The Democrats and their two independent allies, Angus King of Maine and Bernie Sanders of Vermont, currently hold 48 out of the 52 seats in the U.S. Senate. It should be relatively easy, considering President Trump’s weakness, to net three Senate seats in the 2018 midterms and take control of the the upper chamber of Congress.

Unfortunately, this task will be next to impossible. Most people focus on the disparity in how many seats from each party are up for election, but that’s not the real obstacle. It’s true that the Democrats will be defending 25 seats while the Republicans will only be defending eight. It’s also true that some of the Democratic seats will be hard to hold. I’m thinking of seats like Joe Manchin’s in West Virginia, Claire McCaskill’s in Missouri, and Heidi Heitkamp’s in North Dakota. There are other seats from states that Trump won that can’t be taken for granted. Bob Casey in Pennsylvania is probably safe, but can we say the same for Tammy Baldwin in Wisconsin or Sherrod Brown in Ohio?

But even if the national political climate in the fall of 2018 is so hostile to the Republicans that they can’t win any Democratic seats (which is a real possibility), the seats held by Republicans are almost all as safe as any Senate seats can be. Look at the following list and try to find three seats the Democrats can win:

John Barrasso of Wyoming
Bob Corker of Tennessee
Ted Cruz of Texas
Deb Fischer of Nebraska
Jeff Flake of Arizona
Orrin Hatch of Utah
Dean Heller of Nevada
Roger Wicker of Mississippi

The lowest hanging fruit is definitely Dean Heller’s seat in Nevada. Nevada is the only state on this list that Hillary Clinton won. The next logical target is Jeff Flake’s seat in Arizona, but Flake has been adept at creating distance from Trump. These are the two best targets but even if the Democrats won them (and held all of their 25 seats), they’d only get to a 50-50 split.

So far, it looks like the Dems will make Ted Cruz the third target, and that makes sense. Clinton did better in Texas than she did in Iowa and the Dems have a pretty good candidate in Beto O’Rourke. Besides, there aren’t any other promising options. You can look at Nebraska. Bob Kerrey and Ben Nelson won Senate elections there in the not-too-distant past and Fischer is a freshman without much of a positive record to tout. That seems like the longest of long shots, though.

If Orrin Hatch were to retire things could open up in Utah where Trump has never been popular. Perhaps a strong third-party challenge from the right could split the vote and give a Democrat a chance. Perhaps a third-party victor might caucus with the Dems. This, too, seems like a very unlikely result.

I have trouble envisioning a scenario where Bob Corker, John Barrasso or Roger Wicker could possibly lose.

The last option would be in Alabama where the Republican Party is embroiled in so much scandal and where there is a growing fight among Republicans for control of Jeff Sessions’ seat which is currently held by Luther Strange on an interim basis.

In truth, the Democrats are not going to pursue a strategy aimed at winning Senate seats in Wyoming, Alabama and Mississippi. They’ll be playing defense in Ohio and Wisconsin. But it’s just as important to play defense in Missouri, West Virginia and North Dakota, and that is more consistent with a strategy that could attract support in places like Nebraska, Utah and Tennessee.

Either way, though, so long as the House strategy is to go after affluent well-educated suburban seats, the message will be ill-suited for making any progress in these Senate races. In other words, the strategy for winning the House is not consistent with a strategy to win the Senate.

How either strategy can meld with a strategy for winning the Electoral College in 2020 is an open question.

The way I see it, the Democrats can double down on Clinton’s strategy and win if they gather much more strength among the well-educated and affluent in the suburbs without losing a corresponding amount of support in small towns and rural areas. Trump will do most of their work for them in the suburbs, but they have to hold or reverse the line in the less populous areas.

On the other hand, even without further erosion in the suburbs for the Republicans, the Democrats can win if they can do better in rural areas. And this strategy is more consistent with the four goals the Democrats have. They want to win the Senate, the House, and the presidency. They also want to win state legislatures. The suburban strategy is a good one for the House and it could work, possibly, for the presidency, but it won’t help and will potentially hurt with the Senate and the state legislatures.

I don’t think the Dems can create one holistic strategy that serves every purpose, but they ought to recognize the obvious shortcomings of a strategy of just focuses on solidifying gains among well-educated, affluent suburban whites.

I’ve identified an anti-monopoly, anti-consolidation push as one that can work in all areas and serve all goals without getting bogged down in cultural battles. I’m open to other ideas. But I want people to be clear about the shape of the battlefield and why the Democrats can’t accept the realignment that occurred in 2016.

Too many people view all of this through the lens of being either for or against core civil rights issues. What results is a belief that the only option is to accept our losses and seek to consolidate our gains. But that isn’t the only way to look at this, and it’s a very risky and probably losing strategy. It shouldn’t be necessary to point this out, but it’s doubtful that the most efficient way to protect people’s rights is for the left to continue to be a minority party in most of the country. Any plan that accepts that future or that has a doomed strategy for avoiding it is a threat to people’s civil rights that should be taken just as seriously as what the Republicans are doing to us.

What I find frustrating is that the American people can be as pissed off at Republicans as they’ve ever been and come out to vote against them in the 2018 Senate elections in droves, and the Democrats won’t win control of the Senate unless they can beat Ted Cruz or Bob Corker or Roger Wicker. But just because this frustrates me doesn’t mean that I don’t have to account for it. Leaders on the left will make all kinds of promises and raise ungodly amounts of money from people who want to believe they can make a difference. What happens when they discover that the game is so rigged against them that they’ve lost again despite getting millions of more votes?

At some point, the left will abandon electoral politics and turn to something darker. It will be hard to blame them, frankly. But true visionary leaders need to anticipate this and tell people uncomfortable truths. You can’t win the battle for the Senate by convincing the most people or winning the most votes. You can only win it by taking down senators who serve is blood red seats. That calls for new thinking, and if you don’t want that thinking to involve watering down or selling out on core principles, you need a strategy that transcends those issues.

That’s where my thinking has been since November 2016, and I’ll take any allies I can get.