the liberals could get over 160 when the smoke clears, but i'm conceding a minority. what happened?

i had the numbers for the ndp and liberals and bloc pretty close, and my seat counts are pretty good everywhere except ontario, although i don't think i overexaggerated the ontario numbers by more than the models underexaggerated them; i made the right correction but overcompensated on the greens splitting the rural vote, which didn't happen. the bradley effect seemed to be a real thing. it seems like i got the 905 right, but we'll need to wait. i mentioned that if i got something wrong, it would be on the greens not splitting the vote, so it's not like i'm that surprised, but i argued it shouldn't affect the outcome - and was wrong on that point.

i more got nailed on the crumbs.

so, i had the ndp at 20 and the bloc at 30; actual numbers were 23 or 24 and 32. that's 5 or 6 seats i had given the liberals. i had the conservatives at 13 in the east and they got 14; i'm going to be off by a handful out west, too.

these are minor errors, but they made my big error - the greens running low in ontario - matter when i argued it shouldn't.

so, my number was 195, and i pointed out that i might be off by about 15 in ontario. in fact, i was off by about that, then missed a number of seats that the conservatives unexpectedly picked up due to ndp support switching to the conservatives, including in kenora and essex. that was something i didn't factor in - that the conservatives may also benefit from low ndp support.

so, i was wrong about the greens splitting the vote, but anticipated it; i was right about the liberals benefiting from weak ndp support, but i missed anticipating the conservatives also benefiting from weak ndp support. final numbers are not in, but if i lowballed the conservatives by a few points, it's because i had the ndp higher than they actually got. that cost me another few seats in ontario.

i'm sleepy. we'll do this better after i take a nap.

but, of all the claims i made, i got most of them right. and, i think i beat the models, and that my criticisms of the models were correct, even if i went too far, and i didn't always provide a better proposal; i was right to point out they were wrong, but my corrections had problems of their own.