When trying to make predictions I have to keep some facts in mind surrounding election polling and prognostication:

1. It has not happened that one party has won all races that were in the Toss-Up category in RCP poll averages.
2. Since 2014, polling has had an overall Democratic bias. 2010 and 2012 by contrast had Republican bias.
3. Larry Sabato’s Crystal Ball predictions have consistently underestimated Republican Senate performance since 2014, albeit slightly. He has never been more than two Senate seats wrong against Republicans.
4. Sabato has not been wrong on more than two Senate races.
5. Biden’s polling is underwater in most states with competitive races, in most of these cases in double-digits.
6. Nevada has a history of Republican bias in polling.
7. Ohio and Wisconsin have major history of Democratic bias in polling since 2014.
8. There have, since 2010, been at least two upsets in every election cycle in the Senate. An upset here being defined as the result goes a different way than RCP polling averages.

This election regarding governors may be a major backlash election for the length and severity of COVID restrictions imposed by certain Democratic governors. This may be felt strongest in Nevada for a reason I will bring up later. Although I have stuck with the two Senator upset, trends can change. I, for instance, believed that Trump would win in Florida and Ohio but I just had the hardest time conceiving that he would lose despite winning those states, even though national polling made me suspect that he would lose. I have tried my best here to control for biases in my analysis and differentiate between who I want to win and who I think will win. I have compiled a lot of past data that reflects polling biases in major races for these states. D + means that polls were biased in favor of Democrats compared to the actual result, and R + means polls were biased in favor of Republicans compared to the actual result. This year the polling makes me feel justified in predicting one less upset than usual and in continuing to believe that overall polling has a Democratic bias, but maybe there’s a change in the polling winds like there was in 2014 and maybe conventional wisdom here on the Senate will kick me in the teeth. I will consider my efforts against personal bias to be victorious if I get it right, if my errors do not consistently overestimate by party, and if I do consistently overestimate it is on the other side. We’ll see by the end of the day what happens, now won’t we? This will also be posted to my Substack: mikeholme.substack.com.

ARIZONA

Gubernatorial Poll Bias (2014): D +4.4
Presidential Poll Bias (2016): R +0.5
Senatorial Poll Bias (2016): D +2.9
Gubernatorial Poll Bias (2018): R +1.4
Senatorial Poll Bias (2018): R +3.3
Presidential Poll Bias (2020): D +0.6
Senatorial Poll Bias (2020): D +3.3
Biden Approval Rating Average: -15
RCP Average (Governor): Lake (R) +3.5
RCP Average (Senator): Masters (R) +0.3

Winners: Lake, Kelly

Arizona is a mixed bag on polling bias, with it being difficult to establish consistent bias one way or another. If I had to pick one competitive Senate race in which the Republican challenger falls short alongside New Hampshire, it would be Arizona. Kari Lake, however, I think will be the next governor of Arizona on account of a charisma chasm between her and Katie Hobbs. Mark Kelly may be in a strong enough position to hold on as perhaps his record IS independent enough for Arizonans and understandable personal sympathies remain for what happened to his wife, former Congresswoman Gabrielle Giffords, for him to pull through.

COLORADO

Senatorial Election (2014): 0
Gubernatorial Election (2014): R +1.4
Presidential Election (2016): R +1.9
Senatorial Election (2016): D +2.2
Gubernatorial Election (2018): No Average
Presidential Election (2020): No Average
Senatorial Election (2020): No Average

Biden Approval: -8

RCP Average: Bennet (D) +5.7

Winner: Bennet

Colorado is a state that has moved from purple to light blue in recent years, and this election probably is going to reflect that. Frankly, I like Republican Joe O’Dea, and he will probably perform considerably better than Ron Hankins would have, but the state’s current political environment seems too great for him to overcome. O’Dea only came within the margin of error in a Trafalgar poll, but others have shown Bennet up around 7 to 8 points. The incumbent Democrat is not sufficiently offensive for Colorado voters for even someone like O’Dea to take him down. If O’Dea wins this one, this midterm is catastrophic for Democrats.

FLORIDA

Presidential Poll Bias (2016): D +0.8
Senatorial Poll Bias (2016): D +4
Senatorial Poll Bias (2018): D +2.6
Presidential Poll Bias (2020): D +4.2
Biden Approval: -15.8
RCP Average (Governor): DeSantis (R) +12.2
RCP Average (Senator): Rubio (R) +8.8

Winners: DeSantis, Rubio

Governor DeSantis and Senator Rubio currently have poll margins that are insurmountable. In none of the latest polls for either have their Democratic opponents come within five points. This plus Biden’s unpopularity in the state, rampant speculation that DeSantis will win Miami-Dade County, and the state’s increasing Republican bend make this an easy prediction.

GEORGIA

Gubernatorial (2014): D +3.4
Senatorial (2014): D +5.1
Presidential (2016): D +0.3
Gubernatorial (2018): R +1.6
Presidential (2020): R +1.3
Senatorial Runoff (Loeffler vs. Warnock): R +1.1
Senatorial Runoff (Perdue vs. Ossoff): R +0.5
Biden Approval: -15
RCP Average (Governor): Kemp +8.3
RCP Average (Senate): Walker +1.4

Winners: Kemp, Walker

The reelection of Brian Kemp to me seems a good repudiation of baseless cries of both widespread voter fraud and voter suppression and a victory for those who lack crybaby mindsets about elections. He is too far ahead in polling to lose right now. Stacey Abrams will retain her status as a “Superstar Loser”, a concept expounded on in an Atlantic article I will link below. Herschel Walker, I think, will be propelled forward by the momentum of the midterm and his charisma despite him having some flaws as a candidate and an alleged past that does not speak well for him.

https://www.theatlantic.com/technology/archive/2022/11/beto-orourke-stacey-abrams-superstar-loser/671959/

KANSAS

2014 (Governor): D +5.9
2014 (Senate): I +11
Biden Approval: -29
RCP Average (Governor): No RCP average due to lack of polls, but last two showed Democrat Laura Kelly up.

Winner: Schmidt

Kansas has a bit of a history of underpolling Republicans, but an exception to this was the 2018 race that Kelly won against Kris Kobach, a highly controversial and flawed nominee in a difficult year for Republicans. There isn’t much in RCP poll averages I can put up here since Kansas is not often heavily polled in national races, but the last Senate race had two polls with Marshall up ahead by only 4 and 2, but he ended up winning by 11.4. One of the worst polling errors occurred here in the underestimating of Kansan Pat Roberts’ ability to win reelection in 2014 against Independent Greg Orman. Orman was averaged 0.8 ahead, but Roberts won by 10.8. I would not be surprised at all if Republican Derek Schmidt won the election, in fact I’m counting on it.

MAINE

2014 (Governor): D +3.5
2014 (Senate): D +7.8
2016 (President): D +1.6
RCP Average (Governor): None provided, but all polls have Mills in the lead, most by over 5.
Biden Approval: -10.7

Winner: Mills

Although Maine is perhaps less of a Democratic state than it used to be, Democrat Janet Mills will win reelection as LePage is a bit overly Trumpy for them and no poll I have seen points to the possibility of a LePage comeback. BUT…the last Senate election also showed no polls that Susan Collins would win reelection. This perhaps leaves the slightest sliver of a chance that LePage gets back in, but I think Maine voters know that Paul LePage is no Susan Collins. Mills will win, but perhaps by less than what polls are indicating.

MICHIGAN

2014 (Governor): D +2.2
2014 (Senate): R +0.4
2016 (President): D +3.9
2018 (Governor): D +1.5
2018 (Senate): D +1.9
2020 (President): D +1.4
2020 (Senate): D +3.7
RCP Average (Governor): Whitmer (D) +1
Biden Approval: -16

Winner: Dixon

Given the history of poll bias favoring Democrats in Michigan as well as residing resentments over the state’s COVID-19 lockdown policies, I think Tudor Dixon, despite some significant campaign missteps earlier in the race, can overcome and win. The polling momentum is with her.

MINNESOTA

2014 (Governor): D +1.4
2014 (Senate): R +0.3
2016 (President): No RCP average, but Clinton won by only 1.5 when most polls had her over five points ahead.
2018 (Governor): No RCP average, but Walz pretty much performed as expected.
2020 (President): R +2.9
2020 (Senate): No RCP average, but polls underestimated Democrat Tina Smith overall.
RCP Average (Governor): Walz (D) +4.3
Biden Approval: -7.3

Winner: Walz

Minnesota is a state teeming with potential for Republicans, but they just can’t quite clinch major statewide offices lately, and a recent history of polling isn’t exactly encouraging for their prospects on this one.

NEVADA

2014 (Governor): D +18
2016 (President): R +3.2
2016 (Senate): R +0.6
2018 (Governor): R +4.7
2018 (Senate): R +5
2020 (President): R +0.3
Biden Approval: -15.6
RCP Average (Governor): Lombardo (R) +2.8
RCP Average (Senator): Laxalt (R) +3.4

Winners: Lombardo, Laxalt

While it is true that Nevada has a history of overpolling Republicans, the bulk of major overpolling happened in 2018, a year in which Republicans were on defense and their fundamentals were not great in a state that had voted against Trump in the last election. I expect the polls to again have a Republican bias in this race, but not to enough of a degree to which Lombardo and Laxalt will lose. Laxalt has led in five of six of the last polls, including one by a Democratic firm, and the only poll for Cortez Masto showed her up by one. Lombardo has been up in every recent poll except one in which he was tied with Governor Sisolak. What’s more, a recent poll of Nevada’s four House seats shows the Democrats behind in three of them. Nevada was also one of the states in which Trump did better (slightly) in 2020 than in 2016. This one’s, however, still a real tossup given the state’s polling history as well as the late Harry Reid’s machine possibly still being effective, but a recent influx of Latino voters for Republicans may put them over the top. Sisolak also has issues as the Northshore Clinical Labs scandal hit his administration and his lockdown policies are remembered bitterly by many Nevadans. Laxalt and Lombardo I narrowly call winning here, with actual results showing less of a lead for them than current polling indicates.

NEW HAMPSHIRE

Gubernatorial (2014): R +1.7
Senatorial (2014): R +2.4
Gubernatorial (2016): D +1
Senatorial (2016): R +1.7
Gubernatorial (2018): D +3
Senatorial (2020): R +0.9
Presidential (2020): D +0.8
Biden Approval: -10
RCP Average (Senator): Hassan (D) +1.4

Winner: Hassan

There is only one precedent in polling error in New Hampshire that could point to a Bolduc win, and that is the gubernatorial contest in which Chris Sununu won reelection in 2018, and Sununu was then, and remains, popular in New Hampshire. Despite a tightening in the race, Bolduc may just be too extreme for New Hampshire. However, there is an outside chance that inflation, particularly the rising cost of fuel oil, will motivate New Hampshire voters at the polls this year to turn out Hassan. However, I fail to see a scenario in which Bolduc wins without both Masters in Arizona and Laxalt in Nevada winning, and RCP history indicates that neither Republicans nor Democrats win all tossups in an election year. This race is a bit of a bellwether for this election: if Bolduc loses or the race remains too close to call throughout the night, it is a normal midterm. If the race is called for Bolduc early in the night, this is a disastrous midterm for the Dems and even Colorado and Washington become possible losses.

NEW MEXICO

Governor (2014): D +1.3
President (2016): R +3.3
Governor (2018): R +6.2
Senator (2018): R +6.4
RCP Average (Governor): Lujan Grisham (D) +4

Winner: Lujan Grisham

New Mexico is a light blue state, and the only poll having Republican Mark Ronchetti up recently is a Trafalgar poll and he’s up by one…in a state that has a history of overpolling Republicans.

NEW YORK

Governor (2014): D +9.9
Governor (2018): R +4.2
Biden Approval: +1.8
RCP Average (Governor): Hochul (D) +7

Winner: Hochul

Although Democrats have been panicking over Kathy Hochul possibly losing the election with campaign appearances from major Democratic figures, I do not think this comes to pass. I think she holds on by single digits as the Empire State still has strong Democratic fundamentals even if they are being tested right now with crime and an unusually strong Republican candidate in Congressman Lee Zeldin. That Democrats are getting a scare out of this race may either be a great sign for Republicans across the board or reflective of a major fear of the symbolism that will come out of a Democrat losing reelection in a major Democratic stronghold. If Lee Zeldin wins, Democrats will have to reevaluate how they are going about politics…in New York.

NORTH CAROLINA

Senatorial (2014): D +2.9
Senatorial (2016): D +3.7
Presidential (2016): D +2.8
Gubernatorial (2016): D +2
Presidential (2020): D +1.1
Senatorial (2020): D +4.4
Gubernatorial (2020): D +6.6
Biden Approval: -16
RCP Average (Senator): Budd (R) +6.2

Winner: Budd

All the fundamentals are here for Budd. We have an election year that’s getting worse by the day for Democrats, a history of repeated overpolling of Democrats in the state with not one counterexample over the past three election cycles, North Carolina having twice voted for Trump, and Budd is up by over five on the RCP average. He has not gotten below five in the lead in the last five polls. Media stories trying to focus on this race as a possible sleeper are simply trying to save Beasley’s campaign.

OHIO

Presidential (2016): D +5.9
Senatorial (2016): D +2.5
Gubernatorial (2018): No RCP average, but the last three polls had Democrat Richard Cordray leading by three or more points.
Senatorial (2018): D +4.6
Presidential (2020): D +7.2
Biden Approval: -18
RCP Average (Senator): Vance (R) +8

Winner: Vance

Like with Budd, all the fundamentals are here for J.D. Vance as well. A state that twice voted for Trump and has a history of overpolling Democrats. The media stories about an upsurge for Tim Ryan, if history is a guide, are merely stories. J.D. Vance will be Ohio’s next senator.

OKLAHOMA

2014 (Senate): D +5.5
2014 (Senate Special): D +7.6
2018 (Governor): D +6.8
Biden Approval: -41.5
RCP Average (Governor): None provided, but last two polls have Stitt in the lead.

Winner: Stitt

Despite some major trouble for Kevin Stitt surrounding ethics and an unusually strong challenger in Democrat Joy Hofmeister, the fundamentals of the midterm will probably save his keister. The most recent polls suggest a momentum back in his direction.

OREGON

2014 (Governor): D +2.6
2014 (Senator): R +1.2
2016 (Presidential): R +2.6
2016 (Senator): No RCP Average
2018 (Governor): R +1.8
2020 (Presidential): No RCP Average
Biden Approval: -6
RCP Average (Governor): None provided, but last two polls have Democrat Tina Kotek in the lead.

Winner: Kotek

Betsy Johnson’s Independent campaign seemed to really threaten Democratic prospects for continuing to govern Oregon, with Republican Christine Drazan possibly winning because of a split of liberal voters and some disenchantment over the issue of crime in Portland. However, the latest two polls are showing that voters are moving away from Johnson and towards Democrat Tina Kotek. The prospect of a Republican winning the race seems to have spooked enough of them to back off Johnson and continue the long reign of the Democrats over the state.

PENNSYLVANIA

Gubernatorial Poll Bias (2014): D +1.2
Presidential Poll Bias (2016): D +2.8
Senatorial Poll Bias (2016): D +3.6
Gubernatorial Poll Bias (2018): D +2.5
Senatorial Poll Bias (2018): D +1.4
Presidential Poll Bias (2020): 0
Biden Approval: -11.3
RCP Average (Governor): Shapiro (D) +8.3
RCP Average (Senate): Oz (R) +0.4

Winners: Shapiro, Oz

Doug Mastriano has consistently been behind Attorney General Josh Shapiro in polling as he has a lot of baggage, including being present at the January 6th rally. Shapiro wins this one. For the Senate, it is hard to imagine after the disastrous debate performance from Fetterman, which revealed how much his condition was being covered up by his campaign and by his supporters in the media, that Oz does not pull through. The scandal, as they say, is in the cover-up. The only conceivable one is if he rides on the coattails of Josh Shapiro. There have, however, been two polls after the debate performance that still gave Fetterman an edge, including a Marist poll that had him up 6. The majority of polls, however, have Oz at a slight advantage. Given how many people appear to be planning to split their tickets or even not mark the Senate race combined with a history of mildly overpolling Democrats, I think Oz clinches it.

TEXAS

2014 (Governor): D +3.4
2014 (Senator): D +5.5
2016 (President): R +2.7
2018 (Governor): R +3.4
2018 (Senate): R +4.2
2020 (President): D +4.5
2020 (Senate): D +3.3
RCP Average (Governor): Abbott (R) +10.4
Biden Approval: -13.7

Winner: Abbott

What is there to say about this one? Former Congressman Beto O’Rourke, like Stacey Abrams in Georgia, is part of a Democratic class of “Superstar Losers” whose biggest claims to fame are that they came within five points of winning statewide elections in states that are widely thought of as Republican. What’s more, Democrats are appearing to have problems in Latino areas of Texas that before would be easy wins for them.

WASHINGTON

President (2020): D +1.2
RCP Average (Senate): Murray (D) +3
Biden Approval: +1

Winner: Murray

Although Republicans have fielded perhaps their strongest Senate candidate in this election since Republican Slade Gorton lost reelection in 2000, there appear to be too many fundamentals helping Patty Murray stay for another term. The state’s Democratic bent and Biden’s approval rating being ABOVE water in addition to there not being a single poll placing Smiley with a lead makes me conclude that despite a seemingly tightening race that Murray remains, even if the election ends up too close for comfort for Democrats. Seattle and Tacoma will, I think, save her bacon. Washington also is rather lightly polled on races as it is seldom competitive on gubernatorial or senatorial races. I write this as someone who as a resident of this beautiful state really wants Smiley to win and voted for her.

WISCONSIN

2014 (Governor): D +3.5
2016 (President): D +7.2
2016 (Senate): D +6.1
2018 (Governor): No RCP Average, but most polls had Democrat Tony Evers up by a larger margin than his victory of 1.2.
2018 (Senate): R +0.2
2020 (President): D +6
Biden Approval: -15.5
RCP Average (Governor): Michels (R) +0.6
RCP Average (Senate): Johnson (R) +3.6

Winners: Michels, Johnson

Ron Johnson is an enigma for the MSM. They regularly rip on him, yet he beats expectations and he’s doing so again this year. Although Wisconsin voted for Biden in 2020, the state has a history of DRAMATICALLY overpolling Democrats with the sole exception being a light overpolling of Republicans for Tammy Baldwin’s reelection in 2018. Johnson is in the lead, and the fundamentals for the Democrats this year, to put it bluntly, suck. Ron Johnson is either the luckiest senator in politics, or he knows a thing or two that his critics don’t. Wisconsin’s history of poll bias for Democrats combined with the fundamentals for the midterm make me think that Tim Michels will be elected governor as well.