cepth

I can't speak to Chess Assistant, but as a longtime Chessbase owner/user, it lets you:

* Search any database (proprietary or freely downloaded) for games matching positions (partial position matches too)

* Quiz yourself on openings; e.g. the actual memorization of the move orders, which I don't believe ChessRoots is showing

* Generate endless tactics problems to drill on

* Generate annotations for games; e.g. branching moves, variations, text comments etc.

* Graphical markup of positions, to show tactics like pins, skewers, forks, etc.

* ChessBase gives you a perpetual license. You can upgrade if you think new versions are worth it (they usually aren't). I happily used ChessBase 9 for several years before upgrading to 14.

It's really not a fair comparison. If ChessBase is Photoshop, ChessRoots as it is now is just a simple painting tool. This isn't to say that ChessRoots can't be something worth the price down the road, but for now this is not a great comparison.

Chess.com is also something that I've happily paid for for many years. For $99 a year, you get unlimited tactics training, engine analysis (with Stockfish), and a full license to what used to be called Chess Mentor, in addition to opening specific training.

The point about survivorship bias is a fair one.

But, if a given fund (Medallion in this case) is able to thrive/survive during the 1998 Russian financial crisis (which felled LTCM), the 2000-2004 tech bubble bursting, the 2007 quant quake, the 2008 financial crisis, and 2010 flash crash, it would seem to suggest that their statistical arbitrage strategy performs well in all manner of environments.

After their very hefty 5/40 fees, their worst year between 2001 and 2013 was a 21% gain [1].

I understand that there are all manner of epistemic and mathematical problems that prevent us ever from completely disentangling luck and skill, but a conceptual framework like "The Superinvestors of Graham and Doddsville" would seem to apply here. I.e., after a long period of consistent overperformance, the case for skill starts to look much more likely.

[1] http://archive.is/JdQiw

In the book (“The Man Who Solved the Markets”), the author actually provides the Sharpe ratios for their flagship fund (“Medallion”). Medallion’s Sharpe Ratio has almost never been below 2.0, and has been as high as 7.0 (!!!) in some years.

I’ll link to the Wikipedia page below for the math, but these figures would highly suggest that their returns are much less likely the result of luck, and more likely the result of some edge/alpha generation.

https://en.wikipedia.org/wiki/Sharpe_ratio

Yeah, seems like you're correct here.

I went back and re-read the pre-print here (https://www.cs.cmu.edu/~noamb/papers/19-Science-Superhuman.p...). On page 2:

> To reduce the complexity of forming a strategy, Pluribus only considers a few different bet sizes at any given decision point. The exact number of bets it consid-ers varies between one and 14 depending on the situation. Although Pluribus can limit itself to only betting one of a few different sizes between $100 and $10,000, when actually play-ing no-limit poker, the opponents are not constrained to those few options. What happens if an opponent bets $150 while Pluribus has only been trained to consider bets of $100 or $200? Generally, Pluribus will rely on its search algorithm, described in a later section, to compute a response in real time to such “off-tree” actions.

Good catch, and thanks for the correction.

Regarding the effect of stack sizes, I'm not certain on this, but my intuition is that there is some effect on perceived ranges of the other 5 players at the table if stack sizes vary. Since Facebook AI will not be releasing Pluribus code or pre-trained models/weights, we can't be certain, but things like stack-to-pot (SPR) ratio would seem to matter.

Of course, you could always make the argument that human players in a cash game can re-up/refill to the maximum buy-in whenever they're short, but that's another discussion altogether.

As an addendum to this very good comment, that Nature article had a couple of big asterisks.

* Bet sizes were restricted. E.g. humans and the the bot could only bet fixed bet sizes, like 1/4 pot, 1/2 pot, full pot etc. Creative bet sizing is one of the skills that distinguishes top pros.

* Stack sizes were reset after every hand. E.g. every player in the hand was given the same amount of chips at the start of every hand. How you performed previously in the session thus did not matter. Anyone who has played poker knows that this is highly unrealistic. Larger stack sizes convey an ability to bully smaller ones, and stack sizes greatly affect what range of hands you can reasonably play.

The point being, even a supercomputer running the most efficient heuristic based poker decision making programs has not yet been able to beat humans in a game that resembles what a real 6 or 9 person table would reflect.

---

Just as reference, on a four-year-old quad-core/8-thread Intel i7-based desktop with 32GB of RAM, to solve a SINGLE hand in PioSolver (the most popular poker solver) from flop through the river takes my machine about 7 minutes. The game tree alone takes up 4 GB of RAM, and in this scenario there are only two players, and each player is restricted to 3 bet sizes.

The idea that this kind of computation can be done on a phone is ludicrous.

A couple of responses here:

First, the issue is that he's winning at a rate that is 10+ standard deviations above what the typical winning/profitable player is achieving. These results are accomplished across a significant sample of sessions (39 live streamed ones), and across almost a year.

If you consider some of the players considered the best "live readers" in the history of the game (Ivey, Negreanu, etc.), none have won at the rate that he has.

Secondly, he's playing $1/3 and $5/5 poker in a city (Sacramento) that is not exactly a gambling mecca. You would expect someone who is as good as he is to play higher stakes games. $1/3 and $5/5 are stakes offered at nearly any casino.

Third, he's performed well against some elite pros. Matt Berkey (who regularly plays the highest stakes in Vegas) was in a streamed game against Postle, and Postle performed extremely well in heads-up hands. Given that Berkey is playing for hundreds of thousands of dollars on a nightly basis in Vegas, and seems to be a profitable player in those games, are we supposed to believe that Postle somehow has identified body language tells that no one else has discovered?

This is not even considering other facts like:

* Postle used to consult for the production team behind the streaming for the casino

* Postle used to claim to run some kind of app development company, and has since deleted his LinkedIn

* The extreme winning sessions seem to only have taken place during streamed games

Sure, it's possible that there's a benign explanation here, but it's looking far and far more unlikely.

A 30-minute (minimum) time delay on a live stream is usually a legal requirement in most jurisdictions, enforced by the state gaming commission.

Most poker rooms also allow you to use your phone, as long are you're not a participant in the current hand. The alleged cheater (Mike Postle) is said to have hid his phone under the table, in his crotch area. Obviously, it's going to be pretty hard for other players and the dealer to check for that at that angle.

In the end, the no-phone rule in most poker rooms is going to be largely an honor-system based thing.

Edit:

Jason Somerville, who founded one of the most popular poker streaming companies (Run It Up), has described how stringently the Nevada Gaming Control Board inspects casinos that are hosting streams. E.g. multiple inspections of the streaming equipment, a requirement for a guard at the streaming booth, etc. (https://www.youtube.com/watch?v=ZS47DB94-vk). He mentioned that California does not seem to have the same level of on-site scrutiny.

Edit2:

On TwoPlusTwo (the popular poker forum), long time players at Stones Gambling Hall (where the alleged cheating took place) mentioned that their poker room did not institute a no-phone policy until sometime into the alleged cheater's crazy run of success.

There’s a grey area between what is (should be?) unacceptable in a professional setting, and what constitutes a crime. Even in particularly egregious cases of sexual misconduct, it’s difficult to think of how many victims chose to go to trial. Civil settlements are much more common. Think Roger Ailes at Fox.

The local prosecutor in whatever jurisdiction this incident occurred in is likely to decline prosecuting the case. Prosecuting instances of sexual harassment (assault in this case?) Is not typically a prosecutor’s first choice for how to use their office’s limited time and resources.

Maybe she could’ve filed a claim with the EEOC? But, that’s a long drawn out process as well. Most victims of harassment may just choose to try a company’s internal HR processes, since an EEOC claim may require spending inordinate amounts of time, and maybe money if you hire professional help.

From the dozens of Twitter and personal blog testimonials of DataCamp instructors (contractors), it seems that but for collective pressure and organizing from the instructors, DataCamp would’ve just swept this whole incident under the rug.

I’ve personally rolled out a Freshworks implementation, and am quite happy with their products.

They offer Freshdesk for a customer support portal/ticketing system, and Freshchat for live chat. Total cost is a fraction of what you’d pay for Intercom, and the products are continuously adding new features.

I don’t think the parent commenter is far off here: https://en.m.wikipedia.org/wiki/Demographics_of_China#Ethnic....

Granted, Han Chinese in the modern sense is a bit of a constructed identity (https://www.reddit.com/r/AskHistorians/comments/6e312e/is_th...), but it’s a construction that has held up in China.

With the exception of Guangdong and Hong Kong, you aren’t going to find many TV stations or newspapers that are carrying content in non-Mandarin Chinese. Yes, when I visit older family in the countryside, they may still speak their local dialect, but that is becoming increasingly rare: https://www.theatlantic.com/china/archive/2013/06/on-saving-....

Unlike in the US where African and French influences produced creole, and American pop culture has strong African and Latin influences, I don’t think minority group culture in China has anywhere near the same pull. Yes, they may trot out ethnic minority performers once a year during the CCTV New Year’s Gala, but you will not find “minority” singers charting on best seller lists, or “minority” actors in leading entertainment roles.

A 90%+ majority ethnic group is on par with what we see in Western Europe. Do Europeans see themselves as primarily multi-ethnic states? The self-description of a “melting pot” that is commonly used in the US, Canada, and even Central America don’t seem to be nearly as prevalent in Europe.

Condé Nast seems to have figured out a workable majority-subscription revenue model.

https://www.wsj.com/articles/conde-nast-to-put-all-titles-be...

> “The New Yorker, which introduced a metered paywall in late 2014, generated about $115 million in paid-subscription revenue in 2018, up 69% from 2015, people familiar with the matter said. That revenue includes consumers who subscribed to the digital and print editions, although subscribers today no longer have the option of subscribing solely to the print magazine. The New Yorker’s regular renewal price for a print and digital bundle is moving to $149 a year from $119. The magazine will publish 47 issues this year.”

Granted, there’s a big difference between publishing 47 issues of long form writing a year, and 365 issues of a much longer daily-reported newspaper.

EDIT: not to mention that the Times and other big newspapers probably see themselves as informing the public discourse. It probably wouldn’t fly to put general newspapers behind a paywall, though the FT has greatly increased profitability with this strategy.

This is a pretty substantial achievement, if you look at an estimated percentile breakdown: https://www.reddit.com/r/chess/comments/54c1nv/player_rating....

FWIW, I’ve found that on Lichess I’m rated at least 100 points higher than what I would be on Chess.com (the most popular chess site), and maybe 150 points higher than what I’m rated by the US chess federation, which mainly rates live, not online, games.

A player with a rating of 2000 by the US federation would be called an “expert”, just below the 2200 rating that earns you the title of “master”.

What part of TSA screening do you object to? The ID check, or the actual screening of bags?

It’s true that TSA security has been foiled by various “red team” exercises, sometimes in comical/embarrassing fashion. And, there is probably bloat in the agency, as well as problems with the contracting process for companies selling equipment. No one likes having to take off their shoes, myself included. Maybe policies around liquids and gels should be revisited.

At the same time, I think it’s unknowable how a return to pre-9/11 airport security would fare in today’s world; namely a world where private companies were “responsible” for screening, with the FAA providing oversight. See for example: http://www.cnn.com/2004/US/01/27/911.commis.knife/.

> “At the time of the attacks, commission staffers testified, the FAA prohibited passengers from boarding aircraft with knifes having blades longer than 4 inches, tear gas, mace and similar chemicals.

> Except for guns, large knives, explosives and incendiaries, carriers and screening contractors were allowed to use "common sense" in determining what was prohibited on planes, the staffers said.”

In today’s world, I’m not sure leaving airport security to the “free market” is necessarily a good idea. If there are new types of explosives, will private companies be able/willing to buy new equipment to deal with “improbable” threats?

TSA also deploys agents around the world to screen flights coming inbound to the US that originate in countries that can’t necessarily perform their own screenings. Do we expect private companies to pay Americans to live abroad? Or will they rely on poorly paid locals to do the job, maybe without background checks (if it’s even possible).

From a game theoretic perspective, I think it’s reasonable to assume that TSA provides some deterrent effect. Stringent security procedures do reduce the action/choice set of would-be terrorists. Post-9/11 flights within the US, and to the US, have not been affected by terrorist attacks, incompetent shoe and underwear bombers aside.

Ancillary to the job that TSA does, do you object to heavily armed police or military at airports? Because incidents like these: https://en.wikipedia.org/wiki/Brussels_bombings, https://en.wikipedia.org/wiki/2016_Atat%C3%BCrk_Airport_atta..., https://en.wikipedia.org/wiki/2011_Frankfurt_Airport_shootin..., do happen.

—-

FWIW, in Israel, the security process is far more stringent. You have pre-airport physical checks, interviews, passing of luggage through pressurized chambers etc. If you fly El-Al anywhere in the world, there is an extra headache associated with additional screening. But, the record of an El-Al (0 successful hijackings) leads many who travel to Israel to choose El-Al.

TL;DR, everyone wishes there was a way to have secure flights, without having to undergo invasive screenings.

FYI earlier this month, they actually did announce the results of a 1000 game match:

https://www.chess.com/news/view/updated-alphazero-crushes-st....

AlphaZero won 155 games, lost 6, and drew 839 games against Stockfish. Granted, this was against Stockfish 9.

This implies that AlphaZero was roughly +52 Elo rating against Stockfish 9 (https://www.3dkingdoms.com/chess/elo.htm).

Stockfish 10 is currently rated ~32 points higher than Stockfish 9 (http://computerchess.org.uk/ccrl/4040/rating_list_all.html). If we were to do very crude transitive reasoning, you'd expect AlphaZero to still beat Stockfish 10.

EDIT:

So apparently the +155, -6 score was against Stockfish 8. Stockfish 8 is rated by the CCRL list at 3379, with Stockfish 10 rated 85 points stronger than 8.

Worth noting that AlphaZero was only given 4 out of 9 hours of total training time when playing against Stockfish 8 (https://chess24.com/en/read/news/alphazero-really-is-that-go...), but I guess we can't make any real conclusions about AlphaZero vs Stockfish 10.

EDIT 2:

So apparently AlphaZero also "defeated" Stockfish 9, but the preprint of the upcoming paper in Science doesn't seem to provide a crosstable.

It seems that Stockfish 8 was given a 44-core machine to play on, and was not constrained in terms of time spent per move etc.

The vast majority of "action" available online is "6-max" (6 player) or "full-ring" (9 player). On a PokerStars, WSOP.com, or Party Poker, you're going to find that there are maybe 1/10th or 1/20th the number of headsup tables as higher capacity tables.

The development of "GTO" (game theory optimal) play in Texas Hold 'Em is certainly a first step in the direction of computers playing poker. However, there's still quite a long way to go.

Poker Snowie, one of the cutting edge "GTO" programs, is based off of NNs (https://www.pokersnowie.com/about/technology-training.html). At the same time, there are some glaring weaknesses in the software, namely that it can only offer suggestions at specific pot size bets (0.25, 0.5, 1, 2). The authors themselves concede some other weaknesses (https://www.pokersnowie.com/about/weaknesses.html).

Worth mentioning that Amaya, the owner of PokerStars, has posted job openings for "AI researchers" (http://www.starsgroup.com/careers/job/Poker-AI-Research-Engi... & https://www.pokernews.com/news/2017/10/pokerstars-to-hire-ar...). Some people think that the position may be to help PokerStars detect/combat bot use, but others think that there may be a (arguably bleak) future where players have the option to compete against Amaya-created bots online.

IMO, at the highest levels of chess, there has always been a focus on "positioning" (positional chess) over material. World Champions like Capablanca, Botvinnik, Karpov, and Kramnik all play/ed in a style that was postionally sound, and at times boa-constrictor like. If you want to be a grandmaster today, you have to be able to understand/execute concepts like giving up material to establish a fortress ( https://en.wikipedia.org/wiki/Fortress_(chess) ).

The World Champion that played in the most sacrificial/attacking style, Mikhail Tal, was famed for giving up pieces to generate attacking momentum. Contemporary analyses of his play have found that some of these sacrifices were unsound, and some were actually the "best move" in a given position.

I don't think it's feasible to expect human players to be able to calculate at the ply/depth that AlphaZero (or other chess engines) is able to. See this example from the latest World Championship (https://www.chess.com/news/view/world-chess-championship-gam...). A "forced" win in 30 moves was available on the board, but it would've required that Caruana make moves that cut against the "principles" regarding piece placement ("positioning") that are drilled into chess players.

I think a simple reality is that the search depths that AlphaZero (and to a lesser extent other chess engines) are dealing with are simply beyond human capability. A human player trying to execute the sacrifices that AlphaZero did (https://chess24.com/en/read/news/alphazero-really-is-that-go...) would be taking a stab in the dark. In most positions, they wouldn't really be able to calculate all the variations, or foresee how the endgame would play out.

I think that the endowment effect may be the more relevant behavioral economics concept here.

https://en.wikipedia.org/wiki/Endowment_effect

> "When you say "ship at the negotiated rates" that Amazon has access to, it's really hard to interpret that as anything but the amazon warehouse rate."

Who negotiates the rates that are available on "Buy Shipping"? I could've been clearer, and added the "Buy Shipping" name specifically, so sorry for any confusion.

My point is simply that an Amazon merchant who wants to sell with a Prime badge previously had no such access to shipping discounts. Now they do.

> Backing down to "some negotiated rate that Amazon has, but not even close to the best one", is a very different beast.

At no point did I say this. UPS has gross margins of ~8%, FedEx of ~20%. Any discount off sticker price is going to affect those gross margins significantly, so the discounts that SFP merchants have access to are very meaningful ones.

Assuming that you accept my premise that "Buy Shipping" rates are up to 20% better than the rates available to even a volume shipper, a SFP merchant now has rates that are much closer to Amazon's own shipping costs than their previous rate.

So no, SFP merchants do get rates that are much closer "to the best one". Amazon's own rate < SFP rates < "sticker/market" rates, and the delta between SFP and market is much larger than between Amazon's own rate and SFP.

> SFP rates are equal to Amazon MF rates, which are better than public rates but certainly aren't Amazon's own rates being passed through.

Refer to your original comment: "This is false. You do not get Amazon rates for shipping using SFP."

I took this mean that you do not get any discounted rate from Amazon for shipping using SFP. If you're now amending your claim to say that a SFP seller has access to "Buy Shipping" rates, then yes we agree. And if you're conceding that "Buy Shipping" rates are cheaper than those available to other businesses, then we agree again.

At no point did I claim that an individual merchant is getting the exact same quote to ship a package as Amazon would. UPS/FedEx wouldn't allow this, since pickup points are different, etc. But, it is certainly because of Amazon's volume that merchants who want to sell with a Prime badge can now get deeply discounted shipping rates. This is what I meant by "negotiated rates for UPS/FedEx/USPS that Amazon has access to".

> I'm not even the only one replying to you that disagrees with your claim

At the time of your comment, who else is disagreeing? This comment (https://news.ycombinator.com/item?id=18695969) is saying that Amazon's negotiated rates "aren't better". Better than what? Even you agree that buying through Amazon's "Buy Shipping" services are better than the baseline rate tables from UPS/FedEx. What I'm saying is that these rates are better than even the ones available to a merchant doing significant volume.

My company easily does 6 figures in shipping spending across UPS and FedEx annually (on non-Amazon shipping). What we're paying for SFP shipping is at worst equal to, and at best 20% below, what our own rate tables quote.

YMMV, but given the variety in what Amazon merchants ship in terms of size, weight, hazardous/insurance status (not to mention shipping origin and destination), don't you think it's a tad glib to tell me I'm "confused"?

Seems like this depends on who you ask? (https://techcrunch.com/2016/07/02/andy-jassys-brief-history-...)

> "It began way back in the 2000 timeframe when the company wanted to launch an e-commerce service called Merchant.com to help third-party merchants like Target or Marks & Spencer build online shopping sites on top of Amazon’s e-commerce engine...At that point, the company took its first step toward building the AWS business by untangling that mess into a set of well-documented APIs."

> "The internal teams at Amazon required a set of common infrastructure services everyone could access without reinventing the wheel every time, and that’s precisely what Amazon set out to build — and that’s when they began to realize they might have something bigger."

> As the team worked, Jassy recalled, they realized they had also become quite good at running infrastructure services like compute, storage and database (due to those previously articulated internal requirements). What’s more, they had become highly skilled at running reliable, scalable, cost-effective data centers out of need. As a low-margin business like Amazon, they had to be as lean and efficient as possible.

> "It was at that point, without even fully articulating it, that they started to formulate the idea of what AWS could be, and they began to wonder if they had an additional business providing infrastructure services to developers."

From The Everything Store:

>"Groups within Amazon were told to use AWS while the services were still immature, a demand that led to another round of consternation among its engineers." (Page 221)

I guess we could say that AWS's first customer was not necessarily Amazon, but the team and infrastructure predated AWS, and AWS was used by Amazon internally.

> Their negotiated rates aren't better, and they are actually forcing sellers to buy USPS postage with Amazon's account instead of the sellers.

It's true that you're forced to buy with Amazon's account, but this is shipping spending that would've been paid for by Amazon anyway. The merchants are not losing out on any volume discounts, since this was never "their" volume anyways.

IMO, this is the way to think about this: if the Amazon-negotiated rate for warehouse -> customer shipping is lower than what the FBA fee would've been (which it often is), the merchant is coming out ahead.

> Seller Prime is a couple years old now and it seemed to backfire on them and it seems they are trying to push sellers off of it by enforcing near impossible to meet standards that even Amazon themselves don't meet.

Yes it's about 2 years old, but it wasn't something that was widely popping up on Seller Central dashboards until earlier this year. I don't know whether you're referring to the ship-out time requirements, but so far my company hasn't had any issues with SFP. Better yet, the merchant controls which items are fulfilled under SFP, and which are FBA.

Here's a related development:

Amazon has been pushing many of its larger Amazon FBA merchants to use Seller Fulfilled Prime (SFP). Instead of these merchants shipping from their own warehouses to an Amazon fulfillment center (FC), merchants ship directly from their own warehouse to customers. Amazon is dangling huge fee discounts to incentivize eligible merchants to use SFP.

The benefit for merchants is that they get to ship at the negotiated rates for UPS/FedEx/USPS that Amazon has access to. Additionally, these SFP merchants still get the Prime badge.

For Amazon, this will likely mean fewer FCs to build, and being able to save on labor, packaging, etc.

To me, it was always evident that Amazon loses quite a bit of money on shipping/handling some larger items. When an FBA merchant sells on Amazon, there's a referral fee and FBA fee. The FBA fee is often absurdly low, and might be lower than the cost to just ship from the merchant warehouse to a FC.

It seems to make intuitive sense that instead of shipping from point A (warehouse) -> point B (FC) -> point C (customer), if you can ship from point A to point C in a reasonable amount of time, it saves all parties money.

ADDENDUM:

It's also worth considering that SFP will be great for Amazon's efforts to build out its own delivery fleets. I imagine that in the future, SFP merchants will be forced/given the option to ship with Amazon's own parcel delivery service. This will be a steady base of business (extreme "dogfooding") to start off with, similar to how AWS originated with internal infrastructure for Amazon.

Just for the sake of accuracy, most of the Northeast states are net givers in terms of federal taxation.

I.e. for every $1 that NY sends to the federal government in federal taxes, it receives something like ~$0.80 in spending: https://www.osc.state.ny.us/press/releases/oct17/100317.htm. Most of the money that NY receives is for Medicaid and other entitlements.

NY receives every year a total of $165 per capita for transit, slightly less than the national average of $168. When considering the cost of labor and materials in NY, that gap is probably much larger in real purchasing terms.

On the other hand, NY state directly spends $639.2 per capita on transportation annually: https://www.usgovernmentspending.com/year_spending_2019NYdn_.... The actual cost of the "pensions" you deride is $1085.2 per capita per year to the state. If you look at total state and local spending, the total costs of pensions are $1758 per capita per year, and transportation is $1788.8 per year. Neither reaches 10% of total statewide spending.

This doesn't directly negate your point about federal money trickling down, but NY is not receiving very meaningful assistance from the federal government.

Cities actually spent quite a bit of money on hiring consultants, site planners, doing impact studies, etc. I don't think it's unreasonable to say mid to high 7 figures.

-https://www.consulting.us/news/290/consulting-firms-recruite...

-https://technical.ly/philly/2017/09/13/site-selection-consul...

This is not to mention that cities likely had to hand over demographic data that would've otherwise been very expensive to collect: https://www.businessinsider.com/amazon-hq2-search-data-2018-....

Look at some of the cities that made the finalist list:

-Columbus (population ~880k)

-Denver (population ~700k)

-Indianapolis (population ~870k)

-Nashville (population ~667k)

-Newark (population ~285k)

For these cities, an additional 25k jobs would have been huge. Scott Galloway has advanced the theory that a large goal of the HQ2 search was to secure competing bids, which Amazon then took back to NYC and DC in an attempt to secure better terms. If these "finalist" cities were simply used as stalking horses with ~0% chances of actually winning, I can't help but feel it was a little wrong to induce them to spend the necessary fees on consultants and planners to put together a decent bid.

> Additionally the window to compute the total net benefit is not bound to 10 years but for many decades, as long as Amazon is there. For instance over a 24 year window (how long they've been in Seattle) the required net additional tax revenue to break even would only be ~$2500 per year.

What discount rate are you using? My quick calculation is already pretty generous to Amazon, since they already get 10 years to generate the 25k jobs. Presumably, some portion of the $1.525 billion is available immediately for campus improvements etc. It's not worthwhile to project 24 years out, especially when considering the time value of money.

> And it's not just salaries that would generate this revenue, it's also property taxes and taxes on services providing for the HQ and employees.

Yes, I mentioned that increased consumption is likely taxable. The question again is whether this consumption would've happened anyway, by other companies, or by Amazon's future employees working at other companies.

> Increased rents and property prices are actually good for local government tax revenue.

Would you be in favor of widespread gentrification? It's not hard in an accounting sense to "optimize" the local tax revenue generated in New York. Just make it a city where cost of living is so high that only the wealthiest can afford to live there. In a city where affordable housing is already in very short supply, and homelessness is still a major problem, I'm not sure how "good" it is for the city to have to spend more on rent assistance.

The "effective property tax rate" in New York is ~0.8% (https://smartasset.com/taxes/new-york-property-tax-calculato...). To generate an additional $1k in property tax, the price of a property would have to increase $125k. Maybe good for the city, certainly not good for the middle-class and poor New Yorkers who have to contend with higher rents and cost of living.

EDIT: Not to mention that it's hard to see how increased rents lead to more revenue for the city. The state may be able to bring in more corporate tax income, but a more valuable property means that property owners can also claim more in depreciation etc.

Some back of the envelope math. Amazon is receiving $1.525 billion in incentives from NY state and city, conditional on creating 25,000 jobs (https://www.nytimes.com/2018/11/13/nyregion/amazon-long-isla...). That's $61,000 per job, paid out over 10 years.

From a purely fiscal point of view, do we expect these jobs to generate $6100 a year in additional tax revenue for the state and city? That would be the "breakeven point". This could come in the form of additional state and city income taxes, consumption that is taxed, etc. Payscale says the average Amazon software engineer makes around $109k (https://www.payscale.com/research/US/Employer=Amazon.com_Inc...).

If we believed that Amazon is actually creating 25,000 NEW jobs, as well as jobs that otherwise would not exist, it seems pretty reasonable to say that the state and city come out ahead net-net. However, I have trouble believing that the "knowledge workers" who will join Amazon would otherwise have been unemployed and underpaid. Factor in the various negative externalities of increased commercial and residential rents for others, potential traffic/congestion issues etc., and it seems like Amazon got a sweetheart deal.

I was made aware of this posting by Packt on Upwork. It seems that they pay their video course authors $1000 and a 16% royalty for a 3-4 hour video course.

This seems to be comically low. Packt seems to have a reputation on HN for being a "lower end" publisher, but this puts those $10 fire sales they have on content in a new light.

If you reference the article in that comment thread above:

"Large technology companies and industrial conglomerates such as General Electric also take a decentralized approach. But retailers tend to favor an integrated model. That way, different divisions can be compelled to make sacrifices, such as discounting goods, to attract shoppers to stores."

Competition sounds great in theory. At Sears, Lampert divided the company into 30 units, each with its own profit and loss statement.

Space in the weekly circular was farmed out to the unit bidding the most, resulting in:

-Toys for boys being advertised on Mother's day, because that department bid the most for space in the circular

-Lingerie being featured next to tools.

---

Because each unit had its own P&L, and its managers were compensated accordingly:

-One Sears tool brand choosing to use an external battery supplier, because it didn't want to pay royalties to a Sears division that made lithium-ion batteries

-Sears stores not opening early for Black Friday in 2011, despite all its peers doing so, because to do so required agreement among the 30 business unit heads

---

It is perfectly reasonable for retail buyers to consider what mix of products to use, and whether to devote shelf space to external or house brands.

It is not reasonable for the instructive structure to be constructed in a way where the company loses customers and sales because each individual unit's decision makers have a financial incentive to do so.

No other major retailer operates/operated this way, and I'd wager that no one will try doing so in the future.

Kindly refer to the comment thread here: https://news.ycombinator.com/item?id=18086998.

No other major retailer creates completely siloed divisions, each with its own C-suite of executives. No other major retailer throws common-sense merchandising tenets to the wind in service of some idea of a "free market" of ideas.