Guesstimate – A Spreadsheet for things that aren't certain(getguesstimate.com) |
Guesstimate – A Spreadsheet for things that aren't certain(getguesstimate.com) |
Update: Matthew (the other cofounder) and I got Guesstimate to a stage we were happy with. After a good amount of work it seemed like several customers were pretty happy with it, but there weren't many obvious ways of making a ton more money on it, and we ran out of many of the most requested/obvious improvements. We're keeping it running, but it's not getting much more active development at this time.
Note that it's all open source, so if you want to host it in-house you're encouraged to do so. I'm also happy to answer questions about it or help with specific modeling concerns.
Right now I'm working at the Future of Humanity Insitute on some other tools I think will compliment Guesstimate well. There was a certain point where it seemed like many of the next main features would make more sense in separate apps. Hopefully, I'll be able to announce one of these soon.
One of the triggers for the financial crisis in '08 was that the Monte Carlo pricers assumed the various risks were much less correlated than they actually were.
For example, they largely assumed that it was unlikely for many mortgages or underlying MBS securities to simultaneously default (low correlation). This is how many AAA rated CDO securities ended up trading at 50%+ discounts.
IMHO, any multivariate Monte Carlo analysis that doesn't show your sensitivity to correlation is essentially useless, since your answers may change completely.
In the second example model (https://www.getguesstimate.com/models/316), Fermi estimation for startups, you would expect many of the inputs (deals at Series A, B, C, amount raised per deal) in real life to be highly correlated with each other since they all depend on 'how well is VC in general doing right now?'
The final estimate of 'Capital Lost in Failed Cos from VC' has a range of 22B to 39B, this seems way too low. The amount of VC money lost during a crisis (like in '01) can easily be an order of magnitude more.
Guesstimate doesn't currently allow for correlations as you're probably thinking of them. However, if two nodes are both functions of a third base node, then they will both be correlated with each other. You can use this to make somewhat hacky correlations in cases where there isn't a straightforward causal relationship.
Implementing non-causal correlations in an interface like this is definitely a significant challenge. It could introduce essentially another layer to the currently 2-dimensional grid. It's probably the feature I'd most like to add, but the cost was too high so far.
I think Guesstimate is really ideal for smaller models, or for the prototyping of larger models. However, if you are making multi-million dollar decisions with hundreds of variables and correlations, I suggest more heavyweight tools (either enterprise Excel plugins or probabilistic programming).
This rational is why this tool shouldn't be used for anything consequential -- like business decisions that can sink you company.
Also, are we able to adjust the distribution of each variable?
EDIT: I think you can actually manually create correlations on sheet so it should be fine
Thank you!
I'd imagine an Excel plugin to do something similar would be valuable.
The first is to use Excel apps like Oracle Crystal Ball or @Risk. These are aimed at business analysists. They're pretty expensive, but also quite powerful.
The other option is to use probabilistic programming languages. Stan and PYMC3 are probably the best now, but hopefully, some others will become much better in the next few years.
That said, this is a pretty small space. The main "business competitor" is probably people just using google sheets or Excel without distributions to make models.
Crystall Ball: https://www.oracle.com/applications/crystalball/ @Risk: https://www.palisade.com/risk/default.asp Stan: https://mc-stan.org/ PYMC3: https://docs.pymc.io
I imagine there a bunch of cases where the defaults would not work like you're trying to do error propagation (all normal distributions) or you're trying to compute interval arithmetic.
Is it the case that if you input a range which span multiple orders of magnitude then you get lognormal rather than normal?
I might not be exactly the target audience, but I would appreciate a more in-depth of the math and heuristics involved
EDIT: I found this on their blog
https://medium.com/guesstimate-blog/lognormal-normal-833bf41...
We have some documentation [here](https://docs.getguesstimate.com/), and some in the sidebar entree.
Generally, we recommend lognormal distributions for estimated parameters that can't be negative. This works when you span multiple orders of magnitude, though it's possible you may want an even more skewed distribution (which is unsupported).
I may be able to make a much longer video introduction some-time soon.
I think if I were to start again or spend much time restructuring it, I'd probably focus a lot more on enterprise customers. That would be quite a bit of work though, so I don't have intentions of doing that soon.
Spot on, it’d need to load some data on final predictions. Or, it could dump the model in a way that another software could use it.
I was thinking of the same thing.
I love that it was a no BS signup and start using. Super clean and easy. It would be great to be able to show data on GIS as well - effectively showing the outcomes geographic representations. Ill see if the data I was looking to work with today will work with this tool meaningfully.
You can use tools like distshaper6 to generate arbitrary distributions, then copy the samples into Guesstimate.
http://smpro.ca/pjs/distshaper/
Guesstimate doesn't yet support an input format for distributions outside of via samples.
It's also just a fancy name for generating a pseudo-random number from a Uniform distribution in [0,1], and reading off the x-axis of the CDF.
[1] https://en.wikipedia.org/wiki/Inverse_transform_sampling
"If 5 people show up at my house tomorrow evening, I'll hold a poker night." 10 people were invited and 4 of them RSVP yes and 2 of them RSVP no. It looks like there's a 95% chance I'm holding a poker night tomorrow.
"The X team has a monthly meeting on the 1st, never fail. They haven't decided on the location yet, just that it's on the North Side." As the team members pick possible locations, the possible locations appear more distinct until one is chosen.
=A5*A6
=interest*principal
[1] https://www.ablebits.com/office-addins-blog/2017/07/11/excel...
Here's another one I did to ballpark a pension plan: https://www.getguesstimate.com/models/11133
Another thing I like is that you can do simple statistical reckoning for it. For my job, I often have to benchmark something several hundred times with or without a patch applied. It can be bit difficult to put "on average x% faster" in context when the benchmark is noisy, but Guesstimate allows you to answer questions like "assuming somebody ran one run of this benchmark with the patch, and one run without it, what's the expected range of performance improvement that they'd see?" with the actual numbers that you get out of the benchmark: https://www.getguesstimate.com/models/11850
Anyways, it's a histogram, so the x-axis is split into buckets. The bar all the way on the left is likely some range of hours from 0 to whatever the bucket size is
> where there isn't a straightforward causal relationship
One way to interpret a global pairwise correlation is simply that the person building the model is being systematically biased in one direction—either being too pessimistic or optimistic. This is a 'non-causal' relationship but often the biggest contributor to variance between the model and the real world.
Philosophically, this is a bit like the difference between 538's modeling approach and Princeton Election Consortium's for the 2016 election—the former gave Hillary a 2/3 chance of winning, while the latter ascribed a ~99% chance.
The risk of leaving modeling error out is that you'll end up with much more confidence than is called for—it feels very different to come up with a point estimate (I'll save $10k this year) vs. a tight range (I'll save 9k-11k this year), if the true range is much wider.
In the former case you know your point estimate may be very far off, but in the latter you may be tempted to rely on an estimate for variance that too low.
> It could introduce essentially another layer to the currently 2-dimensional grid
You could probably get away with doing almost all of this automatically for the user as long as the decide on what the 'primary' output is:
- For every input, calculate whether it's positively or negatively correlated with the output
- Apply a global rank correlation to all the inputs with all the standard techniques, flipping the signs found above as appropriate
- Report what the output range looks with a significant positive correlation (usually the negative correlation case isn't as interesting)
So basically, draw a multivariate correlated standard normal with the intended correlation. Transform the standard normal draws into quantiles of the standard normal (this works the marginal distributions are also normal). Now you have X and Y quantiles for your target distributions and you can draw from them.
The distributional transformations will slightly attenuate the correlation, and the choppier the distributions the more attenuation you'll get. Additionally, while you can do this for 3+ variables at once, there are constraints on the possible sigma matrix describing n-dimensional correlations.
If the variables have no fixed distribution you can use the eCDF of the actual data, so it's even possible to import e.g. population and income data and produce a permutation that gives you the correlation desired.
I agree that it is fairly difficult to do this if you have arbitrary DGPs with complex interdependencies in them--if you correlate on variables A and B, then it is difficult to guarantee the observed correlation between f(A) and g(B) and vice versa--but still you can provide a lot of utility with the copula method.
That is one heavy model. My computer comes to a standstill trying to open this one.
I am curious though what was the result of the analysis?