The Entrepreneur Forum | Financial Freedom | Starting a Business | Motivation | Money | Success

Welcome to the only entrepreneur forum dedicated to building life-changing wealth.

Build a Fastlane business. Earn real financial freedom. Join free.

Join over 80,000 entrepreneurs who have rejected the paradigm of mediocrity and said "NO!" to underpaid jobs, ascetic frugality, and suffocating savings rituals— learn how to build a Fastlane business that pays both freedom and lifestyle affluence.

Free registration at the forum removes this block.

The Kelly criterion: when to start entrepeneuring

quicksort

New Contributor
User Power
Value/Post Ratio
100%
Oct 9, 2015
18
18
34
tl;dr: let's apply the Kelly criterion to entrepeneurship, I don't think I've ever seen anyone do it before. I was thinking about it, then decided to finish UNSCRIPTED before asking here, just in case MJ mentioned it. I was excited to see that MJ mentioned expected value, but then sad when the Kelly criterion didn't show up. Looks like there are no results for it here:
.

I am afraid to take some big risks. The kind where you can lose everything. Russian roulette kind of risks.

But I paid my bills for a short time in college by playing online poker. I love taking all the big risks that I can as long as they are EV+ and "safe" for the amount of money I have.

If you want to make it as a poker player, you never go all in with everything on the line.

You start by grinding up from the lowstakes, because you don't have much money, so it's not "safe" to start higher (unless you want to keep depositing- not gonna happen for a broke student!), and you probably suck and are unprepared to continuously make good, EV+ decisions.

How much is safe? Well, it's hard to say. It depends on the game format. I played sit n goes (SNGs) which were small, single-table "tournaments" with an entry fee and a limited amount of chips. Rewards went to the few players remaining, but I played 1v1 ("heads up") SNGs, or HUSNGs.

We could determine which stakes you could play at by plugging your winrate and the possible outcomes (win/loss) into a formula called the Kelly criterion: Kelly criterion - Wikipedia

Let's say I have been saving my money from a slowlane job.

So when can I quit my job and pursue my own thing, where like a poker player, I am never worried about losing because my lifestyle will almost certainly never take a hit because I know 1. that the expected value of my time is far above my cost of living and 2. that my living expenses + anything else I may spend on my business attempts are no more - and perhaps well below - than the amount I can safely spend according to the Kelly criterion?

If 9/10 businesses fail, according to whoever it is that likes to create and propagate stupid statistics, then how fast can you fail? If you can test and validate/rule out/"hard proof" (7P process from UNSCRIPTED ) one idea per month, then your probability of finding a "hit" in a year is 1-.9^12 = about 72%. In two years? 1-.9^24 = about 92%.

Well, can we do better? As I think I saw MJ write somewhere here, you can do things to increase the odds: NOTABLE! - Moving the Needle on Probability

And it looks like among people here, 10% might be too pessimistic:

Okay, so maybe we have better than 10%. But how many businesses fail AFTER the hard proof stage, assuming you don't make any big execution mistakes that you don't eventually correct?

And how fast can we fail?

If we say we have a 10% chance of any idea passing out validation stage, and assume that:

- 100% of validated ideas will be worth $1m with no extra work needed to realize that $1m (we can get more complicated later and say a validated idea has X probability of success and costs Y dollars and Z months to execute, to also compute an EV of that validated idea... But lets keep things simple for now and assume X, Y, Z all determine each idea is worth $1m)
- you have some ridiculous living expenses like $5k/mo
- you spend $5k/mo on each validation attempt
- every idea takes exactly one month to validate

Then with the Kelly criterion (taken from Wikipedia), the amount of our bankroll (i.e., savings) we can spend per month (which in our example is equivalent to an idea):
f = (p(b + 1) - 1)/b
f = (0.1*(100 + 1) - 1)/100
f = (0.1*(101) - 1)/100
f = (10.1 - 1)/100
f = 9.1 / 100
f = 9.1%

So under the above conditions, where you're spend $10k per month or idea, you must have at least $109,899, let's say $110k saved up.

If it doesn't work out, then you either need to either calculate the amount you'll be able to risk next month (which means you're lowering your living expenses next month- hard to do on a whim for most people) or... You need to start with significantly more than $110k.

In reality, nobody is spending $10k/mo, and $1m sounds too optimistic. But the numbers $10k and $1m seem clean. You can replace them with whatever you like, or we can plug in $2k and $200k if you like, or whatever.

I'm just trying to figure out, and I didn't see this in UNSCRIPTED , how much does it really take? Because it sounds like people think entrepeneurship is "risky" like Russian roulette is risky, when on inspection, it seems to be risky like poker is risky. I.e., it looks like for someone learning, there is virtually zero risk of "failure" (your savings goes to zero and you are homeless or have to get a job) in the sense that most non-entrepeneurs think.

As a poker player, I never considered it "risky" to play poker. It was virtually guaranteed that over any meaningful number of hands, I'd have more money.

I don't think entrepeneurship looks risky either.

I think you could get tons of people to drop what they're doing and become entrepeneurs just like tons of people learned poker a decade ago, if it could be expressed more accurately:
1. How many "hard-proofed" (to use UNSCRIPTED 's language) and CENTS-compliant ideas actually fail for non-execution reasons?
2. How can you maximize the amount of CENTS-compliant ideas you validate and make a go/no-go decision on?
3. Just as you can maximize probability of success by following advice in UNSCRIPTED , can you maximize the amount of ideas you validate in a given period of time? If you could validate two ideas per month instead of one, things get way better. How can you validate faster? Well, there are some b2b ideas where I can only imagine you have to track down decision makers and make calls, etc. You might not be able to validate that often and even when you do, it's hard to be sure people will buy. But then, there are also ideas where you can set up a website and pay for traffic from Google ads over a a week or two. You could probably validate several ideas in parallel that way, maybe 10+ per month instead of just one per month. And if you are taking pre-orders (which you do say you might cancel) or running a Kickstarter, you also can be pretty confident that if an idea passes, it can be executed.
4. How long does it take you all to validate an idea and decide whether to execute or not?
5. For those of you have decided to execute on validated ideas, how many validated ideas failed? Did you find that some types of validated ideas (e.g., expensive b2b service "validated" by asking business owners if they would buy vs product offered for pre-order online where people actually attempted to buy) were more likely to succeed following validation than others?

What do you all think of applying the Kelly criterion here? What do you think of questions 1-5? Do you think I left anything out or am completely wrong about anything?
 
Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.
Last edited:

Ubu_roi

Bronze Contributor
Read Fastlane!
Read Unscripted!
Speedway Pass
User Power
Value/Post Ratio
333%
Jul 24, 2018
84
280
Europe
Very interesting approach.
I think that if you could prove any success formula you would deserve a Nobel...
The difference with poker though is that in poker you have relatively few variables (there are only a limited number of cards after all), but I’m not sure you can really find a formula to predict success in business, which has almost unlimited varialbes . Also, execution is way more important than validating an idea, and it does take time, which would unbalance any time based formula you might find.
I’m curious to see what others think about it though.
 

quicksort

New Contributor
User Power
Value/Post Ratio
100%
Oct 9, 2015
18
18
34
I’m not sure you can really find a formula to predict success in business, which has almost unlimited varialbes . Also, execution is way more important than validating an idea, and it does take time, which would unbalance any time based formula you might find.

True! For these, I wonder if it would be useful to use some kind of "reasonable" bound, based on the type of business. Product? Service?

Or even, you could say something like "if you have a validated [product|human service|saas|etc] idea validated by [google ads and email sign-ups with small sample size|google ads and pre-orders with large sample size|asking three business owners, two of whom are your friends who don't want to hurt your feelings] and make no big, obvious mistakes that go uncorrected, the odds of success that others have seen has been about P."

Also might be useful: if we could say something like "saas services have been harder to reliably validate. Their post-validation success probability has been observed to be P and their time to execute has been observed to be measurable in quarters if not semesters or years", then maybe we could hone in on the types of businesses that could be less costly in terms of time, and bounded, like "this is a business where you can hire a VA to do the work the same day you find a customer who pays up front, the time to execute is measurable in weeks and you can attempt a new one every month if you want until you find something that someone will buy".

With that, we do you think we could then say "if you stick to THESE kinds of businesses, then you can probably execute this many validated ideas per month/quarter/year"?

Then, just as a tutorial for a newbie poker player might say:
"start at these stakes where you can afford to make bets with the money and time you have, and then you can move up to these higher stakes as you win", maybe we could say

"As a newbie, you likely don't have the resources saved up to *safely* bet on these kinds of businesses. if you started with these other lnes, it would be much safer and you can always revisit the ones you want after you pick up a win or two"?
 

quicksort

New Contributor
User Power
Value/Post Ratio
100%
Oct 9, 2015
18
18
34
As a former professional poker and blackjack player, and a current professional investor, I love the *idea* of using the Kelly Criterion for business and entrepreneurship. Unfortunately, there are two big reasons why this isn't going to translate perfectly to business:

1. Too few trials (you likely won't start enough businesses in your life that KC will optimize your results);
2. Inability to accurately measure risk.

You can reduce the impact of #1 by applying KC to each business decision (as opposed to each business), but that just exacerbates the variance in #2, so it doesn't really help.

In general, my recommendation is to focus on risk reduction early in your business (before you start putting any significant money in). Most business have some inherent risk, but in many cases, you can identify and mitigate that risk before making any substantial investment. Or worst-case, you can abandon the effort before making any substantial investment.

You do this by testing the market, identifying the need for your product/service *before* you spend money building or deploying it. In many cases, you can actually get customers to commit to purchasing your product or service before you start spending any significant money. In this way, the bulk of the risk is taken prior to infusing the bulk of your capital into the business.

Of course, the trade-off is that you'll need to put more *time* into the business (there's always a trade-off between money and time), but for many new entrepreneurs, they're willing to risk time instead of money.

If you get creative, there aren't too many businesses where you can't do a significant amount of testing and validation before spending a lot of money.

A business coach friend of mine (Alan Donegan) talks about doing "mini experiments" to validate your product or service before spending much money to reduce risk:


The material you linked from Alan Donegan was very enlightening. Thank you.

As for the rest... I guess you're right. The more I look at this, the more it seems that you'll be spending almost all your time on a business you're attempting to start, so you can't just try starting several and be as effective. And it seems like - especially after going through Alan Donegan's stuff - if you get creative enough, you don't need much money to try an idea and see if it is viable, either.
 
Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

Post New Topic

Please SEARCH before posting.
Please select the BEST category.

Post new topic

Guest post submissions offered HERE.

Latest Posts

New Topics

Fastlane Insiders

View the forum AD FREE.
Private, unindexed content
Detailed process/execution threads
Ideas needing execution, more!

Join Fastlane Insiders.

Top