There are several reasons for Democrats to be optimistic about their prospects in the 2018 Midterm Elections. Historical patterns regularly favor the out party over the governing party in off-year contests. Donald Trump is a historically unpopular President who continues to dig himself into an ever-deeper political hole. The Republicans in Congress have only one major legislative achievement to date and are divided on issues of trade, immigration and limiting assault weapons. But despite all of this, most Democrats are at best cautiously optimistic. Many are anxious about their party’s unity and uncertain about how to win back the voters who either stayed home or defected to Trump in 2016.

Some of their uncertainty derives from circumstances beyond their control. The economy continues to hum along with low unemployment, a strong stock market (despite occasional dips in reaction to Trump policies) and adequate levels of GDP growth. Trump’s base sticks with him even as he acts ever more recklessly because he defends their values and voices their mistrust of over-educated coastal elites. And while the tax bill Trump signed does more for the rich than for his base, it at least does not burden his core supporters financially in the short term.

Democrats are uncertain about how to campaign given all these mixed circumstances. Should they move left to appease the Sanders wing, hoping to draw in new young and nonwhite voters with bold policy pledges like single payer health care, higher minimum wage or loosened Federal marijuana restrictions? Or instead, should they broaden their coalition to encompass more right-to-lifers and NRA supporters? And is it better to nationalize the election a la Newt Gingrich in 1994 or localize it as they did in 2006 when Rahm Emmanuel ran the DCCC’s operations, actively seeking out cops, military vets and the like to run in the moderate to conservative swing seats? So far in this election cycle, the DCCC appears to be taking a more hands-off, wait and see approach in the emerging Democratic primary contests.

There is nothing new about “left and lefter” tensions within the Democratic Party ranks, but they have become more complex due to broader changes in the political ecosystem. In the immediate post-WW II era, the professional operatives uniformly preferred winning over ideology while the grassroots amateurs favored principles over electoral expediency. The difference between the professionals and the amateurs has lessened. The battle over the DNC Chair revealed that even the official party ranks are ideologically divided.

Also, with the proliferation of public polls and social media forums, the professionals no longer have a distinct advantage over the grassroots in information and data. The districts in serious play are common knowledge thanks to sites like Larry Sabato’s Crystal Ball and other online resources. As a consequence, there are more amateurs at the district level and in the political nonprofit world playing the targeting and mobilizing game at a sophisticated level. Meanwhile many professionals from the TV ad era struggle like amateurs as social media, interactive software, and big data have transformed what consultants do. It is simply not clear who knows best in the party, even when everyone agrees that winning is paramount in the Trump era.

Another complication is that the Rovian revolution has reached Democratic shores. Karl Rove turned conventional wisdom around in 2004 when he focused the Bush campaign on mobilizing the party base rather that attempting to woo undecided voters. While independent voters are on average more moderate, or at least more inconsistently partisan, they are also less likely to turn out, especially in primaries and mid-term elections. The success of the Tea Party and now the Trump base in controlling the Republican party’s message has made a strong impression on Democratic progressives. The consequence of this partisan mirroring of the Republican base-oriented strategy is that many progressives now sincerely believe that there is no tradeoff between ideological purity and winning elections—“lefter” policy, they claim, is also smarter politics.

Lastly and most importantly, the party system’s fracture due to recent campaign finance rulings means that there is no recognized final arbiter of party strategy and tactics. Because independent spending individuals and nonprofit entities control valuable resources, they cannot be pushed aside easily. Self-financed candidates, free to spend as much of their own money as they please, can effectively decide on their own what they will run on. The consequence is that absent anyone being in charge, there are lots of affiliated groups and individuals trying to figure it out for themselves and only sometimes in consultation with one another.

While some of this Democratic disarray can plausibly be blamed on the Supreme Court’s recent campaign finance decisions that created these many outside spending entities, the Democrats have also made their own institutional contribution to political disorder. California is ground zero for the Democrats’ chances of taking the House. They have at least 7 Republican House seats plausibly in play. They also have many candidates—in fact too many—in several of the potentially winnable seats.

That would not be a problem under the primary rules that most states have adopted, but California has a system that only allows the top two vote getters to go forward into the November election regardless of party. This means that in certain competitive districts the Democrats could fail to field even one candidate in the November contest if the too many candidates split the vote and achieve fewer votes than two Republican candidates. Early polls suggest this is a real possibility in at least two seats.

Why does California have such a system? Basically, California Democrats are predictably attracted to any proposal that is described by the press as “reform” and makes the voting process more “open.” Many Democrats actually voted for this concept twice, first for a blanket primary system that the Supreme Court eventually threw out and then for this current top two version that passed judicial scrutiny. Were the Democrats warned that this process might result in the majority party being excluded from running in November or that it would mean spending money inefficiently to re-run a primary contest between two candidates from the same party? Yes, but to no avail.

Will this necessarily hurt the Democrats in November? A lot will depend upon whether the losing factions in these contested primaries stay home or flock to the polls to oppose the President and his policies. In general, it is easier in politics to rally around the negative (i.e. opposing Trump) than the positive (i.e. defining your own agenda). The only realistic goal for Democrats in 2018 is negative: even if the Democrats win both the House and the Senate, they will not have enough votes to overcome a Presidential veto or a Senate filibuster. They will need an answer to their “left and lefter” divide eventually, especially if they obtain a so-called trifecta in 2020. In the meantime, their goal will be to block what they don’t like while they figure out what they do.