Hacker News Reader

You have exactly the same effect in play if somebody calls you and ask how long it takes for you to arrive to a meeting point. If you answer 10 minutes, it will be seen as an approximation between 10 and 20 minutes. If you answer 8 minutes, it will be considered as 8 +/- 1 minute.

This is really impressive, this works each time. Especially if you are on ski slopes and you have two groups trying to meet at a given point. 10 minutes == "We can do another run or two". 8 minutes == "We need to pay attention, they are coming right now!"

Per U.S. Census data, relatively few people have a commute in the 5 minute block from 25 to 30 minutes. Only 20-25, and 30-35. The data shows a very clear valley in the 25-30 block.

Suggests people think of their commute as 20 plus a few minutes but rounding down optimistically, or think of it as almost 30 minutes rounding up angrily, never think of it as 27 minutes.

No writeup of the commute data has mentioned this phenomenon, which I suspect was exactly the same thinking you describe.

Do you have a link to this data?

Certainly. See page 3, top right, “Table 1: Travel Time to Work 2019”

https://www.census.gov/content/dam/Census/library/publicatio...

    15 to 19 minutes . . . 14.9%
    20 to 24 minutes . . . 14.1%
    25 to 29 minutes . . .  6.6%
    30 to 34 minutes . . . 13.9%

And you can tell that after saying 30 mins, people next tended to say 45 mins.

// Noticed this while researching for a comment on HN a few weeks back.

> If you answer 10 minutes, it will be seen as an approximation between 10 and 20 minutes. If you answer 8 minutes, it will be considered as 8 +/- 1 minute.

Unless you're talking to a child who has not yet internalized this distinction! My kid frequently asks exactly how much time has passed, and I sometimes have to explain that I don't mean five minutes literally, but figuratively.

child: "When will Daddy be home?" parent: "Any minute now..." child: "I choose THIS minute!!!!"

Another good approach is to explicitly say what you mean: "5-10min", "10-15min", or "8-10min"

Although your answer may be closer to the truth, this paper suggests using a fixed number not divisible by 1 or 5. So, for example, 6 minutes and 23 seconds.

The paper doesn't make any claims regarding interval or ranges whatsoever, only single values.

To replace that with saying 383.512 seconds would signal confidence in the time of a future event, but not in a speaker who is scheduling a meeting with another human.

383.512 is divisible by one, the answer is 383.512 ... :)

But its hard to extend this one to prices and bids. If you give a range of money, people will come back and say "but you said you could do it for [the lowest possible number you gave]"

The answer is to make it clear what the tradeoffs are. For example it could be more cash and less stock. Or lower price but longer lead time. Just offering a range without any other variables would be foolish, and complicated negotiations usually have more factors than just price.

My german friend taught me this. Now I do exactly what you say, if it's an approximate, I say 5-10 mins. If I'm being more precise, I'll say 4 mins, 7 mins etc, which does exactly what you mention: it "suddenly" turns the approximation into a time critical event.

That's a great example, with the arrival time estimates. I wonder whether this effect itself would diminish as more people get to know about it...

And yet, how often are prices at a store still quoted as $4.95? or $10.99? We all know what the psychology of pricing is, and yet it still persists. People are strange and illogical creatures.

Those are "round" numbers by social convention, and our psychology rounds them down.

I think I automatically round 10.99 to 10, 9.99 to 10, 4.95 to 5, etc.

The first case is bizarre to me, unless that's a typo

> Bizarre

They just round more than you: I think your presumption that others should think like you is bizarre. Some people only round to 10 cents. If you are wealthy enough, you could easily round 10.99 to $0.

> I think your presumption that others should think like you is bizarre.

What I actually wrote was "bizarre to me", not just "bizarre" (as in your quote), so I don't think it's fair to say that my comment conveyed such presumption

I think it just proves that x.99 pricing works. It tripped up the poster who was trying to claim they mentally round up.

Which poster was claiming that?

No it's not, 10.99 for some reason works for me

Keeping with the arrival time estimate, and I’m not sure if this answers your question

But I live in a very dense city (Chicago), it’s become almost “set your watch to it” reliable that not only are estimates more reliable from friends who live here when they say “I’ll be there in 8 minutes”, they’re also seemingly more likely to offer such a precise estimate because they know the area, light timings, and traffic patterns and that certain kind of behavior found among drivers in this city.

Compared to when I have friends from out of town renting a car they’ll tell me “GPS says we’ll be there in 10 minutes” I instinctively add another 10 and it’s never failed that they’ll show up and go “that was a long 10 minutes!” And I just smile.

I lived in a part of town that was in the process of infill so a lot of people “knew” it without really knowing it.

It was practically a given that the first time someone visited they would arrive 15-20 minutes later than they estimated. It seemed like an area without a lot of traffic but it was in that uncanny valley where there’s enough traffic where you catch a lot of lights, and you aren’t turning right on red because there are people in front of you, and you may or may not make the left turn light on the first cycle.

Also the whole city is so tuned for one direction of traffic that it takes longer to go across the narrow dimension than the long one.

There was a movie theater that I found I could reliably get to faster on a Friday evening if I drove five blocks past it and doubled back, the lights were that screwed up. 15 blocks faster than 5.

In San Diego we just use a blanket "be there in 20 minutes" no matter where you are going.

...or knew about octal numbers. (Does it even work on old Unix people?)

"ETA 27 minutes plus a few minutes to park"

> One would think a social psychology effect would be less so in a game between quants

One would, but the little I've seen & heard about due dilligence around M&A (and startup investments for that matter) tells me it's nothing close to "quant". People decide they want to buy something, and then try to get it for a good price, and try to have it so that due dilligence works to reduce the purchase price, while not actually finding anything to block the deal. Perhaps in public company acquisitions there's more sophistication, but I really doubt that.

> People decide they want to buy something, and then try to get it for a good price, and try to have it so that due dilligence works to reduce the purchase price, while not actually finding anything to block the deal. Perhaps in public company acquisitions there's more sophistication, but I really doubt that.

That's indeed how it works in my experience. Thay's also why, as a seller, you want to invest a lot of time making sure that the due diligence scoresheet will be as clean as possible - because then, there can be no mechanical way for a prospective buyer to lower the price: only their belief against that of other prospective buyers...

> People decide they want to buy something, and then try to get it for a good price, and try to have it so that due dilligence works to reduce the purchase price, while not actually finding anything to block the deal

Here's a sales secret: this is how most people and businesses buy things most of the time. Key stakeholders (which may not be the big boss) decide they want something then find reasons to justify it. They'll only abandon the deal if there are true blockers either on price or features. And by "big boss" that can be a parent, spouse, Director, CxO, Board, etc. The buyer decides they want a thing and will set about convincing whoever needs convincing. The more you can make them engage emotionally the harder they'll go on securing approval.

If you're a startup selling to companies have one or two impressive "wow" features to make people decide emotionally they want your product, even if they don't seem like important "bread & butter" core features. It makes the rest of the process a lot easier.

Same goes for selling to consumers or even app videos/screenshots: sell people on your product emotionally within the first 5 seconds if you can.

> it's nothing close to "quant"

This is accurate.

For public companies, the quant part has been calculated by the market. Everything that remains is, almost by definition, in the realm of the irrational, unknowable and/or political (usually at the Board level, though government factors could also be involved).

> quant part has been calculated by the market

And then you go on to say "everything that remains is ... in the realm of the irrational", everything you have laid out is in the realm of the irrational. Including the market.

> you go on to say "everything that remains is ... in the realm of the irrational", everything you have laid out is in the realm of the irrational. Including the market.

This discussion devolves into useless semantics if we rely on colloquial definitions. In case it wasn't clear, the predictable, explicable parts of a public company's value are generally priced in by the market. (If they're not, we're in a chaotic environment and nobody's doing classic M&A.) It's the other stuff, which we put in a bucket titled "irrational," M&A bankers address.

This is why the cash-flow models bankers prepare are largely performative. Every buyer builds their own models. And most of the time, the market does the modeling for them. The part the market can't model, the things one seeks to do with control, is the main attraction.

I caveated the comment because private company M&A is different. There, the negotiation does focus on valuation. Quality of assets. Current cash flows. Strength of contracts. Discount rates tied to the actual cost of financing versus an MBA model. While public company M&A dresses financial strengths and weaknesses in the language of partnerships, private company M&A seeks to address political issues (e.g. a senile founder with voting control) through talk of fundamentals.

Thanks, that makes more sense.

Well, the market does follow a set of rules (buy-sell pressure), but it's not necessarily reflective of actual value of a stock. More a heard consensus (which is why we get bubbles).

The other side is also not necessarily irrational, but much less deterministic. For example, there is the question if the board even wants to sell, if they like the potential buyer, if they have a vision for the company that aligns with the buyer and some non-tangibles. Charisma projection and speech-craft definitely has an influence too, as has track record.

The more factors are in alignment with the board, the closer the price is to the stock value. The more hostile the negotiation, the more it moves away.

> it's not necessarily reflective of actual value of a stock

An econ 101 trope about the value of stock ties it to the value of future dividends. This is technically correct, but practically useless. You don't get those dividends unless the Board gives them to you, and investing on the hope that a Board throws you candy is just a hair short of stupid.

The real enforcer of value in the markets is M&A. In M&A, you become the board. If there are cash flows, you can take them. If there aren't, you can't. You're playing a game of looking for things the market and/or management missed or couldn't access. Furthermore, the moment that process starts, the moment negotiations become public, smart people focus on the stock to anticipate those terms. If they think another buyer could see cash flows you don't, they'll bid it up and the Board will take the signal (as will those other buyers and their bankers); if they don't, if post announcement the stock keeps trading below the bid, the Board will take the hint.

PS: in the case of Musk and Twitter, which I believe started this discussion..

It's Musk, so you can throw out the book. He just takes the stock value and then tweaks it so that the offer contains 420 because obviously that's his shtick since some time.

Wonder if the SEC will ever live that one down.

> Musk, so you can throw out the book

Musk's Twitter takeover is a textbook LBO transaction. The negotiations were unorthodox. But that's been true of LBOs since the 80s. Usually the bankers were the ones up to the antics; this time it's the principal.

LBO, Leveraged Byout. I had to look it up. “A leveraged buyout (LBO) is the acquisition of another company using a significant amount of borrowed money (bonds or loans) to meet the cost of acquisition.”

Textbook except for the size of course. Very few LBOs are above $5 billion.

I took this to apply to common things like buying a house, not M&A, but I think when you get a little emotionally invested in buying something the same concepts apply. Specifically, there's a phase where you've decided you want it already, but now you want to get the best price possible. The people selling it know you already want it, so they have all the leverage unless you're completely willing to walk away. In this scenario, setting an arbitrary target price makes sense, but only if you're completely unwilling to negotiate.

>> If you and I are both estimating the value of a good, if our estimates are close it's one thing

Yeah, if I say €193,345 and they say €1 million, that last five euro would look a little stupid. This is a scenario where already being close to knowing their closing figure is everything.

The proper analysis here would be: What's the chance you get your <1% edge versus losing a deal by being stuck on a number too early?

The way M&A usually works is that all sides have bankers/advisers who run Excel spreadsheet calculations that come out with different precise numbers depending on a couple of scenario parameters. Those are put on PowerPoint slides and presented to management. The decision whether to go with one of the precise numbers (or some aggregate thereof), or some close by round number has more to do with bidding tactics, game theory and psychology, so I think this research actually quite interesting.

If you're a manager involved in some bidding process, what this article is telling you could be somewhat loosely rephrased as "use the Excel numbers rather than the ones that stuck in your head after the presentation", which is interesting advice.

I've seen a few M&A experts run their spreadsheets and then the negotiations happened. In reality 90% of the outcome was determined by how competitive the bidding was and how willing the target company seemed to walk away. All spreadsheets went out the window in the first phase.

If your Excel is done properly, you wouldn't have the sigfigs for a precise number.

It seems like it would also affect the mental granularity of any counter offer. If I bid $10,000 someone might be like no way to high how about $8,000! Cause the significant digits were working with are in the 1000s. But if I say $10,542. They might be more likely to focus on the detail and come back with 10,100 or something.

> I don't know how M&A "works".

In general it's the buyer who names a price first. Take the current Twitter acquisition, the board weren't trying to sell the company at all and Elon Musk made an unsolicited bid of $54.20/share.

Other times the board may be actively looking for a buyer and have a price in mind, but it will still typically be up to the buyer to propose a deal. Not least because the number isn't the only thing being negotiated: the buyer will likely have a range of stipulations or conditions attached to the bid so every bid will look different.

> If you and I are both estimating the value of a good, if our estimates

> are close it's one thing and if they are far apart it's another. Are

> rounders losing because their estimates are less accurate, relative their counterparty's?

I belive it is normally the one with the most optimistic evaluation that wins. It is called "winner's curse": https://en.wikipedia.org/wiki/Winner%27s_curse

I also suspect that sometimes the bid price is a signal to others besides the company being bought - and a number that’s easy to digest might help in media, etc.

>One would think a social psychology effect would be less so in a game between quants

Is the presumption that a "quant" bases their decision on logic alone and that's why they wouldn't be susceptible so social psychology? Maybe someone with more background can elaborate but I think behavioral psychology might refute that assumption.