IL-09 Poll Updates
@il9forecast.bsky.social
A project by @coleurbanist.bsky.social to model the IL-09 democratic primary
Pinned
Many updates today including a Biss internal poll but also a domain name! Model is live at illinoisforecasting.com
The main page is now a summary page showcasing district wide odds and expected margins as well as how much they changed since the last simulation run.
The main page is now a summary page showcasing district wide odds and expected margins as well as how much they changed since the last simulation run.
New! Ideology map of Chicago. a precinct with a 0 score will (on average) vote exactly like the city. +10 or -10 will vote 10 points more progressive or conservative in the mayoral race respectively.
Also includes some basic demographic data.
illinoisforecast.com/chicago_ideo...
Also includes some basic demographic data.
illinoisforecast.com/chicago_ideo...
February 15, 2026 at 7:01 PM
New! Ideology map of Chicago. a precinct with a 0 score will (on average) vote exactly like the city. +10 or -10 will vote 10 points more progressive or conservative in the mayoral race respectively.
Also includes some basic demographic data.
illinoisforecast.com/chicago_ideo...
Also includes some basic demographic data.
illinoisforecast.com/chicago_ideo...
The main reason I am hesitant to do this is that, to my knowledge, we don't get favorables with every poll. If we did, I would implement it. While I suspect the Biss internals are close to accurate, I can't guarantee that and I have a lack of supporting evidence.
I have not put favorables in my model and I am wondering if it would change it much.
I suspect it would actually just get more pro-Biss and absolutely nuke Kat’s chances though.
I suspect it would actually just get more pro-Biss and absolutely nuke Kat’s chances though.
Other polling tidbits from Biss’ team
Biss favorables - 59/24
Fine - 46/28
Abughazaleh 36/31
Biss favorables - 59/24
Fine - 46/28
Abughazaleh 36/31
February 14, 2026 at 8:15 PM
The main reason I am hesitant to do this is that, to my knowledge, we don't get favorables with every poll. If we did, I would implement it. While I suspect the Biss internals are close to accurate, I can't guarantee that and I have a lack of supporting evidence.
I guess "eventually" meant less than 2 hours. New update is live that has 2015, 2019, and 2023 mayoral races for both rounds.
Earlier races are harder to find the correct data on.
Earlier races are harder to find the correct data on.
February 14, 2026 at 5:46 PM
I guess "eventually" meant less than 2 hours. New update is live that has 2015, 2019, and 2023 mayoral races for both rounds.
Earlier races are harder to find the correct data on.
Earlier races are harder to find the correct data on.
NEW! Although not polling or forecasting, I have added the results for the 2023 Chicago mayoral race to the website. There is a toggle to see which round you want and it is slightly transparent so that you can see exactly what areas are included.
illinoisforecast.com/Chicago%20Ma...
illinoisforecast.com/Chicago%20Ma...
February 14, 2026 at 4:26 PM
NEW! Although not polling or forecasting, I have added the results for the 2023 Chicago mayoral race to the website. There is a toggle to see which round you want and it is slightly transparent so that you can see exactly what areas are included.
illinoisforecast.com/Chicago%20Ma...
illinoisforecast.com/Chicago%20Ma...
Interesting spike in traffic around 9 am
February 13, 2026 at 6:21 PM
Interesting spike in traffic around 9 am
After the primaries I will take a break and then probably around summer I will start collected data for the Chicago mayoral race.
With the environment, partisan lean of the state as a whole, and the gerrymander the general is basically a foregone conclusion for this race.
With the environment, partisan lean of the state as a whole, and the gerrymander the general is basically a foregone conclusion for this race.
February 13, 2026 at 6:06 PM
After the primaries I will take a break and then probably around summer I will start collected data for the Chicago mayoral race.
With the environment, partisan lean of the state as a whole, and the gerrymander the general is basically a foregone conclusion for this race.
With the environment, partisan lean of the state as a whole, and the gerrymander the general is basically a foregone conclusion for this race.
I have not put favorables in my model and I am wondering if it would change it much.
I suspect it would actually just get more pro-Biss and absolutely nuke Kat’s chances though.
I suspect it would actually just get more pro-Biss and absolutely nuke Kat’s chances though.
Other polling tidbits from Biss’ team
Biss favorables - 59/24
Fine - 46/28
Abughazaleh 36/31
Biss favorables - 59/24
Fine - 46/28
Abughazaleh 36/31
February 13, 2026 at 5:09 PM
I have not put favorables in my model and I am wondering if it would change it much.
I suspect it would actually just get more pro-Biss and absolutely nuke Kat’s chances though.
I suspect it would actually just get more pro-Biss and absolutely nuke Kat’s chances though.
Recently we have had a Fine internal and a Biss internal. Could we maybe also get the Kat internal I have heard she has?
February 13, 2026 at 5:05 AM
Recently we have had a Fine internal and a Biss internal. Could we maybe also get the Kat internal I have heard she has?
Many updates today including a Biss internal poll but also a domain name! Model is live at illinoisforecasting.com
The main page is now a summary page showcasing district wide odds and expected margins as well as how much they changed since the last simulation run.
The main page is now a summary page showcasing district wide odds and expected margins as well as how much they changed since the last simulation run.
February 13, 2026 at 2:41 AM
How does the Biss internal change how the model thinks the gayest precinct votes? a net of 1.7 points in Biss's favor and making him the favorite to win.
February 13, 2026 at 1:47 AM
How does the Biss internal change how the model thinks the gayest precinct votes? a net of 1.7 points in Biss's favor and making him the favorite to win.
Many updates today including a Biss internal poll but also a domain name! Model is live at illinoisforecasting.com
The main page is now a summary page showcasing district wide odds and expected margins as well as how much they changed since the last simulation run.
The main page is now a summary page showcasing district wide odds and expected margins as well as how much they changed since the last simulation run.
February 13, 2026 at 1:26 AM
Many updates today including a Biss internal poll but also a domain name! Model is live at illinoisforecasting.com
The main page is now a summary page showcasing district wide odds and expected margins as well as how much they changed since the last simulation run.
The main page is now a summary page showcasing district wide odds and expected margins as well as how much they changed since the last simulation run.
Left before this poll update
right after
right after
February 12, 2026 at 11:21 PM
Left before this poll update
right after
right after
New update, Changed from neocities to using render. I will eventually buy a domain to use with this website but this makes it look so much nicer because now you can see the map underneath!
il9prediction-and-tracker-1.onrender.com/IL09_precinc...
il9prediction-and-tracker-1.onrender.com/IL09_precinc...
February 12, 2026 at 4:09 AM
New update, Changed from neocities to using render. I will eventually buy a domain to use with this website but this makes it look so much nicer because now you can see the map underneath!
il9prediction-and-tracker-1.onrender.com/IL09_precinc...
il9prediction-and-tracker-1.onrender.com/IL09_precinc...
What I think is rather interesting is that while our models can show some pretty different things, the model @bayespr.bsky.social came up with (left) and mine both seem to agree that there is a large stretch where Sheridan Rd acts as a border between Biss areas and Kat areas in Chicago.
February 12, 2026 at 1:13 AM
What I think is rather interesting is that while our models can show some pretty different things, the model @bayespr.bsky.social came up with (left) and mine both seem to agree that there is a large stretch where Sheridan Rd acts as a border between Biss areas and Kat areas in Chicago.
Where does the vote go in the gayest precinct in the district? The model expects it to be a very close race between Biss and Kat with a slight edge to Kat
February 12, 2026 at 12:09 AM
Where does the vote go in the gayest precinct in the district? The model expects it to be a very close race between Biss and Kat with a slight edge to Kat
I should have gotten a screenshot, but I followed several of the candidates in this race and one followed back:
Laura Fine.
Except, now she has unfollowed this account.
Laura Fine.
Except, now she has unfollowed this account.
February 11, 2026 at 7:21 PM
I should have gotten a screenshot, but I followed several of the candidates in this race and one followed back:
Laura Fine.
Except, now she has unfollowed this account.
Laura Fine.
Except, now she has unfollowed this account.
Minor update to the model that changes a lot. I further discounted weight of internals (down to 50% from 80%) and changed how the house effect was handled in internals. I also boosted a volatility factor I have because I felt like maybe it was too strong.
February 11, 2026 at 5:14 PM
Minor update to the model that changes a lot. I further discounted weight of internals (down to 50% from 80%) and changed how the house effect was handled in internals. I also boosted a volatility factor I have because I felt like maybe it was too strong.
The fix to turnout data is probably the last big change I have for this. I think it is mostly ready unless I change how I want to handle internals. I would like to maybe introduce more uncertainty for internal polls without discounting them entirely.
Also, more candidates should release internals
Also, more candidates should release internals
Huge updates to the model including how it fundamentally works, and some more information on the page.
chicagomaps.neocities.org/IL09_precinc...
chicagomaps.neocities.org/IL09_precinc...
IL-09 Democratic Primary Prediction Model v.0.8
chicagomaps.neocities.org
February 10, 2026 at 8:28 PM
The fix to turnout data is probably the last big change I have for this. I think it is mostly ready unless I change how I want to handle internals. I would like to maybe introduce more uncertainty for internal polls without discounting them entirely.
Also, more candidates should release internals
Also, more candidates should release internals
This file has demographic data on a precinct level, my computed progressive score, probabilities that each candidate will win each precinct and their median result in that precinct.
drive.google.com/file/d/1Cs4l...
drive.google.com/file/d/1Cs4l...
drive.google.com
February 10, 2026 at 4:55 AM
This file has demographic data on a precinct level, my computed progressive score, probabilities that each candidate will win each precinct and their median result in that precinct.
drive.google.com/file/d/1Cs4l...
drive.google.com/file/d/1Cs4l...
Uhhh, realized I was working with inherently flawed turnout data. I ended up fixing it! The map looks mostly the same but Biss does win some more precincts in Chicago.
Huge updates to the model including how it fundamentally works, and some more information on the page.
chicagomaps.neocities.org/IL09_precinc...
chicagomaps.neocities.org/IL09_precinc...
IL-09 Democratic Primary Prediction Model v.0.8
chicagomaps.neocities.org
February 10, 2026 at 4:21 AM
Uhhh, realized I was working with inherently flawed turnout data. I ended up fixing it! The map looks mostly the same but Biss does win some more precincts in Chicago.
The hardest part to make sure is correct in the model is to make sure that turnout is reasonably accurate. Unfortunately, turnout varies wildly.
So another future project is to make sure that I can handle different turnout scenarios.
So another future project is to make sure that I can handle different turnout scenarios.
February 10, 2026 at 2:06 AM
The hardest part to make sure is correct in the model is to make sure that turnout is reasonably accurate. Unfortunately, turnout varies wildly.
So another future project is to make sure that I can handle different turnout scenarios.
So another future project is to make sure that I can handle different turnout scenarios.
I want to emphasize the second half of this. The precinct level shadings are purely for fun. If anything of this has any value (doubtful) it is the districtwide stuff.
Especially since individual precincts can be unpredictable
Especially since individual precincts can be unpredictable
This is probably close to the final version but I may make tweaks here and there as I consider other factors that may be relevant.
Don't take this too seriously. Trying to predict individual precincts is kind of absurd but I wanted to get something that looks plausible for fun.
Don't take this too seriously. Trying to predict individual precincts is kind of absurd but I wanted to get something that looks plausible for fun.
February 10, 2026 at 1:50 AM
I want to emphasize the second half of this. The precinct level shadings are purely for fun. If anything of this has any value (doubtful) it is the districtwide stuff.
Especially since individual precincts can be unpredictable
Especially since individual precincts can be unpredictable
Future projects:
⚫️A function to call on election night
⚫️A function to project final results based on results already in
⚫️more sophisticated logic to simulate different turnout scenarios
⚫️open sourcing the project (this one is honestly the lowest priority)
⚫️A function to call on election night
⚫️A function to project final results based on results already in
⚫️more sophisticated logic to simulate different turnout scenarios
⚫️open sourcing the project (this one is honestly the lowest priority)
February 10, 2026 at 1:44 AM
Future projects:
⚫️A function to call on election night
⚫️A function to project final results based on results already in
⚫️more sophisticated logic to simulate different turnout scenarios
⚫️open sourcing the project (this one is honestly the lowest priority)
⚫️A function to call on election night
⚫️A function to project final results based on results already in
⚫️more sophisticated logic to simulate different turnout scenarios
⚫️open sourcing the project (this one is honestly the lowest priority)
Huge updates to the model including how it fundamentally works, and some more information on the page.
chicagomaps.neocities.org/IL09_precinc...
chicagomaps.neocities.org/IL09_precinc...
IL-09 Democratic Primary Prediction Model v.0.8
chicagomaps.neocities.org
February 10, 2026 at 12:08 AM
Huge updates to the model including how it fundamentally works, and some more information on the page.
chicagomaps.neocities.org/IL09_precinc...
chicagomaps.neocities.org/IL09_precinc...