Elections in Uttar Pradesh begin in just over a month, yet predicting the results of the polls still seems impossible. In 2012, it was a simple question of whether anti-incumbency against former Chief Minister Mayawati would be enough to dislodge her from the post. In the event, most opinion polls at the time correctly predicted that the Samajwadi Party would take over, although they didn’t forecast the massive scale of the victory. This year is proving to be a whole different ballgame.

There are just too many variables, some of which may continue to change down to the wire. The Bharatiya Janata Party hasn’t yet given any indication of a chief ministerial face and looks set to fall back on Prime Minister Narendra Modi’s popularity. While that tends to be a reasonably dependable commodity, it’s still too early to say how the government’s demonetisation measure will impact Modi’s image.

The incumbent Samajwadi Party, meanwhile, has spent much of the last few months squabbling over who should be in charge, with the primary battle being between party chief Mulayam Singh Yadav and his son, Chief Minister Akhilesh Yadav. That fight seems like it is coming to an end, with Akhilesh on top, but there’s no telling if there are more twists ahead.

The Congress started off strong and vocal, but has since gone quiet, in part because those volumes did not translate into much support on the ground. The party now seems to be waiting to cement an alliance with the SP, which would mean dispensing with its chief ministerial candidate and some portions of it strategy.

And finally there is the Bahujan Samaj Party. Despite some public exits, Mayawati’s organisation has been mostly drama-free, but because of its Dalit support base, the BSP is always under-covered and underestimated, so it is hard to gauge its relative chances.

As Akhilesh firms up his control over the SP, though, it seems likely that the alliance with the Congress will also translate into reality. With the Rashtriya Lok Dal and the Janata Dal (United) also on board, that would make a reasonably powerful alliance.

How powerful? A look back at the previous assembly election results offers some idea:

Allowing for very simplistic calculations, considering the complex nature of India’s first-past-the-post-system, it would seem like a combination of the SP, Congress and RLD would have been streets ahead of the BJP back in 2012.

But that election is not directly comparable to this year’s polls. For one, the BJP has seen a huge rise in fortunes ever since Modi and President Amit Shah took the reins of the party in 2013. The idea of the BJP getting just 15% of the vote share this year, as it did in 2012, would not just be unlikely – it would also be unprecedented.

That’s because of the results the BJP picked up in the 2014 Lok Sabha polls, when it managed a whopping 42% of vote share in Uttar PRadesh, translating into 71 of the 80 seats in the state. Even if the SP had combined with the Congress and the RLD in 2014, the alliance would still have been nearly 12 percentage points behind the BJP.

But again, that election isn’t comparable either. Indian voters have shown, time and again, that they vote differently in state and Lok Sabha polls, and the “Modi wave” has tempered somewhat, as results from Delhi and, more importantly, Bihar in 2015 showed us.

In fact, Bihar might offer the most relevant understanding of how these numbers might come together, because of how similar that election was. That year, the Janata Dal (United) teamed up with its chief opponent, the Rashtriya Janata Dal, with the Congress also on board, to take on the BJP and its alliance partners.

Similar questions were raised about whether simply adding the vote share of the JDU, the RJD and the Congress could predict the final tally.

As the chart shows, a combination of the RJD, JDU and Congress – which did not contest together prior to 2015 – did not exactly match a simple addition of their numbers. But the 2015 number was not far off from the 2014 calculation, while the NDA’s final vote share also fell.

Assuming votes cross over as remarkably as they did in Bihar in 2015, this suggests the SP-Congress-RLD combine should also expect that their sum will at least be as much as their parts.

There is, however, one big difference.

In Bihar, the two principal state parties – the JDU and the RJD – came together in the face of what seemed like a BJP onslaught. In UP, there is no question of the SP and the BSP joining forces, even if it is to take on the BJP. That means the final fight, if the SP’s alliance comes together, will still be three-cornered. Moreover, the swing voters being courted by the SP’s alliance will be the same as those the BSP will hope to retain, which, because of vote splitting, would give the BJP an advantage.

But before any of this happens, we still need to see if the SP can pull itself together in time to actually put together an alliance. As long as Mulayam Singh Yadav is around and calling some shots, nothing can be taken for granted.