Things are heating up politically as we approach election season. Jumping right into the numbers, I've done a single-poll mathematical projection for Germany:
As you can see, in this poll, the SPD had a healthy lead. The existing SPD-CDU coalition would not have a majority, meaning a third party would need to be brought in. Additionally, a SPD-GRN-LNK coalition would have a majority. Note however that this estimates the total number of seats at 650. More on that later.
I've also done an estimation for Japan
310 - LDP
30 - NKP
80 - CDP
15 - JCP
15 - ISH
15 - OTH
Note that Japan is hard to estimate due to their strange attachment to using approval polling instead of vote intention polling. Beyond that, even the vote intent polling that is available often does not come close to the results in the same way they do in most countries. As such, a degree of guesstimation has to be done in order to determine numbers like this.
Back to Germany, I've managed to find a website that seems to do election predictions; here is their latest prediction for the first-ballot constituency seats: https://www.election.de/cgi-bin/showforecast_btw21.pl
Note that Germany only elects half it's MPs from these seats, with the other half coming from the proportional lists. This is why parties like the Greens, AFD, and Linke are expected to win many times as many seats as shown, and why the FDP can win near 100 seats despite not having a single 'dot' on the map. Note that the map shows the CDU winning 181 seats, but I only have them at 140. Why is this? Simply; Germany provides overhang seats to prevent a party winning more seats than it "deserves".
Take a look at Bavaria on the map. It is one of the two southernmost states, the one on the right. Munchen (Munich) is its capital city and shown in a small inset. The map currently shows 5 Green victories and 41 seats for the CSU (the Bavarian sister party of the CDU) Polling, however, shows the CSU only at about 35% in the state. By default, Germany assigns the same number of PR, or, List seats to a state, as it assigns geographic seats. This means that of a grand total of 92 seats exist in Bavaria. Of those 92, however, the CSU could win 41. This, however, is 44.565% of the seats, far more than the 35% they would "deserve". As such, the number of list seats gets increases until the CSU is 'only' winning 35% of the total seats. This would require 117 seats, meaning that an additional 25 seats have to be added to the Bavarian list, bringing it's number of list seats up to 71 (remember, it is only "supposed" to have 46 seats). This is a 1.54x increase. As this would unbalance the representation between all the various states, all other states now get a boost to the number of list seats that exist within their states as well, also by a factor of 1.54. This means the grand total of 299 list seats nationwide now becomes 460 seats. This gets added to the 299 geographic seats, for a grand total of 759, far above my estimated total of 650. This would bring the CDU/CSU coalition up to 161 seats for my single-poll projection, and, 'single poll' is exactly why that number does not match the multi-poll german-based election projection; IE, using a single poll can be more inaccurate, but, using a multi-poll system can miss fresh and hot trends.
I'm bringing up the number of geographic seats, expected to be won by party, for a simple reason: What we are seeing is a massive and radical change from the past. Last election, the AfD won 3 geographic seats, the Left Party (Die Linke) won 4, ad the Greens won 1. In both 2013 and 2009, the AfD won 0, the Left won 4, and the Greens won 1. 2005 saw the Left win 3, and the Greens 1, while 2002 saw the Left win 2, and the Greens just 1, which was their first ever. In 1998 and 1994 the Left won 4, and in 1990, they won 1.
Continuing back, in West Germany, we see a grand total of 0 geographic seats being won by a party that is not the SPD or CDU/CSU in elections in 1987, 1983, 1980, 1976, 1972, 1969, 1965, and 1961. In fact, you have to go back to 1957 to find a smaller party winning seats; in this case, the FDP won 1, and the AfD-like "German Party" won 6.
1953 had far more, seeing 14 FDP seats, 10 German Party, and 1 for Zentrum, while 1949 had 12 for the FDP, 5 for the German Party, 11 for the "Bavaria Party" and 3 for Independents.
Meanwhile, in East Germany, the system provided for 0 geographic seats at any point, and so, there is nothing to look at. Weimar Germany also had no geographic seats, meaning you have to go all the way back to 1912, before the introduction of any form of Proportional Representation, to find more results to look at. If you are interested, you can view such a map on that election's wikipage, but, as it does not relate to my point, I will not be doing so myself.
Instead, I look back to the current prediction from election.de and see that the Left is expected to win a record 5 seats. The AfD is expected to go from 3 to 13. The Greens are expected to go from 1, to 19. In fact, 19 would be the largest number of geography seats won by a "small party" in the history of Germany, more than the 14 the FDP took in 1953. Additionally, since 1960, not a single geographic seat in West Germany has been won by any 'small' party, yet the Greens look set to take 15 of them.
Assuming these projections are correct, a total of 37 geographic seats will be won by a "smaller" party in 2021. This would be 37 of 299, or, 12.4%. This beats the 10.3% elected in 1953 (25 of 242) but not the 31 of 242 that won in 1949 (12.8%)
Regardless, assuming this trend holds, it will certainly be of note in the history of elections in Germany, and may indicate a coming realignment in its political system.