- 00:44:50 Retweeted @jbarro 00:40:41 On the bright side, we got through this thing without having to see a Trump sex tape or dick pic.
- 00:46:13 .@jbarro Please don't say this. You are tempting fate. There are still another 47 hours until polls start to close… in reply to jbarro
- 00:47:46 Retweeted @imbou 00:44:29 @abulsme so why are these bozos publishing a poll they don't believe????? https://t.co/SbNybbxsMY
- 00:48:57 .@imbou Would be intellectually dishonest to stop something after you had started just cause you don't like results. Gotta carry it through. in reply to imbou
- 00:50:31 Retweeted @EmilyJillette 00:50:17 How have I gone all day without mentioning how stupid DST is? And time zones. #OneWorldOneTime
- 00:54:46 Retweeted @imbou 00:54:25 @abulsme there is not liking the results and believing they are inaccurate.If you believe it's wrong don't you revise the methodology? in reply to abulsme
- 00:55:28 .@imbou For next time. Sure. But changing during an election cycle just invites criticism that you are adjusting 2 make a result you prefer. in reply to imbou
- 00:55:38 Retweeted @imbou 00:54:54 @abulsme i think it's far more intellectually dishonest to continue not publish something you know is wrong… in reply to abulsme
- 01:00:53 Retweeted @imbou 00:57:13 @abulsme another case of doing something for appearances and covering your ass instead of doing the right thing… in reply to abulsme
- 01:01:44 .@imbou Eh. You could maybe spin up a 2nd analysis correcting things you thought needed to be corrected, but should keep the first for comp. in reply to imbou
- 01:02:38 .@imbou But LAT have defended their methodology, and might not even agree that they know they are doing something wrong. in reply to abulsme
- 01:03:10 .@imbou Maybe map done by different group, or group using all available date including data other than their own. in reply to abulsme
- 01:03:56 .@imbou Just because you are an outlier doesn’t make you “wrong” necessarily, just more data to feed into meta-models. in reply to abulsme
- 01:04:05 Retweeted @imbou 01:03:03 @abulsme how about analyzing the data figuring out the mistake and starting a new series? in reply to abulsme
- 01:04:45 .@imbou Sure. For 2020. Won’t know how far off (if at all) they were from reality until after the election. Changing before then premature. in reply to imbou
- 01:05:47 .@imbou For all we know, they may end up being the only ones to get it right. We’ll know in a couple days. in reply to abulsme
- 01:06:18 Retweeted @imbou 01:04:19 @abulsme the published an electoral vote forecast that is diametrically opposed to the poll? How louder could they say we don't buy out poll in reply to abulsme
- 01:07:41 .@imbou When I give my opinions on what will happen, Election Graphs isn’t the only place I look. If I was way off from everyone else… in reply to imbou
- 01:08:46 .@imbou It would be dishonest & wrong 4 me 2 tweak how I do it until I agree w/ everybody else. Would need to stick to it, change next time. in reply to abulsme
- 01:10:56 Retweeted @imbou 01:10:44 @abulsme I wholeheartedly don't agree.When I did a financial forecast and I thought it was wrong I fixed it or pulled it in reply to abulsme
- 01:13:51 .@imbou Depends on the goal of what you are doing. I’d argue that an internal financial forecast is very different. in reply to imbou
- 01:14:14 Retweeted @peterb 01:12:47 @imbou @abulsme I think this is legit; the poll is data for your forecast, but it's not "the forecast". in reply to imbou
- 01:14:28 Retweeted @peterb 01:13:16 @imbou @abulsme That said i think if you're NOT using your poll as the basis of your forecast, you should be saying what the basis IS. in reply to peterb
- 01:15:09 .@peterb @imbou Agreed. LAT should explain why their map doesn’t in any way match what their tracking poll says. :-) in reply to peterb
- 01:16:13 .@peterb @imbou When I talk about where I think the numbers coming out of @ElectionGraphs aren’t the best picture, I try to explain why. in reply to abulsme
- 01:17:24 .@peterb @imbou Usually like “Right now KS looks close, but it is only cause of that crazy Google outlier poll, so treat with lots of salt.” in reply to abulsme
- 01:19:10 Retweeted @peterb 01:18:50 @EmilyJillette @abulsme We'll always have Swatch Beats. in reply to EmilyJillette
- 01:32:39 Retweeted @imbou 01:31:25 @abulsme I didn't mean an internal forecast. I used to publish external forecasts and if I was wrong which happened i corrected them in reply to abulsme
- 01:32:40 Retweeted @imbou 01:31:51 @abulsme sometimes you may have a methodology you pulled together which on paper is sound but in practice is flawed in reply to abulsme
- 01:32:44 Retweeted @imbou 01:32:06 @abulsme you need to find the flaw and fix it or pull the damn model in reply to abulsme
- 01:34:03 Retweeted @imbou 01:33:06 @abulsme not just keep publishing the model and then publish an explanation that goes against your forecast data! in reply to abulsme
- 01:34:24 .@imbou If you actually believe and have evidence you are doing something WRONG that is different. in reply to imbou
- 01:35:07 .@imbou If you just now doubt your results because they don’t match everybody else, that is not enough. in reply to abulsme
- 01:35:45 .@imbou LAT has defended their method. They don’t believe they are fundamentally doing something wrong, even if they are an outlier. in reply to abulsme
- 01:35:53 Retweeted @imbou 01:35:18 @abulsme well they seem to think it is! Otherwise then why the hell did they just publish that EV forecast that goes against it? in reply to abulsme
- 01:37:02 .@imbou Cause their poll is one data point of many. No reason to expect a match unless they specifically said fcast was based only on poll. in reply to imbou
- 01:37:51 Retweeted @imbou 01:37:36 @abulsme well then as @Peterb said they need to explain in reply to abulsme
- 01:37:58 .@imbou @peterb Agreed. in reply to imbou
- 17:05:35 Retweeted @stevelevine 17:03:13 Forecasting split on the Senate: @FiveThirtyEight: 54% chance GOP holds it. @UpshotNYT: 56% Dems take it. Princeton's @SamWangPhD: 50-50 tie
- 17:12:29 Retweeted @DKElections 2016-11-02 15:42:30 You will definitely want to bookmark our one-of-a-kind beautifully color-coded poll closing times map:… https://t.co/VpyhZfk7ml in reply to DKElections
- 17:15:49 Retweeted @FranPelayo 16:00:22 +453.8K Latinos voted early in Florida, that's an increase of 86.9% compared to 2012
- 17:26:34 Retweeted @LarrySabato 17:23:50 Our final Crystal Ball picks:
Clinton 322, Trump 216
50-50 Senate
GOP holds House https://t.co/Ta6ITI363x - 18:16:48 Retweeted @kkondik 18:16:34 Final Crystal Ball picks — https://t.co/LEXiXNSJ6I https://t.co/GcOkbui8vS
- 18:19:26 Retweeted @xkcdclone 16:58:17 I'm With Her. https://t.co/cLqlEJcctF We can do this. #xkcd https://t.co/M2YcPQmxOW
- 18:58:07 On fire once again -> I'm voting for Hillary Clinton — and against Donald Trump (@ananavarro) https://t.co/FOvfxftW1O
- 20:23:45 Reading – Glenn Beck Tries Out Decency (Nicholas Schmidle) https://t.co/twOl5WLSMl
- 20:24:11 Retweeted @Rainmaker1973 18:39:05 #Phobos over the #Mars' limb as seen by @esa Mars Express https://t.co/MrFZI6iF0B #MoonCrushMonday https://t.co/9WbYQEy7bq
- 20:25:51 Retweeted @NishaChittal 20:23:43 https://t.co/F7lHeY6GTE
- 20:31:39 Reading – Woman who claimed she was assaulted by Trump at a sex party at age 13 FABRICATED her story (Ryan Parry) https://t.co/UJkXNrlJX0
- 20:31:55 Retweeted @DixvilleVote 20:23:17 #DixvilleVotesatMidnight #Election2016 https://t.co/uPPS7fa0JS
- 20:37:50 Retweeted @davidbylerRCP 19:30:13 Basically everyone is predicting an HRC win. There are fun details, but it feels like most of wonk world's pre-election work here is done
- 20:37:53 Retweeted @davidbylerRCP 20:19:50 And I should note – I don't mean every. single. person. I was referring to wonk world – data journalists and other empirical types in reply to davidbylerRCP
- 20:42:26 Retweeted @jimsciutto 18:29:54 Just in: Florida early voting surpasses entire turnout in disputed 2000 election:
2016 – 6,424,595
2000 – 5,861,223
#Election2016 - 20:49:10 Reading – Australian Police Search Woman During Arrest, Find A Baby Koala In Her Bag (AP) https://t.co/J58Nkxbd89
- 21:12:33 Retweeted @laloalcaraz 20:16:24 BY POPULAR DEMAND: The Wall Cartoon, please share and VOTE https://t.co/WCLkF0K8iF
- 21:31:40 Retweeted @KenTremendous 21:01:40 Two days before the election Trump just plum *made up an award,* said he won it, and no one even blinked. We have i… https://t.co/90C9tACzG0
- 21:56:13 Reading – Donald Trump’s success reveals a frightening weakness in American democracy (@ezraklein) https://t.co/XsAxz9xDz9
- 22:05:25 Seven hours until Dixville Notch!
- 23:03:55 24 hours until the first polls in Indiana and Kentucky close. (Still 25 hours until the first full states close.)
- 23:24:57 Retweeted @RadioFreeTom 23:24:45 Trump now attacking all media as nothing but lies. Because you should get your news from Facebook memes.
- 23:29:25 Just to clarify on article below with misleading headline, the woman said she didn’t know who it was at the time, recognized Trump on TV. in reply to abulsme
Leave a Reply
You must be logged in to post a comment.