I played this game during orientation when I started my MBA at MIT Sloan (the game was originated at Sloan). In my second year, I facilitated the game.
The most eye-opening thing for me was that a good quarter of students were frustrated and surprised by the bullwhip effect. I'd guess the general population would be even less understanding of a bullwhip effect. I think that help explains why it seems so many people don't understand why it's difficult, even for the U.S. federal government, to create more supply of hand sanitizer, respirators, etc.
P.S. To be clear: I'm not suggesting the federal government isn't to blame for a lack of supply - we should have had a larger strategic reserve of pandemic mitigating supplies prior to the crisis, if only due to the risk of biological weapons (those preparations would have been just about as useful in a non-human-caused pandemic like COVID-19). Nor am I saying the government can't do more right now. The Beer Distribution Game merely helps partially demonstrate why manufacturers, suppliers, and the government (especially if they themselves understand the bullwhip effect!) don't instantly will pandemic supplies into existence.
The Wikipedia article only has a cursory overview, but I’m strongly reminded of the coordination issues I saw in software rasterizer pipelines, with respect to queuing, load-balancing, and buffering. A lot of novice engineers who approach parallel SW rasterization are surprised when they see production systems & the “buffers” between phases are “one deep” (say, 1024 samples; 128x16 vertices; etc). Explaining why this is both: 1. optimal; and, 2. the stated design goal, is always the first task when onboarding them. (Briefly: we size the phases so they’ve got soft real time guarantees, so that the pipeline is always “smooth”; in a SW rasterizer, you can never “get ahead”, so regardless of your buffering size, the buffer is always “empty” or “full”—never “in between”. Before starting work, you check if your downstream buffer is full, or if your upstream buffer is empty; if so: do something else! That sort of system can be modeled with a buffer of size==1!)
It's the same for supply chain professionals.Not sure howyou played and facilitated it,but the versionI had had a first round with no communication. During which all teams, also those consisting of seasoned supply chain people, just fell victim to the bullwhip effect. everytime. Me to, and I knew the effect and the rules of the game.
This doesn't actually demonstrate that it's difficult, just that it's expensive. If the government is willing to eat any sunk cost from oversupply, everything becomes predictable. In the case of COVID, it's hard to imagine any quantity of hand sanitizer and respirators doing more damage than the harm they prevent.
you deal those with scenario planning, which might be the sole responsibility of pandemic unit.
We use casual loop diagrams to understand effects when math cant be drawn out on abstract problems.
Its insanely expensive to satisfy 99% populations requirement versus 95% of populations requirement but when that happens and if the consequence are severe we manage those supply chains with redundancy and the costs are absorbed with other players. Sure, it would profitable to operate without this but my understanding is when this sh*t hits everything falls.
$15 billion savings in 2018 resulted in $12 trillion being wiped out in two weeks.
For one product, yes. For multiple product sharing resources it gets difficult as over production for one product might directly result in shortages for another.
Production hand sanitizer and masks and such is being ramped up, now that Chinese factories are slowly going back online. Things like that take some time, so.
Heh, in my days working in a pharmacy (hybrid service & product world), it was amusing whenever prescribers (service-oriented) felt we could work harder to create stock.
Sure, I’ll just work a few extra hours to grow the crop of plants we extract this drug from.
Or build some more vats that we brew this drug in.
Ha! Same here. Almost started an IAP session with Senge to let people play it multiple times (with varying demand curves) because it seemed to have a lot of exploration left...
Wow, I didn't realise there are so many Sloanies here. I attended a class where Jay Forrester himself came in to teach for a day (in 1998), in his eighties. Impressive man, yet so humble.
Same here. If any of y'all want to shoot the shit about Sloan, feel free to send me an email. I'm cooped up at home in South Florida and desperate for social interaction!
Players are crew members of a 6-room ship, equipped with weapons and generators supplying power (ammo) - all operated manually.
Each round consists of a planning and execution phase. During the former players listen to a pre-recorded track of incoming threats and declare their actions.
In the latter phase the threats start to arrive and the previously planned actions are executed.
The difficulty here is that the planning phase is timed, so the team has to communicate efficiently.
A single bad decision of a single player may doom the whole team, especially that players can and will get in the way of each other.
I remember doing just that - instead of going to the generator, spending a unit of fuel to charge it and distributing power to the turrets I picked the wrong decision card which moved me back from where I started, so my pawn was pressing the right buttons, but in the wrong room.
To make matters worse on my way back I got in the way of my teammate(and informal captain) and prevented him from recharging the generator later, which was important, because our turretwoman was supposed to be firing all the time.
The aliens destroyed that part of the ship killing her and leaving us defenseless.
I strongly disagree. Senge's book is hand-wavy, includes acres of unactionable fluff and -- this irks me the most -- introduces causal loop diagrams but fails to introduce stock-and-flow diagrams.
Causal loop diagrams are basically toys, useful only during initial hypothesis formation. You need a stock-and-flow model to actually test and elaborate your hypothesis.
The best all-round introductory book I have read in this area is still Sterman's Business Dynamics. 1st edition hardbacks are out of print, but there are second-hand copies and also cheap international editions around. A 2nd edition is expected next year.
Your response reads as though you're disagree with an assertion I didn't make. I almost feel like you read my post as saying that The Fifth Discipline is the -best- book on the topic. That I'm not saying, if for no other reason than because I haven't read every book on the topic.
However, I stand by the assertion that it is a worthy read. I gained a lot from reading it, and if it falls short of being totally comprehensive and standing entirely on its own... well, that's a criticism that would apply to many books.
All of that said, thanks for mentioning the Sterman book. I wasn't familiar with it, and I'm about to order a copy.
My most interesting takeaway, from the causes of the bullwhip effect:
> Free return policies: Customers may overstate demands due to shortages, if customers cannot returns items, retailers will continue to exaggerate their needs, cancelling orders and causing in excess product or materials.
I was annoyed when I saw that stores were refusing to accept returns, thinking that they didn't want to suffer the cost of regretful hoarders trying to make right. Most people I talked to were pleased that hoarders were getting their comeuppance. But its effect on supply went completely over my head, and while I don't know if anyone else was aware, certainly nobody mentioned it.
I’ve never understood how you win the beer distribution game, people only ever talk about the effects on the supply chain it seems. But how do you mitigate?
Communication, as soon as people are allowed to exchange information, results get a lot better. Teams can start planning, based on real customer demand,backlogs and so on. This changes everything.
which makes perfect sense because among the assumptions of free market economics which are required to make the invisible hand do its work is "perfect information": buyers and sellers share perfect information about conditions, costs, markets, supply, demand, etc.
and still there will be bullwhips because the future contains uncertainties like "when is the next coronavirus outbreak?"
Silicon Valley operates on secrets (what do Google Amazon and Facebook do with your data? anything that is not simply further stacking the deck against you?) the elimination of which is the direction reform minded people should go with regulations.
I wrote a paper on what we thought was the optimal strategy for the beer game. My team won a beer game session run at the system dynamics society conference about 10 years ago.
1. Push all the inventory to the end of the supply chain (the retailer).
2. Run the inside of the supply chain on a "pass-through" strategy - order exactly what is known to be coming down the supply chain on that turn.
3. Use a control algorithm based on the outflows from the retailer to control production at the factory (the factory player has to watch the retailer's inventory closely). The best parameters for the control system depend on the exact demand deck.
If that turns out to be th solution, the game was played based on wrong assumptions. You have to measure inventorylevels as well, the solution is to define, across perticipants, the amount that has to be produced based on customer demand (simulated by the trainer/ facilitator).
When I was working at UNC I tried to get us to make the beer game in VR for the business school. You would actually see the empty shelves and then the overstocked warehouse in VR to understand the physical constraints of the supply chain and the issues with the bullwhip effect.
Could easily be updated to the N95 mask game. Especially given what I was reading about melt blown production lines capacity and initial install costs.
I adopted it once for aircraft maintenance. Still pen and paper, but the session was hilarious! Especially as I played it with people directly and indirectly involved with aircraft maintenance for the army.
The second round was the real eye opener, so.
"Due to lack of information, suppliers, manufacturers, sales people and customers often have an incomplete understanding of what the real demand of an order is"
Of course this makes the game interesting and also explains why in real life the only sustainable model for economy is planning. The planning starts from actually measuring/calculating the real demand. Then you model your whole supply chain. Of course this means the state has to have control over and complete understanding. This is hard science http://www.strategplan.com/en/
Also, this allows planning measures for emergency cases like epidemics, war and other disasters.
It's impossible to achieve that with monetary approach.
>Of course this makes the game interesting and also explains why in real life the only sustainable model for economy is planning.
"real life" planned economies have done terribly so far, while price-based capitalist economies seem to have done a lot better.
>The planning starts from actually measuring/calculating the real demand.
Herein lies the problem. How do you measure "real demand" for an entire economy, without prices? Prices convey information about what is really in demand at a given time and without them it's remarkably difficult to get answers that don't involve guesswork or individuals demanding more than they really need.
A small movement at the handle of a bullwhip produces an enormous (and oscillating) movement further down the the whip, it's an analogy for movements in customer demand having large effects further down the supply chain.
It's funny you should mention that, because most PID configurations are not very stable. Once you introduce the Integral and the Derivative, if you're not careful that PID will oscillate into the stratosphere.
In fact, I'd say that PID's that are functioning most efficiently are very nearly tuned to the point of oscillating. Even worse is the kinds of adjustments you have to make to PID's when there is significant lag between their inputs and outputs.
Indeed. Control loops are generally limited by the bandwidth of the system they are controlling, and lag in the system limits the bandwidth you can achieve. Trying to get a PID loop to move a system faster than that system's bandwidth will just create oscillations, and the more tightly the PID is tuned the worse its behaviour when exposed to shocks (especially when the system becomes significantly non-linear in the process).
You can extract more performance out of a system by having a very good predictive model of it and measuring its inputs: This can really improve how you drive the system but it's much more difficult to achieve, and you're still limited by how quickly information moves through the system in terms of how you can react.
Yeah, I think that's a much more promising approach. Have a good model, use it to build an observer, now you can control based on the state of your simulated system.
The Beer Distribution Game is one of the touchstone examples for introductions to Systems Dynamics, a field pioneered at MIT with direct roots in control theory. PID controllers are not news in that part of the world.
A major insight of the field is that lags due to accumulation create wildly unexpected behaviours in systems that humans try to control. Many automated systems struggle too.
The most eye-opening thing for me was that a good quarter of students were frustrated and surprised by the bullwhip effect. I'd guess the general population would be even less understanding of a bullwhip effect. I think that help explains why it seems so many people don't understand why it's difficult, even for the U.S. federal government, to create more supply of hand sanitizer, respirators, etc.
P.S. To be clear: I'm not suggesting the federal government isn't to blame for a lack of supply - we should have had a larger strategic reserve of pandemic mitigating supplies prior to the crisis, if only due to the risk of biological weapons (those preparations would have been just about as useful in a non-human-caused pandemic like COVID-19). Nor am I saying the government can't do more right now. The Beer Distribution Game merely helps partially demonstrate why manufacturers, suppliers, and the government (especially if they themselves understand the bullwhip effect!) don't instantly will pandemic supplies into existence.