one of the things i do here is analyze polls. i have a degree in math, and experience in polling. i have a pretty good track record of avoiding algorithms and developing approaches on an election-by-election basis. elections are unique and always have factors to identify if you want to get a good prediction. standard models and textbook formulas don't work well in predicting elections.

i want to give you a heads up on hungary - the polling is unreliable. the outcome is entirely unclear.

every single government poll has orban ahead. every single "independent" poll is funded by the opposition. an analyst has nothing to work with, here. it looks very much like an iron curtain state.

something polling firms have found in iron curtain states is that even if the polling is clean, the respondents won't be honest with you. a very large percentage of people in these countries remain fearful about anybody calling them and asking them questions, and that works in any and every direction.

you further need to factor in the likelihood of corruption, both by the government and the opposition, which is high.

so, this is currently a toss-up. there is no way to analyze the data i have available to me, other than to label it unreliable. and, despite being an eu member state, i do not have confidence in the elections being free or fair - and that criticism is in both directions.

your guess as to how this turns out is as good as anybody's.