Scirra cog

About Us

We're a London based startup that develops Construct 2, software that lets you make your own computer games!


Browse all our blog posts

Latest Blog Entries

We love brains!

Join us! Joiiinnn ussss! Mooooree brains!

Why Running a Startup is Like Playing Poker

by Tom | 24th, April 2012

I used to play a lot of poker (and occasionally still do for fun). Like any game to get good at it you have to study a lot unless you are fortunate enough to be born with a different psychology that hugely favours specific games or skills (I was not, Tom Dwan was). Poker for me was immersive, exciting and challenging.

Poker is also a very misunderstood game, much to the frustration of the wider poker playing community. On a small Scottish island I once by chance met someone who was involved with legislation of online gambling in the UK. They were unfailing in their conviction that poker was an entirely luck based game. How wrong they are! It's a shame that policy is partially dictated by the misinformed, no wonder the poker community feels hard done by of recent.

After several years of playing and studying it became obvious that it was unlikely to ever go further than a hobby for me. During the learning process of attempting to get good at the game I learnt a lot of useful skills and life lessons. Now I'm involved in running a startup I've discovered that it draws a lot of interesting parallels with the game.

The Grind

If you're serious about playing poker you will know that the glamour and excitement of the game can wear thin. If you've had a bad day and you're trying to progress you still need to come back the next day and play again. This is known as the grind.

A successful grinder

Startups can be the same - sometimes. Casual observers might think you've got it easy but the reality is you need to sit down and work hard day in, day out. Grind is probably too strong of a word for running a startup you love but it's correct in the sense you have to be prepared to work hard and have the right mentality if you want to progress.

I'm lucky as our startup had traction when I joined primarily from Ashley and co.'s previous hard work on the open sourced Construct Classic. However my short month or two stint of working a full time job and coming home and working on Scirra in the evening was definitely a grind. Startups that have little traction - or where the founders are working full time jobs simultaneously - could probably quickly find themselves in a grind.

Expected Value

Good poker players are very familiar with the term EV or 'Expected Value'. It's the mathematical fundamental that winning players should be using for every decision they make. An assumption a lot of non-poker players make is that you need to be a maths genius to become good at poker. This isn't strictly true, being able to calculate EV is a relatively simple calculation and is by far the most important skill.

To calculate your expected value you simply add the results of all the possible outcomes multiplied by their respective probabilities. Let's say someone offers you a bet on a coin flip. It costs $10 to play, and if it's heads you lose your bet and if it's tails you win $10.25.

Assuming a fair coin the possible outcomes are:

  • Heads you lose $10
  • Tails you win $10.25

Each outcome has a 50% chance of happening:

EV = (-$10*0.5) + ($10.25*0.5) = $0.125

If you play this game 1000 times you can expect to win $125. If you don't value your time that highly you might happily sit there all day playing!

Now let's assume our coin flipper is smart. It might be savvy to assume this and that you're missing out on some important information, why else would he be offering losing bets? You notice whichever side of the coin you choose for the bet he starts the coin on the other side before flipping. This gives the coin a 51% chance of landing on the same face (a fascinating read!) Would you now want to take this bet?

EV = (-$10*0.51) + ($12*0.49) = -$0.0775

Play 1,000 times and you're going to lose $77.50! Still, a pretty laborious and slow way to lose money. The point is that EV calculations are a helpful tool, but don't always refer you to the truth. The more information you have available and the more skilled you are at applying the information in a useful manner into your expected value calculations the more your estimate will tend towards the truth. That's where the real skill in poker is.

It is often advantageous for startups to look at decision making with the EV shades on. A simple example would be investing $100 in an advertisement that will yield $90 in sales giving us an expected value of -$10 (okay, we're not summating values of outcomes like in the previous example but the principle still stands). Purely in a mathematical sense this is a bad investment.

Again however when you start to think of EV in a more intelligent way you might decide that the brand recognition and referred sales from the $100 advertisement is worth more than $10. This now means that you believe it is +EV to run this advert at an apparent immediate loss.

Calculating implied value can be extremely difficult as it's often near impossible to measure. If the item we are selling costs $10, and we know that there is a 15% chance that each buyer will refer another customer, with 9 sales we can expect to yield on average $13.50 in referred sales:

9*($10*0.15) = $13.50

So our advert that costs $100 and yields $90 in sales we have now calculated has an expected value of +$3.50 if we also account for implied value.

This is a simplified example, it's probably almost always ill-advised to keep investing in an advert with a negative immediate return! There are likely better opportunities and ways to spend your money. The advantage of the Internet though is that it can sometimes be easy to scale campaigns like these.

Sometimes however you will want to turn down positive EV decisions, this is because of risk of ruin. You might decide that running an advert during the Super Bowl has an expected value of +$500 but it's going to cost you $1,000,000 to run it (all the cash you have) you would be ill advised to take this opportunity. To reduce your risk of ruin you need good money management with a realistic risk taking attitude. If you have $10,000 available you might only want to expose a few hundred dollars a time to significant risk.

There are lots of decisions that can be reinforced by evaluating them with expected value, it's a useful tool startups should be using when making decisions — especially those with a lot of quantifiable data behind them.

Measuring expected value in poker is often made easier by the fact a lot of the data is quantifiable. Utilising qualitative data in expected value calculations can often be a lot more difficult. Poker ultimately is a fairly simple contained environment (although its complexity reaches great depths). The entire world on the other hand is far more complex environment to exist in.

You're To Blame for Your Mistakes

As just explained expected value calculations are a crucial tool for the winning poker player. If you make a mistake and lose a big pot you only have yourself to blame. It's no one else's fault. Your estimates and calculations are off. If your opponent did a totally unexpected bluff you didn't pick up on it's your fault for not recognising they are a player that is capable of making those bluffs on you.

No one to blame except yourself

Running a startup is the same. If a mistake is made you are the one at fault - this is why running your own startup will teach you a lot in a very short space of time. Making mistakes is normal and an important learning process. In both poker and startups you pay for mistakes, so why not use them to your advantage?

There's a lot of Losers

Estimates of how many players are long term losers in the game do vary a lot and 5% gets thrown around a lot. SharkScope is a well-known poker statistics website and they have different numbers backed by a large sample size which says on average 26% of players are winners, and 74% are losers. (Worth noting however is that this is for tournaments only and not cash games. Cash game numbers could be very different.)

Time to try a different strategy perhaps?

It's incredibly difficult to estimate how many startups fail but of course they do. According the US Small Business Administration "Seven out of 10 new employer firms survive at least 2 years, half at least 5 years, a third at least 10 years, and a quarter stay in business 15 years or more" (source).

This isn't the entire picture however as a lot of startups aren't "employer firms" (they often don't have employees): "nonemployer firms" are far more numerous. Also worth noting is that the SBA is only measuring US based businesses, and that sector to sector and country to country you might find widely varying rates of success and failure.

The fact you might fail doesn't matter. No one when they start out thinks they will fail but there comes a point in poker where you have to realise "this isn't going anywhere". The same is crucially important to startups. If after you've done everything you can do it's failing you will save yourself a lot of pain if you cut your losses. "Never give up" is a romantic notion but sometimes you just have to.

It can be an Emotional Rollercoaster

Playing emotional poker is often a costly mistake.

All it takes is one bad day

Your emotions are powerful and you probably underestimate them. Poker players have a term for this — "going on tilt". Tilt is when you lose control of your decision making process and start playing emotionally, often caused by frustration or some perceived injustice. This leads to a long string of very negative expected value decisions. Strangely enough tilt occurs in many other aspects of life yet Poker players seem to be the only ones who recognise it's extremely dangerous potential and label it accordingly.

Everyone tilts, it's a natural part of human psychology. Everyone also has different thresholds of what will cause them to tilt and how they react as a consequence to it. Even if you recognise you are tilting it's often extremely difficult to break your behaviour pattern which is why it should never be underestimated. The effects of tilt can often be leveraged in specific environments, such as casinos where alcohol is freely available.

When running Scirra one of the unexpected adversities I had to overcome was negative criticism. It's far too easy to respond to these emotionally and as a consequence start to make emotional based decisions. This is a dire mistake that can have incredibly damaging repercussions, just like in Poker. All it takes is one Tweet or email and your world can come crashing down around you as tens of thousands of people on the Internet mobilise to express their seething hatred of you. Thankfully this isn't something we have ever experience or plan to experience (who does?) but we've seen it happen to others. When you're in the public eye it is advisable to remind yourself you are also operating in a leveraged environment.

Very Few people become overnight successes

Every now and then a poker player who is clearly outclassed by the field will win a huge tournament (or 'donkament' as they are affectionately referred to by poker players). This is simply an infrequent and rare chance event. Amateurs spectating the tournament will often romanticise it and truly think they can also win, even though the odds are overwhelmingly stacked against them. "This guy did it, so why can't I?" It's why people play the lottery after all! Aspiring to replicate an incredibly rare chance event isn't the best plan of action.

Overnight successes are rare but spectacular

The same applies to startups as well, Instagram being the hot topic everyone is talking about right now. Launched in October 2010 around 18 months later they sell to Facebook for a billion dollars.

The big wins are rare, impressive and attractive. Yet these legends are often perpetuated by the inexperienced and can lead to unrealistic expectations and goals. Ambition is a virtue, but aspiring to the most extreme events will only lead to disappointment and possibly bad decision making.

You Look at Money Differently

New players to poker are often overly passive as they feel uncomfortable with the risk involved of betting with their hard earned money that is in neat little piles in front of them. Thinking of the chips as your cash is a mistake if you want to win.

As a poker player you need to escape from this natural way of thinking about your money and treat it for what it is — a tool.

Dealing with losses can be difficult but it's important to learn to cope with them. In your startup you might run a fairly expensive advertising campaign that fails dismally. The important thing here is to evaluate what went wrong and try to learn from it, simply putting your head in your hands and thinking about what the money could have been spent on is a useless endeavour that will only lead to misery. As a consequence it could also encourage you to be overly passive in future decision making — a potentially damaging strategy.

It's Risky

Some people are naturally risk averse. Some people just don't like risk and understandably so. We didn't evolve jumping over huge bottomless pits to shave 3 minutes off the time it takes to walk there.

Playing poker should probably be less risky than running your own startup if proper money management is exercised, you're making tens of thousands of bets that are (comparative to your bankroll) small. As long as enough of your bets are +EV you will make money in the long term.

Running a startup can be similar but often entails far more risk. Startup founders often have to sacrifice good jobs and lifestyle. Most of all they often sacrifice well paid and very safe opportunities in the future. In both pursuits risk is an opportunity and should be embraced. Recklessness is not savvy risk taking, but if you take calculated risks they often come with good rewards that make the whole thing worthwhile.

It can be the most fun you've ever had

Everything has its ups and downs. Running a startup is no exception. However when you wake up in the morning and are excited by the day ahead that's when you really have something of value in your life.

Poker players and startup founders have a lot in common, mainly that they are people who embrace and understand risk intimately. For me a life without risk would certainly be a dull one! Not quite being able to see round the corner is exciting and often rewarding in many ways, and Scirra continues to be an adventure I feel lucky to be part of.

Now follow us and share this



AJTilley 5,056 rep

Nice post.

I agree with a lot of what you have said, especially the opening about the grind... I am not building a startup but games (as are many i imagine) And I find that day to day I have to grind away building the game. When I first thought about making games I thought it would be fun and challenging (and it is) but it is also repetitive and frustrating.

Tuesday, April 24, 2012 at 2:29:28 PM
Basseman 3,494 rep

Hmm.. Thanks i guess ;)

Tuesday, April 24, 2012 at 2:30:46 PM
newt 84.5k rep

Its all about the numbers. You just have to know the odds.

Tuesday, April 24, 2012 at 2:34:53 PM
Jakub 6,423 rep

I think that actually that's the best blog post so far!
I'm using alot of poker references in eSports and I think that it's a good tool to teach someone about taking risk and separating emotions from statistics.
Generally speaking people have huge problems thinking using statistics because emotions ruin everything.
Life is all about maximizing your odds.
And when it comes to startups I started with freerolls now I'm just finishing the pure grind phase and starting to play some more big tournaments. ;)

Tuesday, April 24, 2012 at 2:43:05 PM
Tom 48.7k rep

@newt to start yes, but it gets a lot more nuanced when you play better people!

Tuesday, April 24, 2012 at 2:43:37 PM
CyberDagger 2,638 rep

When you talked about calculating EVs, I thought of something else entirely.

Tuesday, April 24, 2012 at 3:06:05 PM
Trevor10 4,161 rep

I would never trust programmers enough to gamble online.

Tuesday, April 24, 2012 at 3:17:15 PM
icemelt7 2,546 rep

Beautifully put! Thank you!

Tuesday, April 24, 2012 at 3:36:59 PM
ludodesign 29.8k rep

Oh!.. really interesting. Good post.

Tuesday, April 24, 2012 at 7:22:51 PM
marcgram 10.8k rep

Interesting slant.

I am betting Scirra is going to be in the winning 26%.

Tuesday, April 24, 2012 at 7:33:47 PM
Kyatric 67.1k rep

I call your bluff. All in !

'Durrr' ^^

Tuesday, April 24, 2012 at 7:57:36 PM
Geo 5,675 rep

Thanks Tom, very insightful. Hope I'll need this knowledge soon :).

Heck, now I'm all excited and can't wait for "the day". OK back to the grind now :p (the grind before the grind, that is...)

Tuesday, April 24, 2012 at 11:11:22 PM
GenkiGenga 11.2k rep

"Your emotions are powerful and you probably underestimate them". Wise words Tom.

For what its worth, right now, I think Scirra has a good hand.

Wednesday, April 25, 2012 at 1:36:54 AM
freeg131 2,947 rep

I have to say I am continually impressed by the quality and range of posts on the Scirra blog and the insights you give us into startup life!

Wednesday, April 25, 2012 at 1:44:32 PM
programaths 4,353 rep

@Trevor10, unless this is a "programatician" :-D
Shuffling by swapping may lead you to initial configuration. These are cycles where entropy increase and decrease :
If you calculate the number of swaps needed to go from one configuration to another (kind of hamming distance), you will notice that the number of ideal shuffles is linked to the number of items.
So, you've to find the right algorithm and give him the right parameters. None of these tasks are easy and this is not a developer job ;-)
In a simple "choose two and swap them" algorithm, this number seems to be n/2 but that's probably wrong.

Also humans are quite unable to create random sequences. Their output is predictable.
⇒ 1011010110101001
⇒ 1110101101111000
The first is what most would output, the second is a "natural" result.

Wednesday, April 25, 2012 at 9:13:08 PM

Leave a comment

Everyone is welcome to leave their thoughts! Register a new account or login.