Cognitive Biases in Poker: The Sunk Cost Fallacy

Flight of the Concorde

One of the most expensive aircrafts ever built, the Concorde was a joint project between the British and French governments. Eight years of studies, three years of development, and four years of construction resulted in the creation of six prototypes by March 1969, but the project had ballooned far beyond its original £70 million budget. Further delays and cost overruns meant that the supersonic jet did not enter service until January 1976, by which time the total programme cost had reached a whopping £1.5 billion.

The Anglo-French joint venture was an ambitious one from the outset but it was beset with problems. The escalating costs and technological issues were not the only challenges. There was clear evidence very early on that Concorde would never be commercially viable. It was fuel-inefficient. The sonic booms that it caused made it only appropriate for trans-oceanic routes. There wasn’t enough of a market for supersonic travel. Nonetheless, both governments continued to fund the project, largely driven by a sense of having to complete the job, given the immense investment that had already been made.

Decision-making by humans is not a purely rational endeavour. We are emotional creatures and, as such, we are vulnerable to a number of cognitive biases. The Sunk Cost Fallacy, also known as the ‘Concorde Fallacy’ is one such bias. After committing time, money or effort into something, it becomes harder to cut your losses. The impact of the loss is outweighed by the perceived value of the gain and there is a feeling that a past investment will be ‘lost’ if an original decision is not doubled down upon. At its core, there is a fundamental failure to understand a new choice to invest in something as separate and severed from a prior commitment to that same thing.

EPT Deauville 2013

Twelve years ago, I played my very first EPT in the beautiful seaside town of Deauville. The home of France’s most eccentric film festival, it is an exclusive getaway destination for the rich and famous, a mock-Tudor DisneyLand for aristocrats and the nouveau riche. It was therefore the most perfect and simultaneously the most ridiculous place to hold a big buy-in poker tournament series.

Back then, I was very assured in my poker skills. I play far better and am a more composed and grizzled player today but, relative to the field, the period between 2010 and 2013 marked the peak of my game. It’s hard to describe the feeling as I walked into the casino. I was keenly aware of the step up in both standard and stakes but I wasn’t nervous. I knew that better players would put me to the test but I had faith in my ability to execute under pressure. I knew that the €5300 price tag was the biggest buy-in I had ever ponied up but I had also spent years successfully detaching myself from the value of money.

Day 1 went well as I mostly small-balled my way to an above average stack. Day 2 was a shorter day and played until close the money bubble. I suffered one small setback early on but, thanks to the terrific structure, I had ample time to recover to an average stack by the close of play. Day 3 started very well, a little card rush and a bad call from Freddie Deeb propelling me to a biggish stack with the bubble looming.

Just Don’t Bubble

Hand for hand took a full level as the pesky short-stacks refused to lose their races and I was joined on the rail by my friend Dara O’Kearney and girlfriend Saron. With each hand taking 10 minutes, I was able to go over and chat to them, receiving inspirational pearls of wisdom like “Don’t do anything mental now!” and “Just don’t bubble!”

I took their advice and successfully navigated my way into the money but a stacked and aggressive table was making life difficult. An entire level of rags saw me slip back to half the average stack by the time we were down to 80 players. It was nice that I was laddering but I was very conscious of being sub-20 big blinds for the first time and figured that I needed to use my pristine image if given the opportunity.

I defended my big blind with K♣️T♣️ and check-jammed the J♣️8♥️4♣️ flop, getting a fold from 9♦️9♥️. I re-shoved A♦️4♦️ from the small blind over a hijack open and got a crying-fold from A♠️J♥️. Then, with 23bbs, I was all-in for the first time in the tournament and got my double-up as my Q♠️Q♥️ held against J♣️J♠️. From there, I nurdled my way to the end of the day with 65% of the average and 54 left. It had been a tough few levels but I was very satisfied with my play.

No More Edge-Passing

Scoffing several pain au chocolats (my body reluctantly processed gluten back then) on the morning of Day 4, I strategised with Dara. We used the vernacular of the day but our conversation revolved around what we now call edge-passing. For 3 days, I had refused marginal spots, taken lots of pot-control lines, played very tight out of position and just generally waited for situations where I believed that my edge was substantial.

Going back with an awkward 27 big blind stack, I signaled my intention to open tight, re-shove selectively but also be on the lookout for good cold-4bet shove spots. I was playing with the big boys and it was now time to make some big boy moves. In the first two orbits, I found three such spots, re-jamming KQ suited twice for 23 and 25 big blinds. I also found a cheeky 27-big-blind cold 4-bet shove with J♣️T♣️ over a hyper-active UTG open and Jason Koon 3-bet.

Just after the hand, Dara appeared on the rail, on a break from his side event, and I gave him a quick summary of the hands I had just played. Just as I was proudly re-telling the cold 4-bet hand, the Pokerstars Blogger Howard Swains caught my eye and asked me if I would do a quick interview at the break. I was happy to oblige and during that chat, he asked me about the bogey hand that had been my undoing in several deep spots in recent UKIPT events.

Djangos

I am not in the least bit superstitious but I have always had a diabolical relationship with pocket Kings. It was especially fiendish in the winter of 2012 when they eliminated me from 10 out of 13 consecutive live tourneys. So, I told Howard I had become part PTSD-sufferer, part Pavlovian dog, developing the association where every time I looked down at the Djangos (as I like to call them), I immediately pictured myself lying on the bed, staring at the ceiling of my hotel room ten minutes later.

Howard then asked me whether I fancied my chances and I confessed that it was starting to dawn on me that, with just 40 players remaining, I actually had a real chance to make the final table in my very first EPT. I acknowledged that I was still below average but noted the distinct lack of big name professional players left.

I also expressed how I was happy at how my early aggression on Day 4 had gotten me back into the 30–35 big blind zone, and with that, armed me with a few new weapons. Hope is a dangerous thing and I probably sounded a bit cocky but it genuinely felt like a real shot. I returned to my table and within the first orbit, looked down at… yes, you guessed it… two Kings.

Djangos Unchained

Doing my best to repress the image of the ceiling of my bedroom in the Royal Barriere, I opened my Djangos in the hijack position and was called in the small blind by the dangerous French Winimax pro Aurelien Guiglini. He check-called my 33% bet on a flop of 3♣️2♦️2♣️ and check-called a 40% bet when the 2♥️ hit the turn. The pot was about 180K and I had 195K behind. I figured his range for pocket pairs, threes through nines, with an outside chance that he was slow-playing/cagily playing 1010+ or stubbornly stationing a hand like A10-AQ. The river brought the 7♠️ and he checked again.

At this point, I took a moment to consider the situation. My perception of his perception of me was that I was tight as we had played the latter levels of Day 3 together when I had been particularly card-dead. My perception of him was that he was capable of folding a small-medium pair if my river-bet appeared too much like a value-extractor. I decided that it was better to go larger and polarise myself.

I landed on a bet of 118K, roughly 65% of the pot and gazed blankly down at the felt, hoping for a call. Twenty seconds elapsed before he announced ‘All-In’. I looked up and stared at him in utter shock. This was the last thing that I expected. I had 78K behind and the pot was almost half a million. Rubbing my face, I began muttering things like, ‘I can’t believe you have sevens,’ and, ‘This is so sick. How the fuck do you have sevens?’

Sunk Cost

I eventually composed myself and started to calculate the odds of the situation. I only had to be good about 14% of the time to justify the call but how often would he shove here without pocket threes, pocket sevens or a slow-played pocket aces? However, I knew that I could completely discount the A♠️2♠️ because when the third deuce came on the turn, the player on the button leapt out of his seat screaming ‘Oh mon Dieu!’

Was it possible that Guiglini was making an ultra-thin value bet with Jacks or Queens? Unlikely but yes. Was it was possible that he was trying an audacious bluff? Very unlikely but yes. In the moment, I decided that I was getting the price for ‘very unlikely’. I stood up and called. He showed his pocket sevens and I tossed my cards face-up in disgust.

Three scoops of ice-cream from Deauville’s finest Gelaterie did little to take the pain away. A 38th place finish in my first ever EPT was obviously a great run and I convinced myself that there was absolutely nothing I could have done about the last hand. I told myself that I got coolered out of the tournament. I chalked it down as yet another bad beat for the Djangos.

Loss Aversion and Unrealistic Optimism

Denial offers a comfort blanket but the idea that it protects us in any way is illusory. In fact, it’s the biggest barrier to growth and improvement. At some point a few years later, I had an epiphany about that hand. As a poker player, one of the cognitive biases that you learn to repress is the sunk cost fallacy. One of the things you learn to resist is that temptation to make a curiosity call. The reason is both reek of loss aversion and unrealistic optimism.

When I bet the river, I was prematurely married to the idea that the pot was mine. When he raised, I did not divorce myself from that notion. The sunk cost was a potent trap that led me to make a poor strategic decision. Just as the bankrollers of Concorde sent good money after bad, I too felt pot-committed and I ended up making a decision based on past costs instead of present and future costs and benefits. Instead of seeing them as two distinct decision points, I framed them together, creating a narrative of failure in my mind if I did not follow through with the call.

In reality, the call was almost never going to be good and I should have realised that. Sure, 8 big blinds would only have given me a slim chance but better slim than none. Rather than make a good lay-down, some irrational part of me wanted to prove to a petulant part of myself that I had gotten horribly unlucky. If I really analyze my mindset, it was also about having other people bear witness to my ill-fortune. If I fold my hand, nobody knows how unlucky I actually got and my only comfort is a very private, self-contained feeling that, on the balance of probabilities, I made the right decision.

The thing is, as a poker player and as a human being, that has to be enough.

David is a professional poker player, strategy writer, podcast producer, and poker brand ambassador. He has written over 600 blogs and articles on poker; including news, opinion and strategy. He has been a professional poker player since 2006 and is the producer and host of the 2-time GPI Global Poker Award winning podcast ‘The Chip Race.’ David has also been a brand ambassador for Unibet since 2017.