prev <-> next
[2019-04-02 11:26:20] - ? -Daniel
[2019-04-02 11:26:11] - a: Have you offered to pay any of your peeps in btc
[2019-04-02 11:02:45] - a: Yeah, it's interesting to think about what the narrative around bitcoin might be if there wasn't that huge spike and collapse because the long term view is still a pretty steady and rapid appreciation. -Paul
[2019-04-02 10:50:43] - paul: right *before* the jump last night, i was having a long-term look at the price too. i wrote this down: "the price bitcoin doubles every ~9 months on average, and has been for 9+ years." kinda crazy. (this also means that bitcoin goes 10x ever 2.5 years. same concept, just with 10x instead of 2x: ln(10)/ln(2)*9/12=2.5) ~a
[2019-04-02 10:03:17] - a: I thought of this mostly because of the BTC jump overnight and how it's still on track for some solid growth for 2019 if you look at the "minimum price" metric. -Paul
[2019-04-02 09:55:12] - a: I saw on Twitter awhile ago that an interesting way to look at bitcoin is the yearly lows. If you look at it that way, BTC growth is something like $400 -> $800 -> $3k over the past few years. That's really incredible growth. I know it's an incredibly biased way of looking at it, but it does help to put some perspective in the growth it has seen. -Paul
[2019-04-01 11:22:51] - aaron: i found a way to increase my win-rate: hybrid-aaron-method! i use my old-method: recurse down some. then use the aaron-method as well for speed! it slows down things a lot, from your method, but i noticed a small increase in my win-rate. 1 level recursion: 32% to 35%. ~a
[2019-04-01 10:25:05] - aaron: changing it to a 6-card-hand, though did decrease my win-rate below 32%. i wonder if i'm doing something else wrong. ~a
[2019-04-01 10:21:25] - aaron: changing it to house-6-card-hand brought me down to 2m runtime for 10k-games (50-iterations). maybe i should try to remove all of the modulus and divisions. ~a
[2019-04-01 10:17:15] - aaron: i stored each card as a single integer. that's probably dumb, because i end up integer-divide and modulus-ing it all over the freak-en place. ~a
[2019-04-01 10:13:51] - aaron: 37 seconds, wow! let me try the 6-card method and i'll let you know my new run-time for 10k, but i doubt it'll be 37 seconds. ~a
[2019-04-01 10:01:26] - aDaniel: Still working on writing up the quarterly freedom portfolio recap, but I ran the numbers last night and the FP is officially back to beating the S&P since inception of the FP (10/1/2018). It's -2% to -3% but it's still winning.
-Paul[2019-04-01 09:40:38] - although i don't know, i guess the most efficient storage method i can think of is something like integer arrays where each integer would represent a card... and a java string is basically the same thing. maybe it's not as bad as it seems - aaron
[2019-04-01 09:38:41] - with 50 iterations per card and 10,000 games, my program took about 37 seconds to run and went 3229/6770/1. ...my program actually seems insanely inefficient, it stores hands as strings "KC 6S 3D 2D 6D JS TC QH 2C TS" and simulating different hands involves thousands of string concatentations - aaron
[2019-04-01 09:35:36] - a: no, 6 cards was accurate. letting the house draw up to 10 cards was too pessimistic for the early stages -- it would estimate things like, "well i have a pair of aces and the house has a 9 high, so i have a 6% chance of winning" - aaron
[2019-03-30 13:51:40] - aaron: "draws until the house has six cards" do you mean 10 cards? my method was too slow so i tried your method: i also got ~32%! my version of your algorithm is maybe slower than yours: it took me just under 5m to do 10,000 games with the 50 iterations for each drawn card. what about you? ~a
[2019-03-29 15:17:33] - aaron: hah. i guess ties are less likely than texas hold-em because there are fewer shared cards (i.e. zero). ~a
[2019-03-29 15:13:43] - a: a tie is unlikely but i currently just track them separately; wins/losses/ties. my last run of 10,000 games went 3178/6822/0, although i have occasionally seen 1 or 2 ties mixed in - aaron
[2019-03-29 14:42:58] - 🂡🂢🂣🂤🂥🂦🂧🂨🂩🂪🂫ðŸ‚🂮🂱🂲🂳🂴🂵🂶🂷🂸🂹🂺🂻🂽🂾ðŸƒðŸƒ‚🃃🃄🃅🃆🃇🃈🃉🃊🃋ðŸƒðŸƒŽðŸƒ‘🃒🃓🃔🃕🃖🃗🃘🃙🃚🃛ðŸƒðŸƒž ~a
[2019-03-29 14:41:01] - aaron: also i liked this: Playing cards in Unicode#Playing cards deck. makes printing out the cards fun!
~a[2019-03-29 14:36:19] - oh i missed the part about "drawing until it has 5 cards". i've been simulating out the whole rest of the game. so i guess it's not exactly the same. ~a
[2019-03-29 14:34:40] - aaron: that's interesting, it's the exact algorithm i'm writing haha. (i was calling it "double monte-carlo" earlier) how does the game rules deal with a split-pot? does that count as a player win or a house win? or neither? ~a
[2019-03-29 14:02:20] - i think it underestimates the chances of straights and flushes, but it seems to make slightly better choices than i do, overall. it would of course be a better AI if it predicted what would happen with the actual rules, instead of just simulating an easier-to-simulate game. if you can beat 32.7% i'll be curious what your algorithm is - aaron - aaron
[2019-03-29 13:59:26] - a: the algorithm is to repeatedly take the card with the best win percent. it estimates this by adding a card to its hand, then drawing until it has 5 cards. then it adds the remaining choices to the house's hands, and draws until the house has six cards. it does this 50 times for each card, and sees which card wins the most. - aaron
[2019-03-29 13:44:30] - aaron: what is the non-foreknowledge algorithm? ~a
[2019-03-29 13:40:58] - aaron: Ah, wow, so my guess without foreknowledge was close-ish, but my guess with foreknowledge was way off. -Paul
[2019-03-29 13:38:40] - i would have never guessed the win chance varied THAT much with perfect foreknowledge, my gut feeling was it would take you from something like 20% -> 40% win rate, since the game seemed to so heavily favor the house, and 10 random cards almost always had a random straight or flush. ...but, it turns out there's usually a way to win if you know what's going to come out - aaron
[2019-03-29 13:36:17] - a: i created a rudimentary AI for the 1-2-3-4-5 poker game -- it wins 32.7% of the time. with perfect foreknowledge, it wins 75.5% of the time but this involves doing a lot of unsmart things, like taking a 7 over an Ace for your second card to prevent the house from getting a straight at the end - aaron
[2019-03-29 13:02:35] - paul: it's ok to support a white male regardless. ~a
[2019-03-29 12:35:18] - https://slate.com/news-and-politics/2019/03/pete-buttigieg-gay-diversity-white-male-candidate.html "Is Pete Buttigieg Just Another White Male Candidate, or Does His Gayness Count as Diversity?" A question to the liberals here (or, I suppose, anybody here): Is it okay to support a white male as long as he is gay?
-Paul[2019-03-28 15:35:33] - i was thinking if this were a game in vegas, they'd presumably have payouts based on how good a hand you won with. winning with a full house, that pays 5:1! winning with a straight flush, 20:1! ...but unintuitively, the payout table should be reversed. winning with a flush is trivial, winning with a pair is unlikely and winning with a high card should pay out like 100,000,000:1, the odds are astronomical - aaron
[2019-03-28 15:23:07] - aaron: i can't go through every possibility. it'll be about pick(52, 15) which is like 1e24. too many! it might actually be closer to choose(52,1)*choose(52-1,2)*choose(52-1-2,3)*choose(52-1-2-3,4)*choose(52-1-2-3-4,5) which is like 1e20. ~a
[2019-03-28 15:03:26] - aaron: With perfect foreknowledge? Maybe 60%? I feel like it would be a lot worse without that knowledge, though. I was originally going to say 75%... -Paul
[2019-03-28 15:01:37] - aaron: count out the possibilities i think will also work to determine the correct answer without foreknowledge. ~a
[2019-03-28 15:01:28] - aaron: perfect foreknowledge won't get you the right answer though.
~a[2019-03-28 15:01:28] - aaron: perfect foreknowledge won't get you the right answer though.
~a[2019-03-28 15:00:55] - paul: do you have any sort of blind guess as to what kind of odds the house has? do you think the house wins like 48% of the time (like blackjack) or is it more like 70%-80% of the time? - aaron
[2019-03-28 15:00:01] - a: i calculated an upper bound for the win probability by assuming the player has perfect foreknowledge -- i just naively brute force all 120 choices and see if the player wins any of them. figuring out the optimal strategy, and what the odds should be for a smart player who lacks foreknowledge, is a more difficult but more interesting problem - aaron
[2019-03-28 14:57:38] - aaron: i guess instead of double-monte-carlo, you can just count out all the possibilities. ~a
[2019-03-28 14:57:32] - aaron: I'm going to guess that the house has the odds, but I did basically no work. Pure intuition. -Paul
[2019-03-28 14:56:23] - aaron: how did you calculate an answer? ~a
[2019-03-28 14:55:14] - paul: you get five cards (1 + 1 + 1 + 1 + 1) and the house gets ten cards (0 + 1 + 2 + 3 + 4), you're both trying to make the best five card poker hand you can - aaron
[2019-03-28 14:54:41] - for example, you might get dealt the (2H), (8H AS), (AC JH 9S), (8C JC 8S 5H), and (QH KH 3C 9D JD) and win -- keeping the 2H, 8H, JH, 5H and KH gives you a flush! ...the following game, you might get dealt the (6C), (8C 3H), (5H 9C 9D), (KH 6S AC KD), and (6D 4S KS 5S 6H) and lose -- your trip 6s lose to the dealer's full house, or you can take one of their kings but your measly pair of 6s loses anyways - aaron
[2019-03-28 14:54:15] - aaron: So the house ends up with nine cards to make a hand from? -Paul
[2019-03-28 14:45:01] - the ten cards you get rid of form the house's poker hand -- and the goal is to beat their hand. is this a good bet? can you win more than half the time? and if not, what kind of odds should the house give you to make it a fair game? (e.g 3:1 odds, 5:1 odds) - aaron
[2019-03-28 14:43:59] - a: i've calculated the answer to our poker question yesterday. but for everyone else: you play a poker game against the house, where you gradually accumulate a poker hand. first, you're dealt one card which is yours. then you're dealt two new cards, and you keep one of those two ones. then you're dealt three cards, and you keep one of those three new ones, and so on up to five cards, to form a five card poker hand - aaron
[2019-03-28 14:25:03] - a: " i don't mean to dissuade you from using individual transactions" vs " in short, because of 1-4, ignoring individual trades is best" HMMMM -Daniel
[2019-03-28 14:23:14] - anyways, i don't mean to dissuade you from using individual transactions: it's literally the only way to get an exact answer. but i warn you that it is *hard* and my bar for doing *hard* things when it comes to financial math is pretty high.
~a[2019-03-28 14:18:58] - paul: more information: search for "approximate return equation" in that page. i forget where i found the 33% part, but it's somewhere on that page. ~a
[2019-03-28 14:15:50] - paul: i ignore all quarters where the amount added (or removed) is more than 33% of your starting position. ~a
[2019-03-28 14:15:17] - paul: no not at all. but you might not like what i *do* do: (assuming this was your first time buying disney) since most of your position of disney was bought in Q1, you should ignore Q1 entirely.
~a[2019-03-28 14:08:49] - a: So if you bought Disney on 1/1 and 3/31 they would both be compared to the S&P500 return for Q1? -Paul
[2019-03-28 14:01:19] - paul: i separate out time periods: i can look at information yearly or quarterly. i wouldn't look at a "disney purchase". instead i'd look at the "disney position" quarterly or yearly and ignore any specific trades. ~a
[2019-03-28 13:59:16] - a: But... maybe we're using different terminology, what would you call the Disney purchase? Isn't that an individual transaction? What if you bought half in 2010 and half in 2015? Those are different time periods. -Paul
[2019-03-28 13:50:58] - which . . . again . . . is easier if you ignore individual transactions
~a[2019-03-28 13:50:39] - paul: it does. i always subtract it for the matching time period. ~a
[2019-03-28 13:49:20] - a: "i subtract the s&p500 from all performances" The performance of the S&P over what time period? Doesn't it have to be matching time periods? -Paul
[2019-03-28 13:46:24] - paul: which will mean that bond performance being negative is . . . expected. ~a
[2019-03-28 13:46:06] - paul: yes, i still compare to the s&p500. basically i subtract the s&p500 from all performances. so everything becomes relative to the s&p500. ~a
[2019-03-28 13:35:31] - a: "you can look at your balance in DIS, and compare it to your old balance in DIS" How is that any different than just looking at the stock price (assuming no splits and no dividends, I guess)? That still doesn't offer any comparison to the S&P at all. Saying Disney is up 80% is only half the story when the S&P500 is up 100% (picking a random number there) in the same time. -Paul
[2019-03-28 13:27:52] - paul: "I think Disney is up pretty big for me in absolute terms since I've held it a long time, but is trailing the S&P", you can look at your balance in DIS, and compare it to your old balance in DIS, also taking into account any money you bought or sold in DIS. ~a
[2019-03-28 13:26:34] - paul: 5. in short, because of 1-4, ignoring individual trades is best. ~a
[2019-03-28 13:26:30] - paul: 4. matching sell orders and buy orders up is easy, but determining the performance requires you to know the "cost basis" (the price per share you paid for a thing). this is hard when you have multiple buys and sells on the same symbol: the methods for determing cost basis are all similar, but come up with *very* different results for performance. ~a
[2019-03-28 13:26:24] - paul: 3. this %/year "unit" still doesn't solve another problem: comparing amount of money you have on a thing: if you held 50k on a position that went up 1%/year, and another position where you held 1k on a position that went up 1%/year, it's hard for your brain to meaningfully compare those things. ~a
[2019-03-28 13:26:22] - paul: reasons looking at individual transactions is hard. 1. individual transactions aren't enough information, you still need to look at the "current price" to see performance. 2. matching sell orders and buy orders is easy. still, you to come up with a "unit" of performance (%/yr?). if you have one time where you held a thing for a year and it went up 1%, that's very different than when you held a thing for a day and it went up 1%. ~a
[2019-03-28 13:17:39] - a: Right, but how do you know which of your individual positions are beating the S&P without looking at individual transactions, then? For instance, I think Disney is up pretty big for me in absolute terms since I've held it a long time, but is trailing the S&P. The only way I knew that was from using the TMF scorecard before. Now, I have no insight into that. -Paul
[2019-03-28 13:05:23] - paul: i also *def* compare performance to the s&p500, but my method does not look at individual transactions: it looks at balances, and inflows less outflows. ~a
[2019-03-28 11:41:21] - Xpovos: If you were joking I thought it was clever. Made me laugh. -Daniel
[2019-03-28 11:31:53] - Xpovos: 30 Rock was awesome and completely unrelated (just in case you weren't joking). -Paul
[2019-03-28 11:31:22] - a: I agree it can be tricky, but if I want to calculate the performance of a position against the S&P, I need to look at individual transactions. Otherwise there's plenty of other things I can use which can track the performance not against the S&P... -paul
[2019-03-28 11:29:56] - a: Heh. The gimmick is not new, certainly, but this evolution is cute and brings a different kind of humor. I liked Third Rock for a few seasons, too. Never watched Thirty Rock. I hear it wasn't a sequel/spin off. -- Xpovos
[2019-03-28 11:25:57] - paul: i don't look at individual transactions because that is very hard (i can explain why this is hard if you want), and the times i've done it, i didn't get useful information that i could take action on. i look at yearly/quarterly balances of each of my accounts and have a (simple) equation that's able to measure overall performance on an account that you're adding and removing money from. ~a
[2019-03-28 11:13:17] - xpovos/paul/aaron: parody
~a[2019-03-28 10:58:23] - considering all the options out there and how I can probably get 80% of the way there using pre-existing stuff. -Paul
[2019-03-28 10:58:02] - aDaniel: I'm envisioning downloading all my transaction information, storing it in a csv or something, and then using python to match up the performance of my open positions with the S&P. It seems like it shouldn't be that hard (knock on wood, obviously), but I wonder if it is worth the effort... -Paul
[2019-03-28 10:56:00] - aDaniel: So, the scorecard I've been using for tracking my performance (The Motley Fool's) has been running into some issues lately with its data feed. Since I haven't been able to find anything better which does exactly what I want, I'm wondering about expanding on some python scripts that I wrote to jury-rig something for myself. -Paul
[2019-03-28 10:34:56] - and yes i think they're really funny
i like the surprise party one - aaron[2019-03-28 10:34:06] - i've seen them here and there over the past month. but, i didn't realize how many there were, that's a crazy amount - aaron
[2019-03-28 09:39:19] - Yeah... it still seems like a bug in the Matrix sometimes, though.
-Paul[2019-03-28 07:35:45] - Paul: Quite oddly, I saw one of those comics for the first time yesterday. I understand that's how virality works and all, but the connections are just weird to see forming. -- Xpovos
[2019-03-27 14:00:18] - https://reason.com/blog/2019/03/27/justin-amash-on-running-for-president-it It's happening? -Paul
[2019-03-27 12:26:50] - https://www.instagram.com/nathanwpylestrangeplanet/ These comics seem like they would appeal to Aaron and Adrian's sense of humor. -Paul
[2019-03-27 10:28:51] - Daniel: It wasn't intended to be a strawman, more of a commentary on what I'm sure the reaction would be if the races were reversed... -Paul
[2019-03-27 10:27:19] - Daniel: Ah, I see the confusion, sorry. I'm not addressing any specific incident and I didn't mean to imply Peele was demonizing anything. I just saw the headline and realized that it would be a... possibly polarizing thing where some people pump their fist and cheer and others point out what would happen if the races were reversed. -Paul
[2019-03-27 10:21:06] - Paul: You said "can we at least not completely demonize the other" - are people demonizing casting white people in general? Is this just a straw man? I'm confused by this part of your statement. Peele didn't demonize casting white people. He just seems excited to be in a place to be able to cast minorities? -Daniel
[2019-03-27 09:04:41] - Daniel: "Who demonized it?" Who demonized what? Demonizing the idea of a white director only casting "white dudes" as the lead? Nobody, as far as I know, because I can't recall that being something said by anybody recently because I think everybody knows what would happen.
-Paul[2019-03-26 17:08:33] - Paul: Who demonized it? I think you can demonize the theory of only having white dudes cast and point at examples (Gyllenhal in Prince of Persia) but I don't think we should demonize casting Tom Hanks as Forest Gump. -Daniel
[2019-03-26 16:52:38] - a: Why? -Paul
[2019-03-26 16:52:24] - Can we disagree with it? Sure. Be disappointed by it? Absolutely. Knock yourself out. But I don't think it's right to completely demonize it. -Paul
[2019-03-26 16:51:52] - plea rejected. ~a
[2019-03-26 16:50:06] - https://www.hollywoodreporter.com/rambling-reporter/jordan-peele-says-i-dont-see-myself-casting-a-white-dude-as-lead-us-1197021 I'm 100% fine with what Jordan Peele said about not casting a white dude. And I completely understand why when you reverse the races it takes on a whole new meaning. My only plea is that if we're all fine with applauding one can we at least not completely demonize the other? -Paul
[2019-03-26 16:39:13] - paul: a straight up bribe. unless we're misunderstanding their meaning. ~a
[2019-03-26 12:48:31] - "The Cook County State's Attorney's Office said Tuesday that after reviewing the case, Smollett's volunteer service and his willingness to turn his bond over to the city, "we believe this outcome is a just disposition and appropriate resolution to this case."" Willingness to turn his bond over to the city? Doesn't that sound like a bribe? -Paul
[2019-03-26 11:45:45] - https://www.cnn.com/2019/03/26/entertainment/jussie-smollett-charges-dropped/index.html This is a little surprising. I wonder if we'll ever figure out what happened. It sounded like the cops had a pretty solid case for the charges of making a false accusation, but I guess not? Weird. -Paul
[2019-03-25 16:03:55] - a: I never thought of the fact that forcing candidates to focus on purple states, even if they are somewhat arbitrary ones, might be a good thing because the alternative might just be both sides relentlessly trying to fire up their base. -Paul
[2019-03-25 16:02:54] - a: http://reason.com/volokh/2019/03/22/the-electoral-college-is-just-ok btw, I thought this was an interesting "defense" of the electoral college. I'm not sure I agree with it, but it made me think, "Is politics better if large, purple-hued, swing states play an outsized role in presidential contests, or would we be better served by a politics in which presidential candidates cater to their base?" -Paul
[2019-03-25 16:01:47] - a: Fair point. Maybe it will happen. I guess I think of myself as a pessimist on matters like this. There has to be some redaction, right?
-Paul[2019-03-25 15:57:17] - paul: "i am especially concerned about what would happen if the report was not made available to congress" - second-ranking house republican. ~a
[2019-03-25 15:56:02] - paul: 420 to zero (two weeks ago) makes me think both parties will force somebody's hand. ~a
[2019-03-25 15:52:52] - a: You disagree? I'm purely guessing here as I haven't been following this story very much at all, but is there any reason why the AG would release it? -Paul
[2019-03-25 15:49:57] - i hope you're wrong. ~a
[2019-03-25 15:48:21] - huh. wow. ~a
[2019-03-25 15:45:58] - a: Probably not, unless there's something the Democrats can do to force somebody's hand. -Paul
[2019-03-25 15:45:48] - ok fair. say before october 15, 2020? ~a
[2019-03-25 15:36:09] - a: Um.... ever? I would say pretty high. Any reason we wouldn't get it the next time we get a Democratic president? I understand that's probably "too late", but you didn't put a time restriction on it.
-Paul[2019-03-25 15:33:09] - paul: well ok let me define it differently: what are the odds that the general public gets to read the entirety of the report sans redaction? ~a
[2019-03-25 15:21:18] - a: That we'll find out more? 90%. That we'll find out everything? 1%? I dunno. Didn't we already find out something new today about when Mueller told Barr about obstruction stuff? -Paul
[2019-03-25 15:05:28] - paul: how sure are you? ~a
[2019-03-25 13:04:16] - Anyway, it's not a big deal. I'm sure we'll find out more in the coming days. -Paul
[2019-03-25 12:49:45] - mig: Right, I understand they are different crimes or whatever, but I guess the way I thought it worked was that the report says: "Person A did crime X" or "Person B did NOT do crime Y" or whatever, so it seems strange that we can definitely say Trump did not do collusion but we still don't know about obstruction of justice. -Paul
[2019-03-25 12:40:50] - paul: they are completely separate issues. - mig
[2019-03-25 12:26:08] - mig: I guess I just don't understand how we can know that there is no collusion but not know if there is obstruction or not. Are there different processes or something? -Paul
[2019-03-25 12:12:36] - That in my opinion is still a longshot of happening, but the chances of it are nonzero. - mig
[2019-03-25 12:11:41] - paul: possibly, but Mueller punting any conclusions on that and leaving it to Barr and Rosenstein complicates things a little bit. Congress could then run with the "Barr's a Trump toadie" and drum up some investigative action on that front and maybe have an impeachment case. - mig
[2019-03-25 12:06:25] - mig: Can we similarly say there is no obstruction then? -Paul
[2019-03-25 11:58:51] - paul: collusion is off the table because there are no new indictments coming and Mueller's investigation is over? There's no need to take his word for it. - mig
[2019-03-25 11:43:17] - mig: Okay, so obstruction of justice charges are still possible, but not collusion? How do we know collusion is off the table? Is it just because we're taking Barr's word for it? -Paul
[2019-03-25 11:42:04] - I can see a legitimate gripe about Barr letting Trump off the hook for potential obstruction of justice charges, which apparently was his call. But anybody pinning their hopes and dreams that some grand Trump/Russian Conspiracy was going to be revealed just had those dreams crushed. - mig
[2019-03-25 11:36:29] - daniel: The big thing from the summary is that there will be no more indictments coming from Mueller's side. Meaning, ultimately, no one in the Trump campaign is going to be charged with colluding or coordinating with any Russians. That's not spin, that's a statement of fact.. - mig
[2019-03-25 11:20:33] - Daniel: Okay, thanks. I'm trying to keep an open mind until we know more concrete facts, but it seems similar to what I assumed was going to come out of this from the beginning: Trump personally probably did a lot of questionable things and surrounded himself with shady characters, but this wasn't some grand conspiracy by Putin to get a Manchurian candidate elected. -Paul
[2019-03-25 11:17:26] - Paul: He was put in by Trump. There are a lot of people (probably mostly D's) who think Barr is definitely biased. I'm not sure about people talking up the report before it came out. I guess there were people who thought that if Mueller wasn't indicting Don Jr / Ivanka / Trump himself then that counted as good news for Trump and co? -Daniel
[2019-03-25 11:11:57] - Daniel: Right, and while I personally believe that's probably accurate, wasn't he nominated by Trump and therefore not entirely unbiased? Also, people were talking up how good the report was for Trump before Barr even had his summary, weren't they? -Paul
[2019-03-25 11:07:17] - Paul: I think Barr's summary is all we have seen? Or is there more. Barr's summary says there was no collusion but not sure how much to trust that. -Daniel
[2019-03-25 10:55:36] - mig: Also, you never replied to the email, but I'm trying to plan some Starcraft 2 this Thursday evening.... -Paul
[2019-03-25 10:54:54] - This is probably a dumb question, but I haven't really been following this at all. Why do we know there is no collusion since almost nobody has seen the report yet? -Paul
[2019-03-25 10:15:50] - a: as far as impeachment potential, it seems like it. - mig
[2019-03-24 22:32:57] - mig: i don't know. is it? ~a
[2019-03-24 20:03:49] - a: https://www.nytimes.com/2019/03/24/us/politics/mueller-report-summary.html its over? - mig
[2019-03-21 14:04:04] - a: Heh, it's funny that you think I post a lot. Facebook is always telling me that people haven't heard from me in awhile and I always worry that I should be posting more. -Paul
[2019-03-21 14:01:17] - paul: i am. you post a lot, i only missed a few of these that had this data. ~a
[2019-03-21 13:55:42] - a: Ultimately I am happy not to own any (anymore), but wouldn't at all be surprised to see them both be big winners. -Paul
[2019-03-21 13:54:44] - a: So torn on AMD and NVDA. In some ways I like the opportunity in front of them, but I've also seen lots of chipmakers get stuck in a race to the bottom where they have trouble maintaining their edge because there's always competitors looking to do the same thing but cheaper. -Paul
[2019-03-21 13:52:50] - a: https://paulvsthemarket.com/fantasy-investing-feb-2019-standings/ I guess you aren't a regular reader.
-Paul[2019-03-21 13:32:59] - hah.
~a[2019-03-21 13:32:50] - paul: good stock market challenge day so far, right? 2 percent on both challenges. so, hey, that reminds me. did you make the other challengers picks public somewhere? ~a
[2019-03-21 13:31:34] - a: You're sneaking up on me in 2019. Down to a single digit lead courtesy of your chipmakers. -Paul
[2019-03-21 12:44:34] - anon: mriou, hah. now that is a name i haven't heard in a long time. ~a
[2019-03-21 12:43:42] - xpovos: yep! the 1/4th is you. ~a
[2019-03-21 12:18:16] - a: I'm the proud owner of the 1/4 that isn't Paul.
-- Xpovos[2019-03-21 12:17:32] - Paul: Alphabet/Google being so huge is a major concern for me on the bet. Apple's performance with the iPhone was after a strong success with the iPod, but they were really still quite small, all things considered. And they'd had a relatively long period of under-performing. There's no way I'd say yes on a ~500 point alpha bet. Oh... that's not fair, language. -- Xpovos
[2019-03-21 12:15:13] - Keeping track of bets is what MRIOU is all about.
[2019-03-21 11:43:37] - But this kinda goes back to the idea of whether votes "count" that I know I'm not in the majority on (my vote doesn't count?)
-Paul[2019-03-21 11:42:37] - https://www.politico.com/magazine/story/2019/03/20/elizabeth-warren-electoral-college-226044 I guess this is the argument for the EC, although I found it a little disappointing. This was a good point, though: "All votes are counted toward the outcome in every state. Voters from Republican, rural areas in California, for instance, aren’t disregarded; they are simply outnumbered." -Paul
[2019-03-21 11:28:46] - i wish we had fewer silly laws. ~a
[2019-03-21 11:28:21] - paul: i think it's probably true about most countries. ~a
[2019-03-21 11:19:03] - a: So is everybody who enters an NCAA bracket pool for money, right? I feel like Americans have a history of breaking silly laws. -Paul
[2019-03-21 11:16:00] - paul: nothing, ethically, imo. but legally, i think everyone who is using the platform may be breaking the law? ~a
[2019-03-21 11:13:46] - a: Yeah, but if the ledger is decentralized and the currency used can be decentralized.... what's the problem? -Paul
[2019-03-21 10:54:31] - paul: yeah, these exist. augur is one system i've heard of. google "prediction market", there are a bunch of others. here are some more. most of them decentralized. a problem, is legality. you're talking about something that is illegal in certain jurisdictions. ~a
[2019-03-21 10:50:31] - paul: yes, the other two are stock market challenges. ~a
[2019-03-21 10:49:02] - a: And then other people can browse and decide to take that bet in full or in part and the app resolves it until it reaches his limit. All exchanges could be done in bitcoin. Heck, wouldn't the blockchain be a good way to track the bets? -Paul
[2019-03-21 10:48:03] - a: Ah, crap. I remember the self-driving car one. Are the others all stock market challenges? I've recently thought a good Alexa skill or app idea would be something to keep track of bets and (more importantly) make it easier to make them. Like, Andrew could throw out, "I think GOOG outperforms the market by 50 points of alpha over 5 years and I'm willing to bet $100". -Paul
[2019-03-21 10:42:00] - paul: i have all of the bets i've made written down. i have four open bets and 3/4ths of them are with you.
~a[2019-03-21 10:35:29] - Xpovos: Also, it's just hard to outperform the market when you're already one of the biggest companies. -Paul
[2019-03-21 10:35:07] - Xpovos: More seriously, and I think there's like a 1% chance we remember this bet 5 years from now, but I would take the under on 50 points of alpha too. I don't see Stadia being a game changer and while Google should have some interesting growth from YouTube and Cloud, they are also looking at declining market share in ads. -Paul
[2019-03-21 10:33:52] - Xpovos: So, iPhone like returns means 475 - (-12) = 487 points of alpha over 5 years? I'll take the under.
-Paul[2019-03-21 10:22:44] - daniel: i understand what you mean by "cap" now, thanks. (sorry, i thought you meant a cap of a very different kind. maybe one where population over a certain amount wouldn't affect your ec count?) ~a
[2019-03-21 10:08:22] - Paul: So based on that (roughly 500 points of alpha over five years for APPL), yes, GOOG will beat 50 points of alpha over the next five. -- Xpovos
[2019-03-21 10:07:40] - Paul: That's tough, because I'm naturally very cautious when it comes to such things. 50 points of alpha, even over 5 years, seems like a lot. But that said, the market continually surprises me. I said iPhone level returns. APPL was worse off before the iPhone than GOOG is now, but APPL stock rose ~475% in the 5 years after the iPhone. The market (S&P500) during the same period declined about 12 percent! -- Xpovos
[2019-03-21 09:35:25] - One thing I've found fascinating the more I read about it is how.... unscientific? medicine can be. Or at least how flawed it can be. I also wonder if the piling on against anti-vaxxers has a negative effect at some point. People who are against vaccines probably already have a bit of an underdog "us vs them" mentality, so I worry that everybody piling on makes them dig in more. -Paul
[2019-03-21 09:33:28] - https://blogs.scientificamerican.com/cross-check/dear-skeptics-bash-homeopathy-and-bigfoot-less-mammograms-and-war-more/ I stumbled across this and found it pretty interesting, despite not necessarily agreeing with the author on everything. It's not too long, but basically talks about why skeptics should look more at things like medicine and war instead of homeopathy. -Paul
[2019-03-21 09:22:09] - Xpovos: Hmmm, we don't have to put money on it, but I don't really think GOOG ouptperforms the market by that much over the next 5 years. What's the over/under you would put the alpha (outperformance) at over the S&P? 50 percentage points over 5 years? More? -Paul
[2019-03-21 09:20:38] - a: Disagree on your comment on if Wyoming was a battleground state because it's still only 3 electoral votes. Why campaign in Wyoming when you could campaign in Ohio or Florida to get like 10 times the votes? Nevada and New Hampshire were considered battleground states last election and I don't recall a lot of people caring about them. -Paul
[2019-03-20 19:34:04] - Paul: Obviously, I like MSFT as well, and I think both (MSFT and GOOGL) outperform the market over the next five years. I think GOOGL has the homerun capacity, I think MSFT just does well. -- Xpovos
[2019-03-20 19:03:54] - daniel: Ok so, I get what you are saying. I just don't agree that it's necessarily "bad'. - mig
[2019-03-20 17:26:55] - But its not just straight proportional though somewhat. -Daniel
[2019-03-20 17:26:14] - a: So I think of that as "capped" but whether someone else would use that word, shrug? -Daniel
[2019-03-20 17:25:52] - a:https://www.archives.gov/federal-register/electoral-college/allocation.html (EC count is senators + state representatives) https://en.wikipedia.org/wiki/Reapportionment_Act_of_1929 - fixed the number of reps. So every state has a minimum of 3 and everyone else has to divy up the rest. -Daniel
[2019-03-20 17:15:13] - capped it? i didn't realize there was a fucking cap. ~a
[2019-03-20 17:06:44] - a: Thats a cool site, would be nice if you could move back in time to compare the graph across decades. -Daniel
[2019-03-20 17:05:24] - a: EC votes kind of scale with pop but like the house of represenatives rep count its skewed now that they capped it. Though it does get redistributed over time yes. Can't go under three though. -Daniel
[2019-03-20 17:02:08] - 1 person, 1 vote, was a lie our parents told us. ~a
[2019-03-20 16:59:46] - paul: battleground-ness is relative to the people in the state though. ohio and florida are only battlegrounds because half of the people in the state are republicans and half are democrats. if wyoming was a battleground state, they'd get a 3-votes per person. which is fucked imo. ~a
[2019-03-20 16:56:51] - a: It's funny, because if you look at that map, you might think that Wyoming is all powerful and gets too much from the EC, but that's hardly the case. Few people care how their 3 votes go. Instead it's often the battleground states of Ohio and Florida and whatnot (both lower on the list) that get the attention. -Paul
[2019-03-20 16:51:15] - So if you live in California and don't like the Democratic candidate, most voters feel like their vote won't count at all. -Paul
[2019-03-20 16:51:09] - daniel: http://www.ravi.io/how-much-is-one-vote-worth ok, maybe it's worse than i thought. ~a
[2019-03-20 16:50:37] - Because it scales with population, in my mind it's not so much the differences in how much a vote is worth relative to an electoral vote, the problem is more the winner take all system of it. -Paul
[2019-03-20 16:48:43] - daniel: yeah, uhhh, i don't think that's right at all. electoral votes are scaled by population (sort-of) every 10 years. so i don't think the problem gets worse as population centers move. ~a
[2019-03-20 16:45:25] - mig: A made up one? Or just personal internal? Long ago it might have been like a CA voter was like .85 of a WY voter or something. Now it might be like .15 of a WY voter. I totally made up those numbers because I don't want to go do the math right now but hopefully illustrates what I mean. -Daniel
[2019-03-20 16:27:08] - daniel: "out of balance" what criteria are you basing that on? I don't necessarily agree things are out of balance. - mig
[2019-03-20 16:09:37] - a: That's fair. The line has to be drawn somewhere. I'm sure my lack of concern over my vote being worth less is because I already have such a low opinion on how much my vote counts.
-Paul[2019-03-20 15:54:07] - paul: that is the case with senators. i'm not against senators. i guess i draw the line at senators? ~a
[2019-03-20 15:49:45] - a: "i don't think a montana resident or a wyoming resident should get more of a say than i do. will everyone agree on that, no i guess not" Isn't that the case with senators, though? Are you against Senators? -Paul
[2019-03-20 15:34:48] - It's agreed! My version is best. -Paul
[2019-03-20 15:05:05] - paul: i also would not like your version. ~a
[2019-03-20 14:55:31] - Paul: It would be a more extreme version of what we have now. Since I already think our EC is out of balance I don't think I would like your version much
-Daniel[2019-03-20 14:47:46] - Daniel: Honestly? In some ways I think I would prefer a system like the EC but every state had the same number of electoral votes over the weird EC we have now. At least then it would be more clear as to what it is trying to accomplish. -Paul
[2019-03-20 14:32:14] - Daniel: Although I honestly don't know if it even accomplishes that very well. -Paul
[2019-03-20 14:31:37] - Daniel: Right, that was part of what I was getting at with my comparison to the Senate. It's not fair to the voters in California that their vote is like 0.00001% or whatever needed to elect a Senator that has the same power as the one from Wyoming where those voters count like 0.01%... if that makes sense. -Paul
[2019-03-20 14:27:03] - a: Its unfair on a individual level but was trying to make things more "fair" on a macro level. However the balance has gotten pretty out of wack due to population movement / concentration. -Daniel
[2019-03-20 14:26:04] - a: Its certainly not perfect and I'm probaby more on the anti EC side but I think I understand the idea of it and why it was put in originally. -Daniel
[2019-03-20 14:25:31] - a: I would agree that the EC is unfair. I think the point of the unfairness is to try and balance out the perspective at a federal level so that the voices from MT or WY are still heard and have influence. I think its kind of an anti hive mind device. If all the east coast thinks a way then it can be good to get some MT perspective and the EC forces you to pay attention (some) to that. -Daniel
[2019-03-20 14:08:12] - a: I would rather compare GOOG to MSFT or AMZN than the market, since that seems to be a better pure comparison for what I am talking about, but I'll let Xpovos wade in if he is actually interested in a bet. I suspect GOOG does beat the market for the next few years, but it's not in the Freedom Portfolio so I don't feel too strongly about it. -Paul
[2019-03-20 14:02:50] - mig: i'm not a sore loser. the EC is unfair. i don't think a montana resident or a wyoming resident should get more of a say than i do. will everyone agree on that, no i guess not. ~a
[2019-03-20 14:01:56] - xpovos say to buy googl, so he's (maybe) putting money where his mouth is too. a way you measure success: does googl beat the market as of 2019-03-20? ~a
[2019-03-20 14:00:50] - A lot of the crying about how "unfair" the EC is mostly sore loser-ism, regardless. - mig
[2019-03-20 14:00:43] - paul: i have no opinion of stadia. i won't make that bet, try xpovos. ~a
[2019-03-20 13:55:41] - paul: My take on the EC is I'm against pure democracy when it can be reasonably avoided. A lot of systems in our government have things in place that prevent the majority from just getting what it wants (House vs. Senate, filibuster, Bill of Rights). The EC, to me is no different. - mig
[2019-03-20 13:55:18] - a: Like which? Waymo is the only one I know of which wasn't pretty much a complete failure. And I get that the whole point of moonshots is that it's likely to fail but you might learn something from trying. What did they learn from trying that has been impactful? -Paul
[2019-03-20 13:54:23] - a: How about this? I'll put my money where my mouth is. I don't know exactly how to measure it, but I will put 5 bits down saying that Stadia won't be a big deal 2 years from now. Maybe not that it is shut down, but that it hasn't revolutionized the gaming industry and few people use it. Maybe a comparison of Xbox Live Gold subscribers to paid Stadia subscribers? -Paul
[2019-03-20 13:51:41] - paul: "None of their moonshots seem to have worked out" they've had more visible failures, i'll agree with that. the past five years, they haven't done much super visible stuff, i'll agree there too. however, i think many of their moonshots worked out brilliantly. ~a
[2019-03-20 13:50:32] - a: They had an amazing.... I dunno, 20 year stretch? Which revolutionized so many things and like you said, I still use a ton of their products today. But very few of them were made recently. I guess I have a Pixel phone that I like which technically counts, but as much as I like the phone it's not some massive improvement over other phones. -Paul
[2019-03-20 13:48:22] - a: Right, just to be clear, I'm not arguing MSFT has been more innovative than GOOG for all time. I'm saying that it feels like GOOG is resting on their laurels lately. None of their moonshots seem to have worked out. They seem to have canceled more projects than they've released. They're followers in areas like the Cloud and Home Assistants... -Paul
[2019-03-20 13:39:14] - ok i guess technically i brought up the 10 year thing. oops. ~a
[2019-03-20 13:36:22] - paul: i don't even care about innovation as much as you do i don't think. android has revolutionized my life, and that'll be as true 10 years from today as it is true today. microsoft has . . . what's the opposite of revolutionized? ~a
[2019-03-20 13:34:42] - paul: i'm not focused on a 10 year mark, that is entirely you. i'm focused on who comes out with good shit i use every day. i use google shit every fucking day. same with amazon. same with netflix. microsoft? no fucking way. ~a
[2019-03-20 13:34:23] - a: In that same time their primary competitors have changed a lot. If I was comparing GOOG, AMZN, AAPL, MSFT in terms of which companies have significantly innovated in the past few years and look set to continue innovating in the next few, I might put GOOG last (or maybe ahead of AAPL) and that seems crazy to me consider how innovative they were before. -Paul
[2019-03-20 13:32:37] - a: Fair enough, you can remove AWS if you want. I'm a little less focused on the strict 10 year mark than you are, I think. My point is that Google hasn't really changed its core business much in ~10+ years and hasn't come out many significant (to me) products in the past ~5 years. -Paul
[2019-03-20 13:30:23] - a: But to my latter point, most of those ARE more than 5 years old. -Paul
[2019-03-20 13:29:49] - a: Okay, so you made me look it up. Here is what I found: 10+ years: Android, Youtube, Maps, Gmail, reCAPTCHA (5). <10 years: Dart, Flutter, Golang, Fi, Drive, Chromecast (6). You were right, although I was shocked Drive was so recent considering docs and sheets are so old... -Paul
[2019-03-20 13:17:13] - paul: "most are 10+ years old" except the ones I labeled as such, no they were not. I looked them up. This is all in your head. ~a
[2019-03-20 13:14:23] - paul: i heart aws, but that was 12 years ago.
~a[2019-03-20 13:04:16] - a: Google just keeps half-heartedly trying things and then shutting them down a few years later. Google Reader, Google Glass, Google Plus... I mean, the comments section on ArsTechnica for Stadia was almost entirely people joking about how they're looking forward to the shutdown announcement in 18 months... -Paul
[2019-03-20 13:02:28] - a: Whereas Amazon has started up AWS and trail blazed home assistants and is starting up an advertising business and bought Whole Foods and is working on delivery.... -Paul
[2019-03-20 13:01:43] - a: But if you look at the past 10 years and compare Google with Amazon or even Apple and Microsoft, they basically haven't done much and are playing second fiddle in a lot of areas. They're still basically "just" an advertising company that makes most of their money from the main Google property. -Paul
[2019-03-20 13:00:29] - a: Maybe it's because I'm looking at it from an investing perspective, so I'm discounting things like programming languages, and I would concede that maybe 10 years was too aggressive (I typically think of it as 5 years or so that Google seems to have stagnated)... -Paul
[2019-03-20 12:58:33] - a: I didn't look up all of those, but most are 10+ years old. No doubt Google has been one of the most innovative companies ever, but they actually haven't done much in the past 10 years. -Paul
[2019-03-20 12:57:49] - Daniel: Google Sheets was 13 years ago. Stradia I think we need to see where it goes. Remember Google Glass? -Paul
[2019-03-20 12:53:38] - paul: dart and flutter (omg so amazing), golang (seriously, the best), google fi (i use this every day), google drive (i use this every day), chromecast (i use this every day), android/youtube/google-maps/gmail (i use most of these every day, technically not in the past 10 years but a few were just barely over 10 years, and a few of them were *only* technically acquisitions), reCAPTCHA, god, i know i'm probably missing some good ones. ~a
[2019-03-20 12:45:59] - Paul: I mean stradia?
-Daniel[2019-03-20 12:43:13] - When were google docs / sheets invented? Those are pretty strong. -Daniel
[2019-03-20 12:42:08] - a: Right, but the 90s were a long time ago. I definitely agree Microsoft wasn't the best most innovative company back then, hence why I commented how it's crazy how things have changed. I'm talking about the Nadella era that we're in now (and have been for the past few years). Let me ask the question of you: What new innovation has Google had in the past 10 years? -Paul
[2019-03-20 12:41:05] - daniel: i often send odt/ods/odp files to other people that use microsoft because i know that they can open them. if it's a non-computer person though, i'm worried i'll scare them off with my scary files. ~a
[2019-03-20 12:40:08] - daniel: yes. ~a
[2019-03-20 12:39:36] - Can word open an odt file? I know libre / open office can open .doc. -Daniel
[2019-03-20 12:35:14] - the fact that many people (not everybody) still use docx/xlsx/pptx files drives me nuts. i'd much prefer people just shared google drive files. or odt/ods/odp jeeze. ~a
[2019-03-20 12:33:32] - paul: i'll agree that they've been good stewards of github so far. i'm not convinced that won't change considering their track record. i'll also agree that office of 1999 was pretty sweet, and we probably have good office choices now because of that, but their lack of open standards in the 90s (which we argued about on this message board decades ago) *may* have hurt us in the long run. ~a
[2019-03-20 12:27:12] - a: They've also done a pretty slick job of transitioning companies to subscriptions to things like Office, which you might not want to give them credit for depending on your thoughts on Office.
-Paul[2019-03-20 12:26:08] - a: I assume you mean exclude all gaming stuff, but I've heard really favorable things about Azure (their cloud solution) and while it likely doesn't count for something they made, they own LinkedIn and GitHub and apparently have been good stewards of both so far. -Paul
[2019-03-20 12:18:27] - paul: microsoft is what-now? god, i hate every time i have to use anything microsoft. yesterday i was using microsoft onedrive live (think google docs, but shitty) and i was so angry at how horrible it was. any time i use onedrive, live anything, office365, outlook anything, or sharepoint i want to hurt someone. please, tell me one good thing microsoft has made in the last 10 years? (other than the xbox console series) ~a
[2019-03-20 12:05:17] - Xpovos: If I wanted to bet on a future where there is a "Netflix of gaming" that kills consoles and uses the cloud for everything, I would definitely put my money on Microsoft ahead of Google. It's crazy to me how much those companies seem to be going in opposite directions now. Google doesn't seem to innovate much anymore and Microsoft is tearing things up. -Paul
[2019-03-20 12:03:28] - Xpovos: I can see it, but I think Microsoft has the big head start in gaming in general and stuff like Game Pass specifically. They also have the cloud infrastructure to quickly match anything Google tries to do, whereas Google has virtually no gaming presence right now. I keep wondering what blockbuster games they're planning on offering. -Paul
[2019-03-20 11:53:37] - daniel: how would you do this without giving more weight to some voters than others? ~a
[2019-03-20 11:50:11] - a/Paul: Yes, GOOG on Stadia. Because if it works it's going to be iPhone level revolution on a market of roughly the same size. That's worth getting in front of. -- Xpovos
[2019-03-20 11:32:54] - Is there precedent for unwinding mergers that have already gone through? I guess its the same as just breaking apart an existing company? Somehow seems odd though. -Daniel
[2019-03-20 11:28:17] - a: I think as that balance has shifted more and more over time though it definitely needs rethinking and adjusting (eliminating?) but I don't think its a terrible idea on the face. -Daniel
[2019-03-20 11:27:42] - communal government. -Daniel
[2019-03-20 11:27:35] - a: I think the case for the EC (at least the one I would make) is to try and provide some balance between representation by individuals and representation of the parts of the country. I think if the US was only CA, TX, FL, NY, and PA then we would be worse off as as country. So to be inclusive to the smaller less dense parts of the country lets give them a voice (Small States Matter ~ Black Lives Matter?) so they get represented in our...
[2019-03-20 11:19:53] - jokes. ~a
[2019-03-20 10:57:33] - https://medium.com/@teamwarren/heres-how-we-can-break-up-big-tech-9ad9e0da324c Completely unrelated, but I'm looking for a headline to talk about this for PvtM and... well... these are three of the FANG companies and I would love to use a similar acronym so WHY did she have to leave Netflix off her list?
-Paul[2019-03-20 10:57:30] - yeah ok, that makes sense. ~a
[2019-03-20 10:54:27] - a: Let me phrase that a little cleaner: I think the ideal solution probably is to move to eliminate the EC and just do everything by national popular vote, but I also would like to hear the best arguments for both sides and not make any rash decisions. -Paul
[2019-03-20 10:53:24] - a: No, not necessarily. Just trying to provide the most reasonable "pro" case I can. I'm a little nostalgic for the EC, and I don't think it's this huge problem that needs solving, but I also don't care much to defend it and am fine with us having an honest discussion on how best to replace it. -Paul
[2019-03-20 10:44:03] - paul: it seems weird to argue for giving extra weight to people living in certain states than people living in other states. do you believe in this argument? or are you arguing the logic of it? ~a
[2019-03-20 10:34:00] - a: Well, yeah, the population. -Paul
[2019-03-20 10:33:47] - paul: i.e. if the wyoming legislature, and wyoming executive, etc want a certain president, they still have to get their population to vote for that president. the state itself can't do shit. ~a
[2019-03-20 10:33:31] - a: Obviously it's a little different, because California still has a huge number of electoral votes vs Wyoming, but it evens things out a bit. -Paul
[2019-03-20 10:31:40] - paul: the ec doesn't give power to the states. it does give power to the population in certain states. is that what you mean? ~a
prev <-> next