Jump to content
The Political Lounge

VCCZAR's 2022 Predictions


vcczar
 Share

Recommended Posts

As many of you know, I created an algorithm that accurately predicted the 2020 Presidential Election, excluding ME-2. 

I'm now attempting to predict the 2022 Midterms. This algorithm, like the presidential one, will be updated until Election Day. I must say I am much less confident in my ability to predict midterms. If the election were today, this is what the algorithm is showing.

Note: This is all from the Democrat perspective.

Bold marks where my algorithm selects a different winner from most polls:

Dem Race Win? Algorithm
Murray (WA) D 11.86666667
Cortez Masto (NV) D 1.7
Kelly (AZ) D 3.6
Bennet (CO) D 8.1
Busch Valentine (MO) R -12.66666667
Franken (IA) R -8.666666667
Barnes (WI) R -0.1333333333
Ryan (OH) R -4.766666667
Demings (FL) R -3.433333333
Warnock (GA) R -0.4
Beasley (NC) R -1.433333333
Fetterman (PA) D 2.9
Hassan (NH) D 3.4
Kotek (OR) D 2.366666667
Hobbs (AZ) R -0.8666666667
Lujan Grisham (NM) D 5.566666667
Polis (CO) D 7.466666667
Kelly (KS) R -3.066666667
O'Rourke (TX) R -7.7
Walz (MN) D 3.5
Whaley (OH) R -5.533333333
Evers (WI) D 1
Crist (FL) R -8.766666667
Abrams (GA) R -6.733333333
Cunningham (SC) R -12.53333333
Shapiro (PA) D 2.866666667
Lamont (CT) D 5.833333333
Mills (ME) D 4.633333333
Whitmer (MI) D 4.766666667
Sisolak (NV) R -0.2333333333

 

 

 

  • Like 2
Link to comment
Share on other sites

My US House Predictions

I can't think of a way for my alogorithm to work for the US House, so I'm just going to say my prediction for House seats is, 

1) If my algorithm predictions lean more Democrats than most pollsters, and most pollsters are still predicting a narrow GOP takeover, then I'll say Democrats keep the House. Generally, my algorithm runs more skeptical, which is probably why it got the 2020 election. 

Link to comment
Share on other sites

On 9/4/2022 at 1:13 AM, vcczar said:

As many of you know, I created an algorithm that accurately predicted the 2020 Presidential Election, excluding ME-2. 

I'm now attempting to predict the 2022 Midterms. This algorithm, like the presidential one, will be updated until Election Day. I must say I am much less confident in my ability to predict midterms. If the election were today, this is what the algorithm is showing.

Note: This is all from the Democrat perspective.

Bold marks where my algorithm selects a different winner from most polls:

 

Dem Race Win? Algorithm
Murray (WA) D 11.86666667
Cortez Masto (NV) D 1.7
Kelly (AZ) D 3.6
Bennet (CO) D 8.1
Busch Valentine (MO) R -12.66666667
Franken (IA) R -8.666666667
Barnes (WI) R -0.1333333333
Ryan (OH) R -4.766666667
Demings (FL) R -3.433333333
Warnock (GA) R -0.4
Beasley (NC) R -1.433333333
Fetterman (PA) D 2.9
Hassan (NH) D 3.4
Kotek (OR) D 2.366666667
Hobbs (AZ) R -0.8666666667
Lujan Grisham (NM) D 5.566666667
Polis (CO) D 7.466666667
Kelly (KS) R -3.066666667
O'Rourke (TX) R -7.7
Walz (MN) D 3.5
Whaley (OH) R -5.533333333
Evers (WI) D 1
Crist (FL) R -8.766666667
Abrams (GA) R -6.733333333
Cunningham (SC) R -12.53333333
Shapiro (PA) D 2.866666667
Lamont (CT) D 5.833333333
Mills (ME) D 4.633333333
Whitmer (MI) D 4.766666667
Sisolak (NV) R -0.2333333333

 

 

 

 

My prediction was worse in Winning states, but better in voting shares, if you remember. But i think both of us did a good job.

Image

 

Edited by Entrecampos
Link to comment
Share on other sites

1 hour ago, Entrecampos said:

My prediction was worse in Winning states, but better in voting shares, if you remember. But i think both of us did a good job.

I wasn't really aiming for voting shares so much (since I measure on polling momentum). The margins are kind of more of a guide for me. In fact, I'm not going to present them when I put my final predictions in 2022 midterms or in 2024. 

  • Like 1
Link to comment
Share on other sites

My September 10th Update

Here's my second week of predictions. It should get more accurate each week since it is based on momentum. This week momentum really shifted to Democrats as 4 seats that were predicted to go R are now D (those in Blue are the flips). The algorithm doesn't attempt to measure popular vote, only which candidate is showing more strength + momentum going into election day. I also take into account state biases, incumbencies, etc. 

I don't really expect Barnes, Demings, and Beasley to hold these flips. I also think Warnock will flip to R. Hobbs might flip D. Next week's data will help and then the week after will probably show my best results. 

Dem Race Win? Algorithm
Murray (WA) D 9.533333333
Cortez Masto (NV) D 3.166666667
Kelly (AZ) D 4.1
Bennet (CO) D 8.1
Barnes (WI) D 0.06666666667
Ryan (OH) R -0.3666666667
Demings (FL) D 0.3
Warnock (GA) R -0.06666666667
Beasley (NC) D 0.9666666667
Fetterman (PA) D 4.633333333
Hassan (NH) D 3.6
Hobbs (AZ) R 0.28
Lujan Grisham (NM) D 4.846666667
Polis (CO) D 6.28
Kelly (KS) R -1.986666667
O'Rourke (TX) R -4.553333333
Walz (MN) D 4.913333333
Whaley (OH) R -4.453333333
Evers (WI) D 0.7466666667
Crist (FL) R -4.82
Abrams (GA) R -5.52
Shapiro (PA) D 4.213333333
Mills (ME) D 5.713333333
Whitmer (MI) D 5.646666667
Sisolak (NV) D 0.18
Link to comment
Share on other sites

My September 19 Update:

Changes are highlighted. GOP flips three in my forecast, while Democrats flip one. The algorithm # is not a popular vote estimate, it is just an estimate of how strongly this race is going for or against the Democratic candidate based on polls, momentum, biases, incumbencies, etc. I also take into account that shared ticket could help or hurt a nominee. For this reason, Crist, Warnock, and Ryan are particular undermined even when they poll well. Colorado and Washington is amost not even close enough to keep in my updates, as I'm only keeping track of states that both parties have a good shot at winning. 

Dem Race State Win? Algorithm
Kelly (AZ) AZ D 5.7
Hobbs (AZ) AZ D 3.08
Bennet (CO) CO D 8.2
Polis (CO) CO D 8.413333333
Demings (FL) FL D -3.133333333
Crist (FL) FL R -6.253333333
Warnock (GA) GA D 1.533333333
Abrams (GA) GA R -3.753333333
Kelly (KS) KS R -1.986666667
Mills (ME) ME D 5.713333333
Whitmer (MI) MI D 5.746666667
Walz (MN) MN D 5.413333333
Beasley (NC) NC D -0.8
Hassan (NH) NH D 7.133333333
Lujan Grisham (NM) NM D 7.913333333
Cortez Masto (NV) NV D 1.866666667
Sisolak (NV) NV D 1.313333333
Ryan (OH) OH R -4.133333333
Whaley (OH) OH R -8.386666667
Fetterman (PA) PA D 5.033333333
Shapiro (PA) PA D 5.246666667
O'Rourke (TX) TX R -4.253333333
Murray (WA) WA D 10.7
Barnes (WI) WI D 0.7333333333
Evers (WI) WI D 2.08
Link to comment
Share on other sites

Here's 9/24 update. Once we get two weeks away, I'll probably have a daily update, like I did with the Pres Election. 

This week shows only one change: Barnes is losing in Wisconsin. Overall, there was a slight shift to the GOP but not in every state. 

At this point, I think the only tossups are both AZ races, FL senate race, GA senate, KS gov, NC sen, both NV, both WI. 

I think DeWine's support is going to make it impossible for Ryan to win in OH. 

I'm almost considering taking CO off this list because both Bennet and Polis now seem certain to win. Same with Murray in WA. I keep Whaley on the list mainly to help judge Ryan since ballot sharing races impact one another and I factor that in. If the Whaley-DeWine race wasn't on the ballot, then Ryan could win. ME and NH races seem like they're probably out of GOP reach

Note: Remember the algorithm isn't a popular vote margin predictor, it's just it's number showing the strenght of the Democrat in that state against the Republican. 

 

Dem Race State Win? Algorithm
Kelly (AZ) AZ D 3.7
Hobbs (AZ) AZ D 0.4966666667
Bennet (CO) CO D 9.533333333
Polis (CO) CO D 11.08
Demings (FL) FL R -2.8
Crist (FL) FL R -6.17
Warnock (GA) GA D 0.95
Abrams (GA) GA R -5.253333333
Kelly (KS) KS R -1.653333333
Mills (ME) ME D 7.413333333
Whitmer (MI) MI D 6.163333333
Walz (MN) MN D 5.746666667
Beasley (NC) NC R -0.9666666667
Hassan (NH) NH D 7.116666667
Lujan Grisham (NM) NM D 7.246666667
Cortez Masto (NV) NV D 0.6166666667
Sisolak (NV) NV D 0.7466666667
Ryan (OH) OH R -3.716666667
Whaley (OH) OH R -12.00333333
Fetterman (PA) PA D 4.783333333
Shapiro (PA) PA D 4.913333333
O'Rourke (TX) TX R -4.836666667
Murray (WA) WA D 10.7
Barnes (WI) WI R -2.05
Evers (WI) WI D 0.6633333333
Link to comment
Share on other sites

8 hours ago, vcczar said:

Here's 9/24 update. Once we get two weeks away, I'll probably have a daily update, like I did with the Pres Election. 

This week shows only one change: Barnes is losing in Wisconsin. Overall, there was a slight shift to the GOP but not in every state. 

At this point, I think the only tossups are both AZ races, FL senate race, GA senate, KS gov, NC sen, both NV, both WI. 

I think DeWine's support is going to make it impossible for Ryan to win in OH. 

I'm almost considering taking CO off this list because both Bennet and Polis now seem certain to win. Same with Murray in WA. I keep Whaley on the list mainly to help judge Ryan since ballot sharing races impact one another and I factor that in. If the Whaley-DeWine race wasn't on the ballot, then Ryan could win. ME and NH races seem like they're probably out of GOP reach

Note: Remember the algorithm isn't a popular vote margin predictor, it's just it's number showing the strenght of the Democrat in that state against the Republican. 

 

 

Dem Race State Win? Algorithm
Kelly (AZ) AZ D 3.7
Hobbs (AZ) AZ D 0.4966666667
Bennet (CO) CO D 9.533333333
Polis (CO) CO D 11.08
Demings (FL) FL R -2.8
Crist (FL) FL R -6.17
Warnock (GA) GA D 0.95
Abrams (GA) GA R -5.253333333
Kelly (KS) KS R -1.653333333
Mills (ME) ME D 7.413333333
Whitmer (MI) MI D 6.163333333
Walz (MN) MN D 5.746666667
Beasley (NC) NC R -0.9666666667
Hassan (NH) NH D 7.116666667
Lujan Grisham (NM) NM D 7.246666667
Cortez Masto (NV) NV D 0.6166666667
Sisolak (NV) NV D 0.7466666667
Ryan (OH) OH R -3.716666667
Whaley (OH) OH R -12.00333333
Fetterman (PA) PA D 4.783333333
Shapiro (PA) PA D 4.913333333
O'Rourke (TX) TX R -4.836666667
Murray (WA) WA D 10.7
Barnes (WI) WI R -2.05
Evers (WI) WI D 0.6633333333

What does this lead to for the senate? 51-49 R?

Link to comment
Share on other sites

Yeah, so I'm going to create a way to predict who will control the US House by just focusing on specific House Seats. 

Currently, Dem chances for keeping the house are about 30%, which isn't good, but that's a better % than pundits gave Trump of winning in 2016 (about 25%). That is, it's unlikely but not improbable that they keep the house. Meanwhile, Democrats have about a 70% chance of keeping the Senate. That is, the chance for the GOP to take the Senate is identical to the chance that the Democrats keep the House. [Note: 538.com % used]

These are the 7 house races that are still in the range of competitive but are all likely to go GOP. Five Thirty Eight predicts that two of these going Blue will mean Dems probably keep the House. However, my algorithm is usually more cautious on Democratic turnout than pundit forecasters. I will say that 1 Democratic victory means GOP wins the House, and 2 Democratic victories means that it will be a tossup (and I will make a human, non-algorithm prediction on who wins the House on Election Day), and 3 Democratic victories below will mean Democrats keep the House. 

  • Axne (IA-3)
  • Vasquez (NM-2)
  • Wild (PA-7)
  • McLeod-Skinner (OR-5)
  • O'Halleran (AZ-2)
  • Smith (CA-27)
  • Riley (NY-19)

Let me repeat that I'm far less confident in my midterm algorithm than in my highly successful presidential election algorithm. Pres elections are a lot easier to predict. 

Link to comment
Share on other sites

23 minutes ago, vcczar said:

Yeah, so I'm going to create a way to predict who will control the US House by just focusing on specific House Seats. 

Currently, Dem chances for keeping the house are about 30%, which isn't good, but that's a better % than pundits gave Trump of winning in 2016 (about 25%). That is, it's unlikely but not improbable that they keep the house. Meanwhile, Democrats have about a 70% chance of keeping the Senate. That is, the chance for the GOP to take the Senate is identical to the chance that the Democrats keep the House. [Note: 538.com % used]

These are the 7 house races that are still in the range of competitive but are all likely to go GOP. Five Thirty Eight predicts that two of these going Blue will mean Dems probably keep the House. However, my algorithm is usually more cautious on Democratic turnout than pundit forecasters. I will say that 1 Democratic victory means GOP wins the House, and 2 Democratic victories means that it will be a tossup (and I will make a human, non-algorithm prediction on who wins the House on Election Day), and 3 Democratic victories below will mean Democrats keep the House. 

  • Axne (IA-3)
  • Vasquez (NM-2)
  • Wild (PA-7)
  • McLeod-Skinner (OR-5)
  • O'Halleran (AZ-2)
  • Smith (CA-27)
  • Riley (NY-19)

Let me repeat that I'm far less confident in my midterm algorithm than in my highly successful presidential election algorithm. Pres elections are a lot easier to predict. 

And as a follow up to the above. I just posted these in. Specific polls for house seats are very few, so prediction won't be precise. Neverthess, I currently show GOP winning al of these house races. I'm hoping more polls come out, but they're so infrequent as to be useless. The only inaccuracy in my 2020 prediction was ME-2. I had it going for Biden based on very limited polling. I fear my House predictor will be off, but I'll have one anyway. 

Link to comment
Share on other sites

Here's October 1 update. At some point this month, probably when we are about 1 month from Election day, I'll start daily calculations, but I'll only post when something shift to the other party. Democrats still seem better off than the GOP, despite the GOP likely to win the House. This might be the worst midterm performance ever for an out-of-power party facing an unpopular incumbent. It tells you a lot about how much MAGA is detested because it certainly isn't excitement about Biden or Pelosi and establishment Democrats:

1. House Elections: GOP takes the House. There are shockingly few House polls. It makes it virtually impossible to make any sort of estimate. However, the few polls we have do not show significant evidence that Democrats will score an upset at the moment. 

2. US Senate Elections: NV, NC, GA, and WI are all so close that they could flip a few times before election day. Barnes has clear momentum against him, unfortunately. Nevertheless, Democrats are looking at an improved Senate 51-49.:

5X7y9z.png

3. Governors: Democrats still gain, but they won't split the nation. When was the last time we had more Dem govs than GOP, by the way? I must say that the NV, AZ, and WI races are extremely close in my algorithm. KS has way too few polls to be accurate. These may go back and forth the next few weeks. The rest might be set. 

b4LA.png

 

 

  • Like 2
Link to comment
Share on other sites

13 minutes ago, ShortKing said:

I do not have an algorithm or anything, this is based on absolutely nothing beyond vibes, but I kinda feel like Laura Kelly has a better shot than Katie Hobbs

There is about 1 KS Gov poll per month, and not by the same company, so it makes it almost impossible to estimate. However, the most recent poll was an A-grade poll (using 538's grades for polls), and it has +2 for Kelly, although she's losing or is even in the other polls. 

Link to comment
Share on other sites

We're about a month from Midterms, so I'm going to start posting my forecast anytime something flips. Some things to know:

  • I am expanding the number of Gov and Sen Races I'm analyzing because some seem like they may be closer than had been the case earlier. 
  • I am not going to analyze the US House seats until election day. There's just not enough polling for House races, so I will make a gut decision, rather than let my algorithm pick. 
  • I may actually make two predictions on election day. If I disagree with my algorithm, I'll say where I disagree with it. 
Link to comment
Share on other sites

10/7 update: My Forecast now has AZ Gov going GOP

Senate: 51-49, D

Govs: 23-27, R

The algorithm shows Warnock-GA and Cortez-Masto-NV almost at an exact tie. I've NH Gov because the NH Sen race will be close-ish, although NH Gov will not be close, but it helps in calculating. I've done the same with Wyden-OR because the OR Gov race is now showing a very close race. 

Dem Race State Win?
Kelly (AZ) AZ D
Hobbs (AZ) AZ R
Bennet (CO) CO D
Polis (CO) CO D
Demings (FL) FL R
Crist (FL) FL R
Warnock (GA) GA D
Abrams (GA) GA R
Kelly (KS) KS R
Mills (ME) ME D
Whitmer (MI) MI D
Walz (MN) MN D
Beasley (NC) NC R
Hassan (NH) NH D
Sherman (NH) NH R
Lujan Grisham (NM) NM D
Cortez Masto (NV) NV D
Sisolak (NV) NV D
Ryan (OH) OH R
Whaley (OH) OH R
Kotek (OR) OR D
Wyden (OR) OR D
Fetterman (PA) PA D
Shapiro (PA) PA D
O'Rourke (TX) TX R
Murray (WA) WA D
Barnes (WI) WI R
Evers (WI) WI D
Link to comment
Share on other sites

5 hours ago, vcczar said:

10/7 update: My Forecast now has AZ Gov going GOP

Senate: 50-50, D w/ VP

Govs: 23-27, R

The algorithm shows Warnock-GA and Cortez-Masto-NV almost at an exact tie. I've NH Gov because the NH Sen race will be close-ish, although NH Gov will not be close, but it helps in calculating. I've done the same with Wyden-OR because the OR Gov race is now showing a very close race. 

 

Dem Race State Win?
Kelly (AZ) AZ D
Hobbs (AZ) AZ R
Bennet (CO) CO D
Polis (CO) CO D
Demings (FL) FL R
Crist (FL) FL R
Warnock (GA) GA R
Abrams (GA) GA R
Kelly (KS) KS R
Mills (ME) ME D
Whitmer (MI) MI D
Walz (MN) MN D
Beasley (NC) NC R
Hassan (NH) NH D
Sherman (NH) NH R
Lujan Grisham (NM) NM D
Cortez Masto (NV) NV D
Sisolak (NV) NV D
Ryan (OH) OH R
Whaley (OH) OH R
Kotek (OR) OR D
Wyden (OR) OR D
Fetterman (PA) PA D
Shapiro (PA) PA D
O'Rourke (TX) TX R
Murray (WA) WA D
Barnes (WI) WI R
Evers (WI) WI D

Update 10/7 again! 

See the post above. I realized I forgot to update the generic congress polls. This flipped Warnock's seat from D to R as the generic congress has momentum towards the GOP right now.

Link to comment
Share on other sites

On 10/4/2022 at 6:22 PM, Dobs said:

I'm going to be a contrarian and suggest that blue Ohio and red Pennsylvania will exist contemporaneously.

Ohioan-turned-Pennsylvanian here.  Not a chance.  Ohio grows redder every day, while Pennsylvania has a strong urban base.  A traditional Republican might have won in PA, but Dr. Oz is SUCH a stunningly bad candidate while Fetterman has that blue collar appeal.  I'd bet my house that PA goes Fetterman while Ryan loses in Ohio. 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...