Mark_C Posted July 6 Author #26 Share Posted July 6 1 minute ago, Waspie_Dwarf said: Not true. The expected result of EVERY constituency is predicted using statistical means based on historic voting patterns. It is for this reason that the predictions can be very accurate nationally, but incorrect at an individual constituency level. Fair enough and its not certain they did or didn't come here after all, this doesn't negate the normally high accuracy of their nationwide predications and their woeful failings here yesterday. 7 minutes ago, Waspie_Dwarf said: I'll give you another example... the constituency of Old Bexley and Sidcup (where I went to school). Widely predicted to be a Labour gain it was actually a Conservative hold. Is this evidence that the Conservatives were cheating? The exit polls are a lot more accurate than opinion polls. Opinion polls gauge intention, exit polls gauge past action. Link to comment Share on other sites More sharing options...
Setton Posted July 6 #27 Share Posted July 6 2 minutes ago, Waspie_Dwarf said: Hartlepool MAY not have been in an area specifically polled, I' don't know for sure, but then neither did the Hartlepool Mail. Even if it was one of the areas it would not have been all polling stations in the area and would only have been a few hundred people asked out of more than 35,000 voters. Yeah I edited my post to reflect this. Thanks. Link to comment Share on other sites More sharing options...
Setton Posted July 6 #28 Share Posted July 6 (edited) 5 minutes ago, Mark_C said: Fair enough and its not certain they did or didn't come here after all, this doesn't negate the normally high accuracy of their nationwide predications and their woeful failings here yesterday. The exit polls are a lot more accurate than opinion polls. Opinion polls gauge intention, exit polls gauge past action. I've looked up the actual exit poll for Hartlepool. Ipsos, based on their limited sampling predicted a 91% chance Reform would win and 9% chance Labour. Let's take another example. You're playing cards and really need an Ace (a 1/13 chance, assuming you're playing with a full deck). You draw an ace. Should those playing with you assume you cheated? Edited July 6 by Setton 1 Link to comment Share on other sites More sharing options...
Saru Posted July 6 #29 Share Posted July 6 @Mark_C - just taking a brief glance at some other constituencies: Moray West, Nairn and Strathspey: Exit poll prediction: 95% conservative. Actual result: SNP Waveney Valley: Exit poll prediction: 99% conservative. Actual result: Green So are the Greens and SNP "cheating" as well ? And there are many more examples. The only issue here is with your understanding of exit polls. 3 Link to comment Share on other sites More sharing options...
Mark_C Posted July 6 Author #30 Share Posted July 6 15 minutes ago, Setton said: I've looked up the actual exit poll for Hartlepool. Ipsos, based on their limited sampling predicted a 91% chance Reform would win and 9% chance Labour. Let's take another example. You're playing cards and really need an Ace (a 1/13 chance, assuming you're playing with a full deck). You draw an ace. Should those playing with you assume you cheated? Thanks for your integrity, its not proof of anything but it does raise eyebrows. Link to comment Share on other sites More sharing options...
Mark_C Posted July 6 Author #31 Share Posted July 6 (edited) 7 minutes ago, Saru said: The only issue here is with your understanding of exit polls. Id say the same, if any cheating took place it was to allow the Labour party to win. So the 2 statistics you provided only add credence to that. Edited July 6 by Mark_C Link to comment Share on other sites More sharing options...
Setton Posted July 6 #32 Share Posted July 6 Just now, Mark_C said: Thanks for your integrity, its not proof of anything but it does raise eyebrows. What are you even on about? Let's try another example. You have a bag of 10 marbles. 9 green, 1 blue. I won if I pull out the blue marble. Assuming you haven't lost any, that's a 1 in 10 chance I win. If I do pick the blue marble, would you assume I cheated? Would it even "raise eyebrows"? Link to comment Share on other sites More sharing options...
Mark_C Posted July 6 Author #33 Share Posted July 6 Just now, Setton said: What are you even on about? Let's try another example. You have a bag of 10 marbles. 9 green, 1 blue. I won if I pull out the blue marble. Assuming you haven't lost any, that's a 1 in 10 chance I win. If I do pick the blue marble, would you assume I cheated? Would it even "raise eyebrows"? No it wouldn't be proof of cheating but it would be highly improbable for you to pull out the blue one. Link to comment Share on other sites More sharing options...
Mark_C Posted July 6 Author #34 Share Posted July 6 14 minutes ago, Saru said: And there are many more examples. Are there any examples of conservatives winning where they'd been predicted to lose? Link to comment Share on other sites More sharing options...
Setton Posted July 6 #35 Share Posted July 6 4 minutes ago, Mark_C said: No it wouldn't be proof of cheating but it would be highly improbable for you to pull out the blue one. No, it would be a one in ten chance. If ten of us played the same game, statistically one would win. If 650 of us play the same game, lots of us should win. By extrapolation, it would be more strange if Ipsos predicted every seat correctly. Link to comment Share on other sites More sharing options...
Setton Posted July 6 #36 Share Posted July 6 4 minutes ago, Mark_C said: Are there any examples of conservatives winning where they'd been predicted to lose? Stockton West. 99% chance Labour gain in the exit poll. Conservatives held the seat. OMG CHEATING TORIES 😮 1 Link to comment Share on other sites More sharing options...
Saru Posted July 6 #37 Share Posted July 6 (edited) 14 minutes ago, Mark_C said: Id say the same Please feel free to elaborate on what you feel I'm not understanding about exit polls. Quote if any cheating took place it was to allow the Labour party to win. An absolutely baseless statement. Quote So the 2 statistics you provided only add credence to that. They do absolutely nothing of the sort. 6 minutes ago, Mark_C said: Are there any examples of conservatives winning where they'd been predicted to lose? Waspie provided one already: Old Bexley and Sidcup: Exit poll prediction: 97% Labour. Actual result: Conservative. Edit: And Setton just posted another one above. Edited July 6 by Saru Link to comment Share on other sites More sharing options...
Mark_C Posted July 6 Author #38 Share Posted July 6 (edited) 10 minutes ago, Setton said: No, it would be a one in ten chance. If ten of us played the same game, statistically one would win. If 650 of us play the same game, lots of us should win. By extrapolation, it would be more strange if Ipsos predicted every seat correctly. Ok lets look at my previous quote again"In the past five British general elections, the exit poll has predicted how many of the 650 or more parliamentary seats would be claimed by the winning party to within an average of four seats. Last time, in 2019, it had the winning party’s total just three seats out" That means they are normally 99.4 % accurate. In Hartlepool the person they predicted would win lost by 100% ( ish ). 5 minutes ago, Saru said: They do absolutely nothing of the sort. They do, its a 2 horse race. The SNP and the greens are non runners. 5 minutes ago, Saru said: Waspie provided one already: Those were opinion polls not exit polls. Edited July 6 by Mark_C added ish Link to comment Share on other sites More sharing options...
Setton Posted July 6 #39 Share Posted July 6 Just now, Mark_C said: In Hartlepool the person they predicted would win lost by 100% Not true. The prediction was 91%. Link to comment Share on other sites More sharing options...
Setton Posted July 6 #40 Share Posted July 6 11 minutes ago, Mark_C said: Are there any examples of conservatives winning where they'd been predicted to lose? Keighley and Ilkley. 99+% Labour gain, Tories held. Link to comment Share on other sites More sharing options...
Mark_C Posted July 6 Author #41 Share Posted July 6 Just now, Setton said: Keighley and Ilkley. 99+% Labour gain, Tories held. I dont understand this language, are these exit poll results or opinion poles versus results Link to comment Share on other sites More sharing options...
Saru Posted July 6 #42 Share Posted July 6 4 minutes ago, Mark_C said: They do, its a 2 horse race. The SNP and the greens are non runners. The SNP and Greens winning seats is not evidence that Labour is cheating! Quote Those were opinion polls not exit polls. No, I double-checked those figures myself - they were from the actual exit poll. Check it for yourself: https://inews.co.uk/news/uk-general-election-2024-results-map-exit-poll-3151119 Link to comment Share on other sites More sharing options...
Saru Posted July 6 #43 Share Posted July 6 4 minutes ago, Mark_C said: I dont understand this language, are these exit poll results or opinion poles versus results It's directly from the exit poll. The prediction was 99+% likely to be Labour and it went to the Conservatives. Link to comment Share on other sites More sharing options...
Setton Posted July 6 #44 Share Posted July 6 1 minute ago, Mark_C said: I dont understand this language, are these exit poll results or opinion poles versus results 🤦♂️ Exit polls and results, like you asked for. So I've got two where the exit poll gave 99+% chance of Labour win and Tories actually won. You've got one and only a 91% chance at that. Will you now admit you were talking rubbish or will you admit that if there was any cheating (there wasn't), clearly it was to benefit the Tories? Link to comment Share on other sites More sharing options...
Mark_C Posted July 6 Author #45 Share Posted July 6 3 minutes ago, Saru said: The SNP and Greens winning seats is not evidence that Labour is cheating! No, I double-checked those figures myself - they were from the actual exit poll. Check it for yourself: https://inews.co.uk/news/uk-general-election-2024-results-map-exit-poll-3151119 Its ok I believe you but fyi that link takes me to a paywall. Link to comment Share on other sites More sharing options...
acute Posted July 6 #46 Share Posted July 6 The punchline to this thread is.... You don't understand how exit polls work! The results are a prediction, based on samples and a bunch of local stats. They are not applicable to every constituency. 1 Link to comment Share on other sites More sharing options...
Setton Posted July 6 #47 Share Posted July 6 1 minute ago, Mark_C said: Its ok I believe you but fyi that link takes me to a paywall. https://news.sky.com/story/exit-poll-what-is-the-forecast-election-result-in-my-constituency-13163180 Link to comment Share on other sites More sharing options...
Mark_C Posted July 6 Author #48 Share Posted July 6 Just now, acute said: The punchline to this thread is.... You don't understand how exit polls work! The results are a prediction, based on samples and a bunch of local stats. They are not applicable to every constituency. The crunch line is that they're usually about 99.4% accurate and this time they were "anomalous" Link to comment Share on other sites More sharing options...
Mark_C Posted July 6 Author #49 Share Posted July 6 10 minutes ago, Saru said: It's directly from the exit poll. The prediction was 99+% likely to be Labour and it went to the Conservatives. I can see that, it was pretty close though in the end, unlike the Hartlepool result. Link to comment Share on other sites More sharing options...
Setton Posted July 6 #50 Share Posted July 6 5 minutes ago, Mark_C said: The crunch line is that they're usually about 99.4% accurate and this time they were "anomalous" Oh FFS. The poll is 99.4% accurate AT THE NATIONAL LEVEL. This year it was accurate AT THE NATIONAL LEVEL Link to comment Share on other sites More sharing options...
Recommended Posts