size] |
|||
[quote author=Virginiá link=topic=479962.msg9336095#msg9336095 date=1704077424 uid=13620] |
|||
I'll try and remember to take a look at the PVI issue. |
|||
[/quote] |
|||
[size=10pt][font=Segoe UI]Thanks Virginia! Happy New Year's to all Atlas posters![/font][/size] |
|||
[quote author=cherry mandarin link=topic=557933.msg9138785#msg9138785 date=1689869783 uid=37311] |
|||
[quote author=Virginiá link=topic=557933.msg9138734#msg9138734 date=1689865673 uid=13620] |
|||
I'm surprised this wasn't pointed out already (or was but I missed it). I did have the PVIs for the CDs in there but the code doesn't use it due to a bug. I'll have to fix that. Thanks for pointing it out. |
|||
[/quote] |
|||
No worries. Do you plan to use the new, "official" figures released by Cook in 2022, or the old version, according to the previous format (how you're currently calculating it)? And I assume the listed PVI for each state and CD will follow the *new* post-redistricting lines when a user selects the 2024 election, but the old boundaries and 2012/2016 PVI values upon switching to the corresponding 2020 option? Thank you for the helpful clarification. |
|||
[quote author=Virginiá link=topic=557933.msg9138734#msg9138734 date=1689865673 uid=13620] |
|||
I designed it to look like the forum & website, which is just 1990s-style minimal/utilitarian layout for everything. That list of radio controls |
|||
[/quote] |
|||
I'm not sure if you mean you made the interface back in the 1990s (which, granted, would explain a lot, haha) or you were simply trying to emulate it while updating it over the past few years. Either way, I assume you intended for the corresponding column name to light up regardless of the state's order in the alphabetical list? For instance, if I switch AL to the Republican side, the R will grow, but if I then move VA into the toss-up category, it will do the same thing to "T". However, "R" will still be bolded, which I think isn't what you wanted, since it now conflicts and clashes with the newly enlarged T in the other column. |
|||
[/quote] |
|||
https://talkelections.org/FORUM/index.php?action=post;topic=479962.50;num_replies=66 |
|||
[quote author=MT Treasurer link=topic=576024.msg9335417#msg9335417 date=1703999637 uid=12707] |
|||
There's little reason to believe AZ/GA will be particularly easy for the GOP to flip back or trend right relative to the popular vote. There's also very little reason to believe PA's long-term trend has suddenly reversed. |
|||
[/quote] |
|||
[size=10pt][font=Segoe UI]Don't you believe the data suggesting that Trump has made sizable gains among non-white and younger voters since 2020, whereas Biden has held up better, or even increased his support, among older and white voters? These trends would also fit neatly in line with what we should expect based on long-term economic, demographic, social, political, and cultural shifts.[/font][/size] |
|||
https://talkelections.org/FORUM/index.php?action=post;quote=9335417;topic=576024.0 |
|||
fixed shading of certain states; adjusted colour scheme for consistency; added further gradients for greater nuance; corrected padding and margins |
|||
in accordance with updated formula adjustments dating back to 2022 |
|||
MAKE GRAPH OF EV COUNT BY PV MARGIN (SET BAR at 270) in new polling worksheet, most recent elections (2012/2016/2020) |
|||
shift in states (trend from 2020 to polling average now) on tableau, matplotlib using |
|||
atan2 produces, yields undefined result? |
|||
<!-- Note to editors: this file is a '''static''' version of [[dynamic]] that reflects the results of the [[ United States presidential election| ]] and [[ United States presidential election| ]] elections. It should NOT be updated to the newest election results. |
|||
Both files are necessary, given that the static version is more suitable for usage on certain wiki pages, whereas the dynamic one is more appropriate for others. Keeping them will significantly reduce the demand for periodically uploading newer versions of key files, as well as preventing the need to constantly update or redirect links on up to hundreds of articles across Wikipedia. |
|||
In addition, this map uses the '''old''' (post-[[2010 United States census|2010 census]] [[redistricting]] cycle) boundaries for the congressional districts of Maine and Nebraska, whereas [[dynamic]] uses the new (post-[[2020 United States redistricting cycle|2020]]) ones. |
|||
Furthermore, the electoral-vote counts for each state on this map reflect the numbers as they stood at the time of the [[ United States presidential election| ]] and [[ United States presidential election| ]] elections, whereas [[dynamic]] uses the new, post-2020 [[United States congressional apportionment|reapportionment]] totals. --> |
|||
<!-- Note to editors: this file is a '''static''' version of [[dynamic]] that reflects the results of the [[ United States presidential election| ]] election. It should NOT be updated to the newest election results. |
|||
Both files are necessary, given that the static version is more suitable for usage on certain wiki pages, whereas the dynamic one is more appropriate for others. Keeping them will significantly reduce the demand for periodically uploading newer versions of key files, as well as preventing the need to constantly update or redirect links on up to hundreds of articles across Wikipedia. |
|||
In addition, this map uses the '''old''' (post-[[2010 United States census|2010 census]] [[redistricting]] cycle) boundaries for the congressional districts of Maine and Nebraska, whereas [[dynamic]] uses the new (post-[[2020 United States redistricting cycle|2020]]) ones. |
|||
Furthermore, the electoral-vote counts for each state on this map reflect the numbers as they stood at the time of the [[ United States presidential election| ]] election, whereas [[dynamic]] uses the new, post-2020 [[United States congressional apportionment|reapportionment]] totals. --> |
|||
<!-- Note to editors: this file is a '''dynamic''' version of [[static]] that should be updated after every election. |
|||
Both files are necessary, given that the dynamic version is more suitable for usage on certain wiki pages, whereas the static one is more appropriate for others. Keeping them will significantly reduce the demand for periodically uploading newer versions of key files, as well as preventing the need to constantly update or redirect links on up to hundreds of articles across Wikipedia. |
|||
In addition, this map uses the '''new''' (post-[[2020 United States census|2020 census]] [[redistricting]] cycle) boundaries for the congressional districts of Maine and Nebraska, whereas [[static]] uses the old (post-[[2010 United States redistricting cycle|2010]]) ones. |
|||
Furthermore, the electoral-vote counts for each state on this map reflect the numbers according to the new [[United States congressional apportionment|reapportionment]] plan that will be in place during the 2020s decade, whereas [[static]] uses the numbers from the [[ United States presidential election| ]] and [[ United States presidential election| ]] elections. --> |
|||
<!-- Note to editors: this file is a '''dynamic''' version of [[static]] that should be updated after every election. |
|||
Both files are necessary, given that the dynamic version is more suitable for usage on certain wiki pages, whereas the static one is more appropriate for others. Keeping them will significantly reduce the demand for periodically uploading newer versions of key files, as well as preventing the need to constantly update or redirect links on up to hundreds of articles across Wikipedia. |
|||
In addition, this map uses the '''new''' (post-[[2020 United States census|2020 census]] [[redistricting]] cycle) boundaries for the congressional districts of Maine and Nebraska, whereas [[static]] uses the old (post-[[2010 United States redistricting cycle|2010]]) ones. |
|||
Furthermore, the electoral-vote counts for each state on this map reflect the numbers according to the new [[United States congressional apportionment|reapportionment]] plan that will be in place during the 2020s decade, whereas [[static]] uses the numbers from the [[ United States presidential election| ]] election. --> |
|||
It bears mentioning, of course, that Nevada is one of the few perennial battlegrounds where polling has narrowly, yet consistently, tended to overestimate Republicans' chances rather than Democrats' over the past few cycles. As a result, it'd perhaps be wise to take its numbers with a grain of salt, at least in the run-up to Election Day itself. If Trump continues leading there by such gaudy margins, though, it's highly unlikely to matter even if he ends up underperforming his polling by a handful of points. |
|||
Unless there are strong arguments for doing otherwise, we should either follow the current naming convention, OR change it for all files to which said convention applies. Where both the existing and proposed conventions are equally suitable, I '''usually''' prefer the former option, since it involves a lot less work and is far likelier to succeed. There are some exceptions, however. |
|||
renaming requested to reduce confusion and achieve naming consistency among maps in the swing state category |
|||
In particular, the current file name could easily be confused with that of 2020 swing state map.svg. This file is the PNG version of 2020 swing states.svg, so I am proposing a name change in order to reflect this fact. |
|||
Furthermore, this change would streamline its name to be consistent with the other images from the same category (2012 swing states, 2016 swing states, swing states, etc.), while also distinguishing it from the separate "swing state map" group, to which it does not belong. |
|||
(''Originally posted 25 January 2024. Last edited as of the [insert date here].'')First written and published as of the |
|||
= State of the Race: 40 Weeks Out = |
|||
Can you believe there are only nine months (and spare change) left to go before Election Day? With the presidential field looking pretty much set for a rematch this time around, here are some of my thoughts a few major events in the campaign to date—along with their possible implications for the ultimate result. |
|||
[list] |
|||
[item]Trump led Haley by more than 24% in the final (post-DeSantis dropout) New Hampshire primary polling average. Thus far, it looks like he's only on track to win by about 11 points. That's actually a massive miss![/item] |
|||
[item]Although the Iowa polls were pretty accurate, they followed the same general trend of overstating his support while underestimating each of his main opponents. In fact, Trump actually underperformed his election-eve average by more than DeSantis overperformed his![/item] |
|||
[item]Meanwhile, Trump's general-election polling lead just keeps on growing. As of this writing, he's currently up by 4.9% over Biden in an average of all seven surveys conducted within the past week. For what it's worth, over his entire history in the political arena, he's never enjoyed a larger lead in his polling average against an eventual general-election opponent. |
|||
I'd always thought Trump would only start consolidating reluctant Republican supporters behind him in large numbers closer to the time when both candidates receive their inevitable convention bumps. But after a bitter, divisive, and hard-fought primary campaign, it seems like Trump's already benefiting—however slightly—from increased unity from within his own party's ranks. could end up being his trump card to play in the election! |
|||
If these numbers continue to hold, don't be surprised if the Democratic clamouring for Biden to drop out eventually reaches a fever pitch on his own side of the aisle over the coming weeks and months, or potentially even grows into a deafening chorus at some point. As I'll explain below, though, I don't think this intra-party "challenge" will necessarily leave him in a weakened position on its own, or even pose a significant risk for his standing in his impending matchup against Trump. However, it certainly could speak to a number of more deeply-rooted issues that the Democratic Party currently faces writ large, chief among them its longer-term struggles to turn out many vital portions of its own base. Biden's current deficit arguably also exposes a few specific problems with his own candidacy—most notably his age and the lack of enthusiasm he inspires among Democratic partisans, as well as his stance on the Israel-Palestine war.[/item] |
|||
Putting on my pundit's hat now ... |
|||
[item]In 2016, he had [url=]his "outsider" magic[/url]. In 2020, he had incumbency. What will Trump have going for him this year? COVID-19? The "Big Steal"? His opponent's age? "Bidenomics"? |
|||
Right now, I just don't buy the case that he's the favourite to beat Biden in November. I've always had this race as a roughly 50-50 proposition, and I stand by my overall toss-up rating just as much today as I did three years ago. Although I consider this election to be a coin-flip, I'd still take the incumbent's chances over his challenger's if I had to choose, in line with the position I've maintained all along. |
|||
At this point, most people have grown sick and tired of Trump. By and large, the American public would like to move on from him. Of course, the same could be said for Biden, too. Faced with the binary choice, many voters are resigned to having to picking the "lesser" of two evils—voting [b]against[/b] the opposing party, rather than affirmatively in favour of the candidate they support on paper.[/item] |
|||
[item]By and large, Democrats have performed well in elections since 2020, whereas Republicans have generally tended to do poorly. I think this has more to do with Trump leaving office than Biden being the president now—after all, this phenomenon could still be observed during Trump's term, just to a slightly lesser extent. |
|||
Of course, the two times Trump was on the ballot were notable exceptions to this trend, and definitely the most consequential ones. If you were to ask for my take on whether Trump's tendency to overperform his pre-election polling is statistically significant or a mere coincidence, I'd lean toward the former option. |
|||
I've expounded on this topic [url=]elsewhere[/url] in the past, but to summarize, I just think the overall balance of evidence generally points in a favourable direction when it comes to the process of arriving at the above conclusion. Furthermore, these arguments are sufficiently numerous, strong, and varied (i.e. drawing from a wide and diverse range of sources) to make me at least relatively comfortable with saying that it is a sound one to land on. |
|||
Of course, this is an entirely separate matter, altogether removed from the question of how much we should attempt to account for these considerations, or factor them into our calculations for the 2024 race. We also don't yet know if they'll turn out to be meaningful indicators for 2024 in any way, or whether they might end up proving predictive of future results at all. |
|||
I do think there are credible arguments that any "shy Trump voter" effect we observed in the 2016 and 2020 elections could be greatly reduced or diminished, if not wiped out altogether, by the time Americans head to the polls this fall. On the other hand, many analysts made the same argument in the lead-up to 2020, yet he ended up outperforming his final numbers by a lot more than the first time he ran. |
|||
As I mentioned above, I do feel that Donald Trump has been able to create a cult of personality revolving around him, to the point where a substantial chunk of his most devoted followers are magnetically attracted to him on a personal level. Many of these supporters will only show up to cast their ballot when he's on it himself. They'll usually tick the box next to the names of other GOP candidates while they're at it, too—whether to try and make it easier for him to score legislative victories by receiving congressional approval for them, or out of simple and pure distaste for the Democratic Party. Larger GOP majorities in both chambers could also pave the way for him to quickly fulfill his campaign promises, implement his agenda as planned, pass his policy proposals, and achieve a number of important partisan objectives. In addition, he has repeatedly declared his intention to "drain the swamp" by changing the system from within, purging the federal bureaucracy, and getting rid of the "permanent governing class", and going after the "deep state". |
|||
However, these low-propensity Trump voters also tend to be sufficiently disenchanted and disillusioned that they generally won't turn out in down-ballot races—even the ones involving hardcore MAGA-type candidates—if they aren't getting an opportunity to directly vote for Trump at the same time. |
|||
This is also a large part of the reason why I've often stated that it's still much too early to proclaim Vivek Ramaswamy the frontrunner to capture the 2028 GOP presidential nod—a claim that I see a number of "terminally online" Twitter conservatives prematurely making. |
|||
Sure, his candidacy may very well end up covering many of the same overarching themes as Trump has emphasized over the years. In fact, Ramaswamy's 2024 campaign pulled from a handbook that might appear, at first glance, nearly identical to the one that Trump's been using. After all, he offered similar policy prescriptions, shared similar worldviews, sang a similar tune, hit similar notes, rang similar alarm bells, sounded similar warnings, preached similar messages, delivered a similar stump speech, and espoused similar messages on the campaign trail. |
|||
However, none of this implies that members of the MAGA movement are guaranteed to shift their support his way en masse once the next cycle rolls around by any means, especially if Trump refuses to provide Ramaswamy with his full-throated endorsement. A carbon copy of their leader isn't necessarily what these voters are after. |
|||
[/list] |
|||
Where do you think the general-election state of play currently stands? Feel free to weigh in and share your own opinions as well! |
|||
__ a quick tangential aside on the construction of electoral forecasting models __ |
|||
When discussing Trump's elections, it's worth noting that [url=]we're dealing with a very small sample size[/url] of just 2. Since we have such a limited trove of data to dig into, it can be especially hard for analysts to [url=]determine the precise set of priors we should ideally be adopting[/url] as we head into the midst of another campaign season. Furthermore, the low frequency at which American presidential elections take place poses a major inconvenience to our efforts to divine the 2024 results. This can make it extremely difficult, not to mention highly risky, to try and extrapolate future outcomes from past examples. |
|||
First and foremost as it relates to the upcoming election is the question of how much we should attempt to adjust for the "Trump bump"—in other words, his perceived penchant for outperforming his general-election polling figures. In fact, is it even worth trying to control for? The only thing we know for sure is that this is far from a settled matter, and it remains an open question that should continually return to forefront of the debate to some degree as the year goes along. |
|||
No matter what assumptions we make, or how we build them into our models, there's every chance we could end up being way off the mark. Nevertheless, it's essential for election forecasters to spend adequate time honing their craft, as the planning and conception steps of the process form the very basis upon which all other elements of their models will be built. Many common errors pertaining to their design frequently occur around this point, yet they go unnoticed a lot of the time. Since this "bedrock layer" essentially represents the foundational levels of a model, a few minor mistakes or poor decisions made here can quickly lead to a disproportionately harmful "cascade effect" that costs its creators valuable time or causes them significant headaches later on. Although smaller setbacks tend to happen more often at a model's execution stage, they can be identified and corrected using far simpler solutions, while usually also being much less consequential over the long run. |
|||
Election modellers should additionally be willing to go to great lengths to try and avoid drawing faulty or otherwise flawed conclusions. One way of achieving this goal involves taking pains to carefully select the best method and approach that most appropriately fits the specific political situation, era, region, population demographics, electoral system, level of government, and candidates in question. Without a robust and systematic strategy for tackling these unique attributes of the electorate they're working with, a model can barely stand on its own. Like a house made from crumply old playing cards, a gust of wind or even a slight breeze could cause it to fall over and collapse. |
|||
It is therefore critical to always take measures aimed at actively reducing the risk of disaster, or at least mitigating its worst effects, at every step along the way. They are particularly influential when smoothly and efficiently incorporated at their earliest available convenience, however. In fact, neglecting to pay proper attention to these minutiae at the model's very inception can easily spell doom for prognosticators. This is because it adversely impacts their likelihood of getting a bunch of minor yet important details correct, thereby rendering it virtually impossible for them to accurately nail the final results. |
|||
Meanwhile, as we continue to look forward to the segments of the campaign that have yet to unfold, let us also take the time to duly reflect upon a bevy of relevant lessons learned from previous cycles. Moreover, keep in mind that the key aspects underpinning the race ahead of us are very much expected to constantly remain fluid, flexible, and subject to change the whole way through. |
|||
At this early stage in the campaign, you really can't know for sure what the "right" or "wrong" answers are going to be. We're truly in the dark here, and there's no way to tell the direction in which the polls will miss, if at all. |
|||
Nate Silver, for one, is on the record as a vocal opponent of pundits trying to "out-guess" the polls. Although there are some exceptions, I generally agree with this principle. |
|||
For instance, remember how a number of pollsters artificially padded the Conservative Party's margin, or in some cases even subtracted from Labour's total, when discussing their surveys during the United Kingdom's 2017 general-election campaign? Survation even declined to publish their final report, believing that their estimates for Jeremy Corbyn's party were so rosy as to be inconceivable. In the end, |
|||
though, it's Labour that ended up beating their polls by quite a bit, whereas Theresa May's Tories fell short and actually lost their majority. |
|||
font size |
|||
pliable to his needs and bending to his will |
|||
help him more smoothly |
|||
chasm in between |
|||
base-responsive GOP politicians |
|||
infrequently back Republican candidates |
|||
voting methods |
|||
sow distrust in elections systems |
|||
more conciliatory tone, takes a less opposed or belligerent tone thusly |
|||
get a more inclined aligned naturally amenable Congress aid will bend to his will |
|||
among likely voters |
|||
in line with expectations! |
|||
as noted earlier |
|||
a.k.a. made it so that |
|||
pull |
|||
in other words |
|||
what also bears repeating is the fact that |
|||
I would like to point out |
|||
keep in mind |
|||
at all, carries, yet largely |
|||
reasonably, rationally, or realistically speaking, however? |
|||
craft message that can appeal to a broad swathe slice of the electorate |
|||
to define himself in the eyes of voters! |
Revision as of 09:46, 28 January 2024
!!! GO DIRECTLY TO VISIT GERMAN SPANISH TAB AND WINDOW FIRST!!!
MAKE SURE ALL PAGES SAVED, INCLUDING MAIN, BOTH EDIT OPTIONS, HISTORY PAGE, ANY THING ELSE TO OTHER LINKS
MAKE SURE ALL BOOKMARKED
ARCHIVE ALL
GO TO HISTORY OF EACH
OVERWRITE File:US CPVI after 2016.svg
I have never been the most active of users, but I anticipate that I will contribute at a high rate in the future.
I have never been the most active of users on here, but I anticipate that I will contribute to the Commons at a high rate in the future.
Prior to that time, I edited Wikipedia on-and-off using an account that has now been lost.
For instance, I would like to fix the map and legend for File:2016 battleground states.svg. It lists NV in the wrong category, despite being coloured correctly on the map itself. ME is coloured on the map, but nowhere to be found in the legend. The same can be said for AZ. Furthermore, CO is listed (and coloured) as a <2% state, but it should be in the <5% bucket instead. In addition, there is no reason for any of GA, OH, or IA to be included in this map if neither TX nor UT are.
I also feel like I made a fair number of useful and helpful improvements to existing images in advance of the permissions change last year (be it improving their visual effects, correcting outdated information, standardizing shading among various maps of the same category, or anything else), and would like the opportunity to continue further doing so freely.
clarity and concision: renaming to reduce confusion and make the title consistent with the other dozens of swing state maps within the same category
clarity and concision: renaming to make the title consistent with the other dozens of swing state maps within the same category; reducing confusion, as "map" is redundant (any image of swing states will necessarily have to be a map)
- as a near-exact (but somewhat outdated) duplicate of __.svg, by which it has now been replaced and superseded.
- _ contains a number of errors. For instance, New York, Michigan, Nevada, Maine, and others (???) are all coloured incorrectly on this map. It is also poorly cropped, with wonky margins.
- No pages on any of the existing wikis currently link to this map. In addition, none of the relevant pages would require (or even prefer) using a PNG file over an SVG, rendering it entirely redundant and therefore irrelevant as well. Even if they were to eventually need one, a higher-quality version could easily be created from the existing SVG file—one that is more up-to-date rather than being riddled with mistakes, and that has a transparent background, a superior colour scheme, greater dimensions, and nicer-looking borders.
Right, but is there any argument for keeping it?
Is it actually superior to _ in any way whatsoever?
In my personal view,
the existence of this file merely serves to cause greater confusion for editors, surrounding which version to use
and whether _ is
superior in any way (e.g. if it were to show newer figures than the SVG file does) at all
For the record, I am indeed of the opinion that Biden will outperform his eventual polling margin among younger voters, and underperform among the elderly once again—similar to my expectation that the racial voting gap will remain larger than they suggest, since surveys generally tend to get slightly ahead of themselves when it comes to demographic trends like these.
Of course, it's slightly likelier than not that Biden should be able close the gap a little bit among the core Democratic constituencies where he's currently struggling to maintain enthusiasm. As he consolidates the party's historical base back behind him after the GE campaign truly gets underway, at least a few undecided younger and non-white voters should come around once all is said and done.
However, the swings are indeed real, and I anticipate that [b]most[/b] of the movement currently reflected by the polls will materialize in the end.
https://en.wikipedia.org/enwiki/w/index.php?title=User:A_Red_Cherry/sandbox&oldid=1190971722
https://en.wikipedia.org/enwiki/w/index.php?title=User:A_Red_Cherry/sandbox&oldid=1193718380
https://en.wikipedia.org/enwiki/w/index.php?title=User:A_Red_Cherry/sandbox&oldid=1190857380
[quote author=Virginiá link=topic=479962.msg9336095#msg9336095 date=1704077424 uid=13620]
If you hover over a map, you should see a little "edit this map" text block appear beneath the map interface. You can click on it, and it will be loaded into the editor.
[/quote]
[size=10pt][font=Segoe UI]Thanks. I was wondering, though, if this could be done with maps that have not been the subject of a forum post, but that we nonetheless have the code for. In those cases, will we have to publish it on a thread somewhere, in order to be able to access the "edit this map" button?[/font][/size]
[quote author=cherry mandarin link=topic=557933.msg9137807#msg9137807 date=1689796401 uid=37311] The thing is, the PVI figures are (apparently) now "wrong" or "outdated", because (last year) the Cook Political Report supposedly changed the weighting that they gave to certain factors' consideration, shifting the balance to 75% in favour of the most recent presidential election and only 25% for the one before that (as opposed to the older, more even split of 50/50), for whatever reason.
I don't know if Virginia wishes to keep her current method of calculating things, or the new one that Cook uses nowadays—in fact, I've personally gone by the statewide margin relative to the national margin in terms of *total* votes, not just the 2-party share, which I think more accurately reflects where any given state stands, once you take third-party supporters into consideration too.
Either way, the PVI for all 5 CDs is, once again, still incorrect. You can let me know how you'd like them to be calculated (old or new boundaries), and I'll gladly do that for you. :) [/quote]
[size=10pt][font=Segoe UI]For reference, here's the overview and run-down that I wrote for the PVI issue at the time, as well as the minor glitch regarding the boldface highlighting of party names using the selection menu:[/font][/size]
[quote author=Virginiá link=topic=479962.msg9336095#msg9336095 date=1704077424 uid=13620] I'll try and remember to take a look at the PVI issue. [/quote]
[size=10pt][font=Segoe UI]Thanks Virginia! Happy New Year's to all Atlas posters![/font][/size]
[quote author=cherry mandarin link=topic=557933.msg9138785#msg9138785 date=1689869783 uid=37311] [quote author=Virginiá link=topic=557933.msg9138734#msg9138734 date=1689865673 uid=13620] I'm surprised this wasn't pointed out already (or was but I missed it). I did have the PVIs for the CDs in there but the code doesn't use it due to a bug. I'll have to fix that. Thanks for pointing it out. [/quote]
No worries. Do you plan to use the new, "official" figures released by Cook in 2022, or the old version, according to the previous format (how you're currently calculating it)? And I assume the listed PVI for each state and CD will follow the *new* post-redistricting lines when a user selects the 2024 election, but the old boundaries and 2012/2016 PVI values upon switching to the corresponding 2020 option? Thank you for the helpful clarification.
[quote author=Virginiá link=topic=557933.msg9138734#msg9138734 date=1689865673 uid=13620] I designed it to look like the forum & website, which is just 1990s-style minimal/utilitarian layout for everything. That list of radio controls [/quote]
I'm not sure if you mean you made the interface back in the 1990s (which, granted, would explain a lot, haha) or you were simply trying to emulate it while updating it over the past few years. Either way, I assume you intended for the corresponding column name to light up regardless of the state's order in the alphabetical list? For instance, if I switch AL to the Republican side, the R will grow, but if I then move VA into the toss-up category, it will do the same thing to "T". However, "R" will still be bolded, which I think isn't what you wanted, since it now conflicts and clashes with the newly enlarged T in the other column. [/quote]
https://talkelections.org/FORUM/index.php?action=post;topic=479962.50;num_replies=66
[quote author=MT Treasurer link=topic=576024.msg9335417#msg9335417 date=1703999637 uid=12707] There's little reason to believe AZ/GA will be particularly easy for the GOP to flip back or trend right relative to the popular vote. There's also very little reason to believe PA's long-term trend has suddenly reversed. [/quote]
[size=10pt][font=Segoe UI]Don't you believe the data suggesting that Trump has made sizable gains among non-white and younger voters since 2020, whereas Biden has held up better, or even increased his support, among older and white voters? These trends would also fit neatly in line with what we should expect based on long-term economic, demographic, social, political, and cultural shifts.[/font][/size]
https://talkelections.org/FORUM/index.php?action=post;quote=9335417;topic=576024.0
fixed shading of certain states; adjusted colour scheme for consistency; added further gradients for greater nuance; corrected padding and margins
in accordance with updated formula adjustments dating back to 2022
MAKE GRAPH OF EV COUNT BY PV MARGIN (SET BAR at 270) in new polling worksheet, most recent elections (2012/2016/2020)
shift in states (trend from 2020 to polling average now) on tableau, matplotlib using
atan2 produces, yields undefined result?
It bears mentioning, of course, that Nevada is one of the few perennial battlegrounds where polling has narrowly, yet consistently, tended to overestimate Republicans' chances rather than Democrats' over the past few cycles. As a result, it'd perhaps be wise to take its numbers with a grain of salt, at least in the run-up to Election Day itself. If Trump continues leading there by such gaudy margins, though, it's highly unlikely to matter even if he ends up underperforming his polling by a handful of points.
Unless there are strong arguments for doing otherwise, we should either follow the current naming convention, OR change it for all files to which said convention applies. Where both the existing and proposed conventions are equally suitable, I usually prefer the former option, since it involves a lot less work and is far likelier to succeed. There are some exceptions, however.
renaming requested to reduce confusion and achieve naming consistency among maps in the swing state category
In particular, the current file name could easily be confused with that of 2020 swing state map.svg. This file is the PNG version of 2020 swing states.svg, so I am proposing a name change in order to reflect this fact.
Furthermore, this change would streamline its name to be consistent with the other images from the same category (2012 swing states, 2016 swing states, swing states, etc.), while also distinguishing it from the separate "swing state map" group, to which it does not belong.
(Originally posted 25 January 2024. Last edited as of the [insert date here].)First written and published as of the
State of the Race: 40 Weeks Out
Can you believe there are only nine months (and spare change) left to go before Election Day? With the presidential field looking pretty much set for a rematch this time around, here are some of my thoughts a few major events in the campaign to date—along with their possible implications for the ultimate result.
[list] [item]Trump led Haley by more than 24% in the final (post-DeSantis dropout) New Hampshire primary polling average. Thus far, it looks like he's only on track to win by about 11 points. That's actually a massive miss![/item]
[item]Although the Iowa polls were pretty accurate, they followed the same general trend of overstating his support while underestimating each of his main opponents. In fact, Trump actually underperformed his election-eve average by more than DeSantis overperformed his![/item]
[item]Meanwhile, Trump's general-election polling lead just keeps on growing. As of this writing, he's currently up by 4.9% over Biden in an average of all seven surveys conducted within the past week. For what it's worth, over his entire history in the political arena, he's never enjoyed a larger lead in his polling average against an eventual general-election opponent.
I'd always thought Trump would only start consolidating reluctant Republican supporters behind him in large numbers closer to the time when both candidates receive their inevitable convention bumps. But after a bitter, divisive, and hard-fought primary campaign, it seems like Trump's already benefiting—however slightly—from increased unity from within his own party's ranks. could end up being his trump card to play in the election!
If these numbers continue to hold, don't be surprised if the Democratic clamouring for Biden to drop out eventually reaches a fever pitch on his own side of the aisle over the coming weeks and months, or potentially even grows into a deafening chorus at some point. As I'll explain below, though, I don't think this intra-party "challenge" will necessarily leave him in a weakened position on its own, or even pose a significant risk for his standing in his impending matchup against Trump. However, it certainly could speak to a number of more deeply-rooted issues that the Democratic Party currently faces writ large, chief among them its longer-term struggles to turn out many vital portions of its own base. Biden's current deficit arguably also exposes a few specific problems with his own candidacy—most notably his age and the lack of enthusiasm he inspires among Democratic partisans, as well as his stance on the Israel-Palestine war.[/item]
Putting on my pundit's hat now ...
[item]In 2016, he had [url=]his "outsider" magic[/url]. In 2020, he had incumbency. What will Trump have going for him this year? COVID-19? The "Big Steal"? His opponent's age? "Bidenomics"?
Right now, I just don't buy the case that he's the favourite to beat Biden in November. I've always had this race as a roughly 50-50 proposition, and I stand by my overall toss-up rating just as much today as I did three years ago. Although I consider this election to be a coin-flip, I'd still take the incumbent's chances over his challenger's if I had to choose, in line with the position I've maintained all along.
At this point, most people have grown sick and tired of Trump. By and large, the American public would like to move on from him. Of course, the same could be said for Biden, too. Faced with the binary choice, many voters are resigned to having to picking the "lesser" of two evils—voting [b]against[/b] the opposing party, rather than affirmatively in favour of the candidate they support on paper.[/item]
[item]By and large, Democrats have performed well in elections since 2020, whereas Republicans have generally tended to do poorly. I think this has more to do with Trump leaving office than Biden being the president now—after all, this phenomenon could still be observed during Trump's term, just to a slightly lesser extent.
Of course, the two times Trump was on the ballot were notable exceptions to this trend, and definitely the most consequential ones. If you were to ask for my take on whether Trump's tendency to overperform his pre-election polling is statistically significant or a mere coincidence, I'd lean toward the former option.
I've expounded on this topic [url=]elsewhere[/url] in the past, but to summarize, I just think the overall balance of evidence generally points in a favourable direction when it comes to the process of arriving at the above conclusion. Furthermore, these arguments are sufficiently numerous, strong, and varied (i.e. drawing from a wide and diverse range of sources) to make me at least relatively comfortable with saying that it is a sound one to land on.
Of course, this is an entirely separate matter, altogether removed from the question of how much we should attempt to account for these considerations, or factor them into our calculations for the 2024 race. We also don't yet know if they'll turn out to be meaningful indicators for 2024 in any way, or whether they might end up proving predictive of future results at all.
I do think there are credible arguments that any "shy Trump voter" effect we observed in the 2016 and 2020 elections could be greatly reduced or diminished, if not wiped out altogether, by the time Americans head to the polls this fall. On the other hand, many analysts made the same argument in the lead-up to 2020, yet he ended up outperforming his final numbers by a lot more than the first time he ran.
As I mentioned above, I do feel that Donald Trump has been able to create a cult of personality revolving around him, to the point where a substantial chunk of his most devoted followers are magnetically attracted to him on a personal level. Many of these supporters will only show up to cast their ballot when he's on it himself. They'll usually tick the box next to the names of other GOP candidates while they're at it, too—whether to try and make it easier for him to score legislative victories by receiving congressional approval for them, or out of simple and pure distaste for the Democratic Party. Larger GOP majorities in both chambers could also pave the way for him to quickly fulfill his campaign promises, implement his agenda as planned, pass his policy proposals, and achieve a number of important partisan objectives. In addition, he has repeatedly declared his intention to "drain the swamp" by changing the system from within, purging the federal bureaucracy, and getting rid of the "permanent governing class", and going after the "deep state".
However, these low-propensity Trump voters also tend to be sufficiently disenchanted and disillusioned that they generally won't turn out in down-ballot races—even the ones involving hardcore MAGA-type candidates—if they aren't getting an opportunity to directly vote for Trump at the same time.
This is also a large part of the reason why I've often stated that it's still much too early to proclaim Vivek Ramaswamy the frontrunner to capture the 2028 GOP presidential nod—a claim that I see a number of "terminally online" Twitter conservatives prematurely making.
Sure, his candidacy may very well end up covering many of the same overarching themes as Trump has emphasized over the years. In fact, Ramaswamy's 2024 campaign pulled from a handbook that might appear, at first glance, nearly identical to the one that Trump's been using. After all, he offered similar policy prescriptions, shared similar worldviews, sang a similar tune, hit similar notes, rang similar alarm bells, sounded similar warnings, preached similar messages, delivered a similar stump speech, and espoused similar messages on the campaign trail.
However, none of this implies that members of the MAGA movement are guaranteed to shift their support his way en masse once the next cycle rolls around by any means, especially if Trump refuses to provide Ramaswamy with his full-throated endorsement. A carbon copy of their leader isn't necessarily what these voters are after. [/list]
Where do you think the general-election state of play currently stands? Feel free to weigh in and share your own opinions as well!
__ a quick tangential aside on the construction of electoral forecasting models __
When discussing Trump's elections, it's worth noting that [url=]we're dealing with a very small sample size[/url] of just 2. Since we have such a limited trove of data to dig into, it can be especially hard for analysts to [url=]determine the precise set of priors we should ideally be adopting[/url] as we head into the midst of another campaign season. Furthermore, the low frequency at which American presidential elections take place poses a major inconvenience to our efforts to divine the 2024 results. This can make it extremely difficult, not to mention highly risky, to try and extrapolate future outcomes from past examples.
First and foremost as it relates to the upcoming election is the question of how much we should attempt to adjust for the "Trump bump"—in other words, his perceived penchant for outperforming his general-election polling figures. In fact, is it even worth trying to control for? The only thing we know for sure is that this is far from a settled matter, and it remains an open question that should continually return to forefront of the debate to some degree as the year goes along.
No matter what assumptions we make, or how we build them into our models, there's every chance we could end up being way off the mark. Nevertheless, it's essential for election forecasters to spend adequate time honing their craft, as the planning and conception steps of the process form the very basis upon which all other elements of their models will be built. Many common errors pertaining to their design frequently occur around this point, yet they go unnoticed a lot of the time. Since this "bedrock layer" essentially represents the foundational levels of a model, a few minor mistakes or poor decisions made here can quickly lead to a disproportionately harmful "cascade effect" that costs its creators valuable time or causes them significant headaches later on. Although smaller setbacks tend to happen more often at a model's execution stage, they can be identified and corrected using far simpler solutions, while usually also being much less consequential over the long run.
Election modellers should additionally be willing to go to great lengths to try and avoid drawing faulty or otherwise flawed conclusions. One way of achieving this goal involves taking pains to carefully select the best method and approach that most appropriately fits the specific political situation, era, region, population demographics, electoral system, level of government, and candidates in question. Without a robust and systematic strategy for tackling these unique attributes of the electorate they're working with, a model can barely stand on its own. Like a house made from crumply old playing cards, a gust of wind or even a slight breeze could cause it to fall over and collapse.
It is therefore critical to always take measures aimed at actively reducing the risk of disaster, or at least mitigating its worst effects, at every step along the way. They are particularly influential when smoothly and efficiently incorporated at their earliest available convenience, however. In fact, neglecting to pay proper attention to these minutiae at the model's very inception can easily spell doom for prognosticators. This is because it adversely impacts their likelihood of getting a bunch of minor yet important details correct, thereby rendering it virtually impossible for them to accurately nail the final results.
Meanwhile, as we continue to look forward to the segments of the campaign that have yet to unfold, let us also take the time to duly reflect upon a bevy of relevant lessons learned from previous cycles. Moreover, keep in mind that the key aspects underpinning the race ahead of us are very much expected to constantly remain fluid, flexible, and subject to change the whole way through.
At this early stage in the campaign, you really can't know for sure what the "right" or "wrong" answers are going to be. We're truly in the dark here, and there's no way to tell the direction in which the polls will miss, if at all.
Nate Silver, for one, is on the record as a vocal opponent of pundits trying to "out-guess" the polls. Although there are some exceptions, I generally agree with this principle.
For instance, remember how a number of pollsters artificially padded the Conservative Party's margin, or in some cases even subtracted from Labour's total, when discussing their surveys during the United Kingdom's 2017 general-election campaign? Survation even declined to publish their final report, believing that their estimates for Jeremy Corbyn's party were so rosy as to be inconceivable. In the end, though, it's Labour that ended up beating their polls by quite a bit, whereas Theresa May's Tories fell short and actually lost their majority.
font size
pliable to his needs and bending to his will help him more smoothly chasm in between base-responsive GOP politicians infrequently back Republican candidates
voting methods sow distrust in elections systems more conciliatory tone, takes a less opposed or belligerent tone thusly get a more inclined aligned naturally amenable Congress aid will bend to his will among likely voters in line with expectations! as noted earlier a.k.a. made it so that pull in other words
what also bears repeating is the fact that I would like to point out keep in mind at all, carries, yet largely reasonably, rationally, or realistically speaking, however?
craft message that can appeal to a broad swathe slice of the electorate to define himself in the eyes of voters!