@RealCarlAllen

Educator, Scientist. Sports & Political Data Analyst. Other social media: RealCarlAllen 馃摉 Book: *The Polls Weren't Wrong* out now

 Ohio, USA
realcarlallen.substack.com/
Joined January 2013
  • Tweets68.5K
  • Following5.4K
  • Followers5.6K
16.7KMedia
I fully expect DMs, EMs, etc to blow up in the next 48 hrs If I don't respond, don't worry. I know who has supported my work and I won't lose sight of that If you'd like to join me in developing a set of unified scientific standards for this field, my book is available in bio
11
1
69
50,150
5 events with probability 83% happen Silence 1 event with probability 17% happen *Uncontrollable screeching* I guess statistical illiteracy keeps me employed
115
Behold The Prediction Market Hater Paradox
116
Just a reminder for the statistical illiterates that if prediction markets assign the correct probability, then events they assign a 15% probability to will happen between 1/6 to 1/7 times
Mark Davin Harris
@markdharris1h
Just a reminder for all the polymarket is destiny stuff that Lula was at 15% the day before the election.
1
1
2
248
"Poll error in Brazil" No there wasn't. Poll aggregators who didn't lie and dismiss undecideds had Bolsonaro around 48. If you compare poll margin to election margin and call the discrepancy "poll error" you are fundamentally ignorant of what polls measure
2
11
1,859
I didn't have a stake in Brazil elections. I did not build a forecast, and I'd say markets pricing Bolsonaro as a medium favorite were reasonable But here were the results from R1 Lula + left wing parties (PSD, Avante) = 51 Bolsonaro + right wing parties (Mission, Novo) = 49
1
122
Unless Republicans win by 18 nationwide, this poll is wrong If you can explain why that is flawed logic, you understand why the assertion that undecideds will split evenly to major candidates (and any discrepancy is a poll error, making the poll wrong) is also flawed.
1
467
What did Carl teach you about high undecided poll data?
2
4
1,168
The direction of photographer error in recent years: *2004: even bigger anti-US bias *2008: huge anti-US bias *2012: anti-US bias *2016: unbiased *2020: US bias *2024: anti-US bias
Nate Silver
@NateSilver538Oct 1
The direction of polling error in recent years: * 2008: Unbiased * 2010: Unbiased * 2012: R bias * 2014: D bias * 2016: D bias * 2018: Unbiased * 2020: Even bigger D bias * 2022: Unbiased * 2024: D bias, less than 2016/2020 * 2025 (e.g. NJ/VA): Big R bias
715
I think the Dem margin in North Carolina will be larger than the Republican margin in Mississippi, and maybe Florida.
3
1
1
25
1,554
The hardest throwing starting pitcher in the history of baseball is also the best pitcher in baseball No wonder boomers are so mad!
1
698
And one of the fastest players in history with 0 career home runs batting 3rd 馃ぃ
343
There are a lot of possibilities this election. I'll divide them into 4 ~equally likely bins 1) Rs hold Senate: narrow wins in notoriously purple states like.. Uhh Texas, Ohio, Kansas, and Iowa Leaves Rs taking a victory lap and Nate Silver loudly declaring "polls was wrong!"
7
1
1
64
89,070
more replies
3) Dems win all states they're favored in, plus KS and/or IA. 4) Dems win all of the above, plus some surprising state(s) like Nebraska, South Carolina in which case polling is dismissed as irredeemably broken I personally think 1 is more likely than 3, but that doesn't matter
3
1
13
12,914
What matters here is - even in generic ranges - no one KNOWS what will happen Lots of people will try to convince you otherwise. Strictly by chance (about 1/4 in fact) some of them will guess correctly Don't be tricked. Smart people quantify uncertainty, they don't wish it away
4
1
24
11,790
People who don't know how to do math crack me up 馃ぃ Watch this body 馃憞 3/5 (even 3 singles) is a .600 OBP, .527 wOBA, 246 wRC+ 1/5 with a HR is a .200 OBP, .403 wOBA, 159 wRC+ Nerds take 3/5 every time. Higher wOBA, higher wRC+. Proved him wrong again #shegone
Fryedaddy/Frito
@shegone03Oct 23
You #nerds sure hate Double Plays don鈥檛 you? You鈥檇 rather a guy go 1 for 5 with a homer and 3 K鈥檚 than a guy go 3 for 5 and hit into a Double Play. Cracks me up! #shegone
1
1
1,200
Carl Allen retweeted
Carl Allen@RealCarlAllenOct 23
Replying to @TheOneCA2021
And the people who calculate "poll accuracy" assume unconditionally across all election types and state makeup that undecideds will always split evenly to the major candidates.
1
324
It makes me SO HAPPY that my mentions blow up whenever someone with a following says something about "Hey, it's kind of garbage math to assume undecideds always split unevenly" Let me inspire you: There's a whole world of discovery waiting once you understand how bad it is.
1
2
13
1,343
If you work in media, political science, or statistics, I'm happy to chat. Depending on your background and comfort with science, I can get you totally caught up with what's going on in 15-45 minutes I'll send you a book, explain my research, whatever you want.
1
3
651
The assumption that poll averages translate to win probabilities illustrates a fundamental misunderstanding of how polls work
Zachary Donnini
@ZacharyDonniniOct 22
Kalshi probabilities and our polling averages have largely converged, except in Ohio, Iowa, Kansas, and Nebraska, where fundamentals are considerably more favorable to Republicans than polling suggests. These states also have histories of polling errors or limited polling data.
1
1
9
1,239
A margin of x% simply does not unconditionally yield a win probability of y% I understand that Nate Silver's probit regression (that said someone has a 99% chance to win if up by 6...80% chance if up by 2) LOOKS appealing But stats aren't always so simple.
313
Load more