Global Infrastructure Partners Gets Big Returns In Sleepy Sector
bloomberg.com
bloomberg.com
Obviously it's "dumbed down", but it still tries to stay realistic & uses a 'systems-based' logical approach for gameplay; the total operation is basically just the sum of all the different inter-dependent systems and how they relate to eachother, how failures cascade, etc...
Disclaimer: I'm involved in the project.
That would mean you don't need space for restaurants, or lots of space for people to pool as they're waiting past security.
It would be very interesting to experiment with changes like that and see how it would look.
Edit: There's now an entire discussion downthread about how airports did and didn't look like before 9/11.
That misses the point of what I'm asking, which I'll rephrase. I'm asking as a simulation game, how customizable is this? E.g. in in SimCity 2000 it initially looks like you can do anything, but really you can't design something that doesn't look & function either like an American suburb or car-centric Los Angeles. There's no way to construct something that looks & functions remotely like the city of Venice in SimCity.
Similarly in this game, can you really design your own airport that works anyway you want from scratch? Or are you just placing all the pieces in your typical international airport in different places on the board, while the NPCs walk in some pre-determined route from parking, to checkin, to security etc.
There's nothing that appears in the very brief trailer that suggests that it isn't highly constrained in the same way SimCity is. But I'd love to know if that's really the case.
Restaurants existed in airports long before 9/11. People arrive early because missing flights costs you hundreds of dollars. Security adds maybe 20 mins to the process.
I'm going to guess that pre-booking is some sort of way to maximize revenue or minimize wasted empty seats.
It also allows them to price discriminate against business travelers who tend to have deeper pockets (flights expensed to their company) and tend to make lots of last minute travel changes (need another day to close a deal, etc). Losing the ability to ramp up the price on these types of travelers would mean higher fares for everyone and likely fewer ticket sales.
I'd much rather cope with the Christmas rush by booking months early than by waiting at the airport for days.
However I've never run an airport so I can also imagine that there are important factors that I don't even know about that make this unworkable.
The hourly flights still exist (now run by American, which inherited the service from US Airways in the merger, which bought the service from Donald Trump's creditors after he defaulted on the loans he used to buy the service from Eastern when Eastern got into financial trouble), though of course now you have to buy a ticket before you go through security.
Plus, they ask you to be even earlier than that just in case it might take longer, so you're stuck there for 45+ minutes after security.
The layour is 3 separate terminals which are each a semicircle ring: http://www.metafares.com/images/airports/MCI_Airport%20Layou...
It is great to fly in and out of. It feel optimized for people to get on and off planes quickly, and get out of the airport fast.
It is awful to have layovers in or have to change terminals in for a flight: you have to leave security for everything for one. And the waiting spaces are cramped and unfriendly.
After 9/11 it's still extremely fast to get in and out of, but the long waiting before flights is unpleasant. - Personally I would love to see airport design focus on passengers and less on retail space. All the proposals to replace KC Airport brag about all the shopping and restaurants... when all most people care about is spending as little time there as possible.
The MCI airport had the misfortune to be designed and built almost right before this all happened. TWA designed the terminal according to their vision of the future of air travel - drive right up to the gate, walk a couple steps onto the plane, and away you go. The new security requirement invalidated this vision almost right away, and made the design so unworkable that TWA moved their hub to St. Louis only 10 years after opening their brand new "airport of the future" in Kansas City.
[1] https://en.wikipedia.org/wiki/The_Skies_Belong_to_Us [2] https://en.wikipedia.org/wiki/Kansas_City_International_Airp...
Perhaps there's only so much that can be done about #1, but regarding #2, I wonder if we could rethink airport security once we have completely autonomous planes. If the plane is controlled completely by autopilot and/or remotely, it's basically on tracks, so I can't see any logical argument why we couldn't then take the same approach to domestic flights as to domestic trains. (You could argue that that's already the case with locked cockpit doors, but there's an argument on both sides there.)
We have definitely had discussions on this exact topic -- potentially allowing absolutely zero security -- and it's something we may revisit in the future. It's a bit of a fine line when it comes to both not being too outrageous on the political-correctness scale, as well as being able to model/simulate the risks in a not-too-overly-rare and still fun way.
Honestly though, these are exactly the kinds of discussions/concepts that we really want the game to allow for! :-)
I remember the days as a kid when you could just walk out to the tarmac, climb a stairs and board a plane. Life was so much simpler, then someone hijacked a plane to Cuba and it all suddenly changed virtually overnight.
The big problem with modeling incidents is going to be that it has very little to do with airport security at all, which is almost all pure theater. For example screening for knives is useless because anyone can make a shiv in an airport bathroom the same way prisoners do. Even explosives are not particularly any more dangerous on a plane than they would be in a subway car or anything like that (in the sense that they can kill 100% of passengers in both cases). Checking for ID only causes the terrorists to choose a person with no priors as the one who carries out the attack.
And attacks are adaptive. Banning twelve ounce bottles when you can have both ten three ounce bottles and an arbitrarily large empty tub is classic theater.
The main reason there haven't been any repeats of 9/11 is that as of 9/11 all passengers are on notice that resisting hijackers is mandatory, and even a terrorist with a knife can't fend off 100 passengers at once.
So much of the security theater also has financial motives. Checking ID prevents the resale of non-refundable tickets. Prohibiting 12 ounce bottles encourages people to buy one inside the airport for five times the grocery store price. Makers of expensive scanning equipment are in favor of mandates or subsidies for expensive scanning equipment.
You could add a lobbying component to the game. If you're an airport you want to defend the profit-generating ID checks against criticism from privacy and immigrant advocates, but oppose mandates for security scanners you have to pay for when their manufacturers lobby for them.
> We have definitely had discussions on this exact topic -- potentially allowing absolutely zero security -- and it's something we may revisit in the future.
One answer might be to let the user choose what country to set up shop in, and then some hypothetical countries might have less security bureaucracy but other disadvantages (like not being able to have direct flights to high bureaucracy countries).
I'd never really thought about shivs before, but I've always felt that the available materials in duty free (after security) must be more than sufficient to blow a hole in the side of a plane (lots of aerosols etc)?
I guess the additional danger comes with a plane lands somewhere and a train has far less potential to cause massive collateral damage. However looking at the tube bombings in London it is clearly possible to make a decent fist of terror using trains.
>The main reason there haven't been any repeats of 9/11 is that as of 9/11 all passengers are on notice that resisting hijackers is mandatory, and even a terrorist with a knife can't fend off 100 passengers at once.
I disagree. The changes to cockpit doors are far far more relevant, it's simply not practical to take control of a plane through threat of violence (plus all pilots are now aware they won't necessarily be hostages - they'll be weapons). Yes a bladed weapon will only let you harm a small number of passengers before being overwhelmed - but a couple of guns probably achieves the same. The clear difference being the chance of depressurising the cabin.
The reason explosives are only a minor threat to planes isn't that it would be hard to get one through, it's that the threat it poses isn't specific to planes.
> I guess the additional danger comes with a plane lands somewhere and a train has far less potential to cause massive collateral damage.
That isn't much of an attack vector when you can't choose the target. Nearly all of "somewhere" is bodies of water and open space.
> However looking at the tube bombings in London it is clearly possible to make a decent fist of terror using trains.
It's not about trains either. Someone could drive a car bomb into a nightclub.
A lot of the smaller attacks that have actually happened would have been dramatically worse if we hadn't lucked into the fact that the terrorists were incompetent. Although there may be a shared causation in that correlation.
> The changes to cockpit doors are far far more relevant, it's simply not practical to take control of a plane through threat of violence
Those are also a help, though somewhat of a risk too if a terrorist actually managed to get into the cockpit.
> Yes a bladed weapon will only let you harm a small number of passengers before being overwhelmed - but a couple of guns probably achieves the same. The clear difference being the chance of depressurising the cabin.
That changes nothing. Given the choice between certain death or likely death while saving a thousand people on the ground, the decision remains clear.
An omnipotent system manager cannot exist on complex systems, that's why we delegate.
If you're affiliated with, partners with or have licensed the name from EA/Maxis, then never mind!
These guys create a firm, pitch their friends to get a seed fund off the ground and then start fundraising from larger institutions. They then hire some real talent and they are off to the races. Not everyone makes it, but the ones who do are featured as geniuses on Bloomberg and the Wall Street Journal. Survivorship bias for sure. Most Hedge Funds and Private Equity firms don't do well, similar to any start-up.
If everyone (including GIP) has modest (much less than the historic 24% of GIP) returns, then we might think the opportunity was simply about untapped inefficiency, and that any team with the massive capital + reasonable human talent would have had the same results as GIP in the past decade.
If GIP continues to outperform vs the competition even as the space gets crowded, we could consider their human talent to be a differentiator.
If all the infrastructure funds are doing great, even in a world of high competition, then again we can consider the access to capital and access to the counterparty stakeholders to make the deal happen to be the differentiator, not the operational or engineering talent of any firm.
Personally I'm willing to believe that operational experience at massive scale is much rarer than simply "great engineering talent" at the individual contributor level, but am open to be proven wrong.
Does it really take an "engineer" to find a trouble-spot and fix it ? Why do you need a billion-dollar fund to grab engineering talent to do what the highly-paid leaders of, say, airports, are already supposed to be doing ? If they can't figure it out, isn't there a whole industry of consultancy firms that send "analysts" through your organization to do the same thing?
If this sector is so profitable, will it cause the same crisis that privatization of hospitals did to healthcare ? Say "Global Infrastructure Partners" expands and "buys" JFK, ATL, LAX, just like private healthcare bought all the hospitals. Now they can start jacking up their landing fees (each year) until air travel becomes unaffordable. What's to stop them from turning this into the ultimate regulatory capture ?
Actually, I find it obscene that governments that can borrow at basically 2% per annum or less are not engaging in these projects, and instead leave 24% returns on the table for the private sector.
Call me a Keynesian all you want, but yes - the austerity stance in the last decade (after the GFC) (particularly in Europe, due to Germany, but also in the US) was a huge mistake.
Ironically, GIP owes its airport holdings to the UK government coming to that conclusion about the market power of the airport operator formerly known as the BAA, and forcing it to sell some of its airports. Its growth potential has depended on the success of lobbying efforts to allow it to expand LCY (was going through at the time they sold it) and ideally a second runway at Gatwick (still hoping for).
The politics might have more effect on its bottom line than buying in better logistics systems but it's probably less sexy an anecdote for an article. I'd definitely think they were better off hiring one really smart process guy as COO than contracting top tier consulting firms to throw dozens of graduates at the airport's operations issues though.
If you read the bios in the article, one of the investment banker guys got his start in Mergers and Acquisitions in the 80s, so it's the kind of approach they would take since it's what they knew and also what the people fronting them the money would be able to relate to. If he had gotten a start in consulting work instead, it would have been a different approach and a different story.
Most popular destinations already have a choice of alternative options (EWR, LGA, BUR, SNA, ONT) and the ones that don't are big only because a major carrier decided to build a hub there (ATL, SLC). Carriers will not jump in with huge commitments until some long-term contracts are in place.
In Southern California, for instance, Santa Monica charges a landing fee, and nearby Van Nuys is suddenly the busiest general aviation airport in the nation.
Two sigma corresponds with a 95% confidence interval, while one sigma is a 68% confidence interval which he is presumably referring to.
I don't want to detract too much from the work that these people did, which I think is actually very cool, but it worries me that their head of engineering can't discuss really simple statistics correctly...
Edit: words mean nothing, see replies
That's exactly how I learnt it in secondary school, the standard deviation, what's the problem?
> Motorola has determined, through years of process and data collection, that processes vary and drift over time – what they call the Long-Term Dynamic Mean Variation. This variation typically falls between 1.4 and 1.6.
https://www.isixsigma.com/new-to-six-sigma/dmaic/15-sigma-pr...
31% outside +/- 1 standard deviation ("2 sigma"),
5% outside +/- 2 standard deviations ("4 sigma"),
0.3% outside +/- 3 standard deviations ("6 sigma"),
100% outside +/- 1 standard deviation,
25% outside +/- 2 standard deviations,
9% outside +/- 3 standard deviations,
- are at a smaller and more local scale that don't attract attention from the likes of GIP - the process improvements are around information flow between applications instead of physical flow
I am really interested in talking to you whatever the domain may be.
An example I've learned of recently:
Garbage truck routing in NYC is grossly inefficient. The routing used to be centralized under a single contractor with rumored tentacles into organized crime. Due to changes in contracting regulations, NYC garbage collection is now includes a mix of independent contractors. This results in multiple trucks visiting the same street on the same night.
This is inefficient in the following ways: - cost of garbage transportation - noise pollution - localized pollution from truck emissions - CO2 emissions from inefficient route planning
I'm fairly certain efficient routing for this is a solved computational / information systems problem and the source of the issue is a lack of coordination. I'm trying to dig in with the Department of Sanitation to see who the different players are.
i want to see an airline that is run by software engineers with aviation experience that rolls its own GDS and revenue management system; one that can scale well beyond its years and can operate on clusters of cheap compute.
my hypothesis is that current airline operations are band-aids on top of protocols and runbooks that were developed around the initiation of the jet age, and fear and fearfully low margins has held back massive amounts of innovation that keep prices higher than they should be.
(and yes, i understand that aircraft are expensive to buy and maintain)
No one really knew how it worked, it was built 20 years ago, and no one was allowed near it for fear of breaking it. Therefore, instead of tuning the algorithm, the analyst added their own (mostly arbitrary) rules on top of the system's calculation.
It reminded me of that conversation in Asimov's Foundation, where some guy goes to some planet and is afraid to see everyone using technology, having lost all the knowledge of how it actually works and how to maintain it, therefore seeing it as some kind of magic.
This was in 2015.
To be fair, airlines wouldn't exist without governmental intervention (aircraft and airport usage is expensive, but the market wants rock-bottom pricing), so this makes innovation tricky. Which is exactly why I'm so interested in this space.
I believe that most engineer would be able to come up with a reasonable way to manage an airport, maybe with more dumps that our heroes, but still I believe that most of us will be able to dig themselves out of the problem.
But then, how do you get into it?
You need tons of money and wealthy/powerful people that trust you, how do you get that trust?
because then you get to keep your youth while you have the respect as well.
Wanting the skip all of that and just get a cool billion dollars because you have a big idea is naive, to say the least.
How does anyone do anything big? Get lucky with parents/work hard and network with the right people/start ambitious projects/some or all of the above.
Wondering how to get all wealthy people to trust you is the wrong question: What you want is ONE wealthy person to trust you. 300 failed pitches don't matter. You just want to succeed once.
Therefore, my strategy is to aim as high as you can, be ok with a bunch of failures, and once you have success, consider whether you can use it to aim even higher, now that you have established a higher baseline. It's better to have one resounding success an 50 failures at convincing people that never failing, but aiming way too low.
I do agree that shooting for the stars is the way to go, but I still find it difficult.
A few random things I learned:
* While a billion dollars may seem like a lot of money, it's actually fairly small in the infrastructure world where you are building massive projects like airports and power plants.
* For that reason, our fund focused on higher-risk, smaller projects in the developing world, which made it a tough sell to the cautious pension and endowment type investors who invest in infra.
* The secret to doing well is deal access. We were always envious of GIP because they had an incredible network.
I think this is why people are dislike big government (me included, but necessary for some critical things) because there is no incentive to optimize. Quite actually, that is punished. Don't spend all your money one year and your budget gets slashed even though you may need that in a few years to upgrade equipment. The more layers of organization you add, the more managers you have who see their job as self-preservation. It looks like these guys run very very flat.
That government can be inefficient is an efficiency problem.
That private actors basically "steal" taxpayer money is a lobby/regulatory problem.
Government does have some blame for the former, but you can't simultaneously regulate wall street (for example) and still blame the government for being in bed with them.
Is that really what two sigma performance means? That sounds more like one sigma worth of failures.
Airplanes don't run out of fuel mid-flight, or crash 30% of the time, my luggage doesn't fail to arrive to its destination 30% of the time, and I've yet to show up to a flight, only to discover that my seat in just a hole in the floor.
Now, the flight attendants misunderstand my choice of mid-flight snack ~30% of the time, but that doesn't really matter, because it's always ~12 grams of peanut-related animal feed, anyways.
It's not always binary like "Did the bag get lost or not?"
To quote Willy Wonka, "They lived happily ever after."
I gave up on my first...
I bet graycat knows, care to chime in? Anyone else have a guess?