For the IPL auction tomorrow, this is a post with me nerding out about cricket auctions and cricket players, and assumes familiarity with cricket and the IPL. For any non-cricket readers, normal programming will resume (…at some point? I’m sorry!)
In a few short hours, a bunch of smart, successful people will gather in a room in Kochi to buy the services of players for IPL 2023. The IPL’s choice of having an auction instead of a draft serves important functions, like its obvious display of fairness in the market. But by far the most important one is the spectacle and the glamour of the thing. The glamorous, breathless TV coverage, the values of rupees being thrown around, the records being broken for most expensive buy every year, are supposed to get you talking about them.
But it’s not just that: IPL auctions are fun. It’s underrated to an extent, but auctions are almost an entirely different sport from the cricket they help put on the field. Putting together a good squad of players in an auction is a problem that needs good knowledge of cricket for sure, but it’s not enough. IPL auctions generate their own share of bad takes (and good ones), but I think a better understanding of auction dynamics can often make them even more fun to watch. As someone who has followed them avidly for a long time (and attended a couple!), I think it’s important to know these things to get a better understanding while watching the auction:
1. Determinism is stupid (and boring)
Auctions are inherently probabilistic events. Anyone who tells you Sam Curran will be sold for 18 Cr tomorrow is wrong. So is anyone who tells you it will rain at 10:15 in the morning. Both for similar reasons: weather systems and auctions are both dynamic events that change as they go. Budgets shrink and expand. Players rise and fall in priority. Hell, if you asked Akash Ambani how much he would bid for a player before he sat at the table, his answer would probably not match what he does at the table.
A better way to think about this is in terms of probabilities. An auction is less like a draft where you decide which players you want and more like a game of poker where you’re trying to play both for yourself and against other players. Especially in a small auction, buying one overpriced player can be the difference between buying one player and two players, and even cost you a good season. Being a good death bowler - a valued commodity - means your likely price is higher. Being a good captain makes you more valuable to specific teams who may be looking for one. But the outcome is a range, not a value. If you miraculously turned back time and conducted the same auction again, I’m pretty confident the outcomes will be different each time.
2.Price vs Value: Why does Ben Stokes go for this much money? Why does Adil Rashid go unsold?
What follows from this is that a player’s price is only marginally reflective of skill or ability or form. Very often, especially in small auctions, there are only a few players at the level of ability you want with the skillset that you want, and very often that means the price is simply a reflection of how the supply and demand is stacked in the market. To use my favourite example: In the 2018 (mega) auction, Mumbai Indians bid 6 Cr for Dinesh Karthik, 4.8 Cr for Wriddhiman Saha, 6.4 Cr for Robin Uthappa, and 7.8 Cr for Sanju Samson before they went and bought the uncapped Ishan Kishan the next day for 6.2 Cr. If you were a man roaming around Jaipur that day with wicketkeeper’s gloves on, there was a fair chance Akash Ambani would have tried to give you money. That wasn’t necessarily a reflection of performance - though these are all good players! - wicketkeepers were just suddenly valuable in a market with a buyer desperate for one specific thing. Someone going for high money is not an indication of them being fantastic. Someone going unsold is not an indication of them being a bad player. It’s the market.
Let me take one specific example: foreign spinners, however good they are, often struggle to get picked in an IPL auction. This is a very particular dynamic: an IPL team has only four foreign slots available in any match. As has been shown in the last few years of the IPL, even relatively no-name Indian spinners are often pretty good (as shown by the pipeline of them: Kuldeep, Markande, Rahul Chahar, Bishnoi…). The difference between <young Indian wristspinner> and Adil Rashid is often not big enough to warrant spending that foreign slot. The difference between Basil Thampi and Nathan Coulter-Nile often is. The difference between Shivam Dube and Ben Stokes often is. If the spinner is a mystery spinner or can bat, this calculus can shift towards them - Rashid and Narine are prime examples - but Adam Zampa being unsold is not a reflection of how good Adam Zampa is. It’s a reflection of the fact that foreign slots are used well on what Indian players are not good at: both batting well and bowling half-decent seam-up.
3. Method in the madness
There are a few things that seem like idiotic choices when on the outside of the auction looking in which have an internal logic, even if you don’t agree with it. The list of players re-called at the end is often a prime example of this: I’ve always wondered why those players go unsold if teams have specifically requested for them. Well, some of them are chosen for other teams to hopefully bid on, so you can get who you want. Some of them are chosen for your own consideration after filling your most important slot. Some of them are just.. random. There to fill out the list. Like all of us.
Teams often have specific slots they’re looking to fill, with an idea of which ones are more important (and which players are more important). This brings an element of game theory into the equation. It might mean letting some players go for cheap (if you really need an all-rounder and have one foreign slot, and Cam Green comes before Ben Stokes - and you’d really prefer Ben Stokes - do you bid?) If a later player is more important to you, you could let someone go unsold with the intention of buying them later if that means saving up money for the more important of the two players.
Teams also often have access to information that may or may not be publicly available, like injury or availability concerns to their main fast bowlers making them do seemingly dumb things like buy up even more fast bowlers. Or a bias for a particular player, or a coach who absolutely does not like a certain player, or the thousand other things groups of humans have.
4. Madness in the madness
There is an internal logic to the game of auctions, but that doesn’t mean they are some kind of rational event played by rational actors. In my opinion they’re prone to the same kinds of fun irregularities that any markets are: bidders are people and can get carried away or bid emotionally, it’s very easy to get fixated and lose the plot, and hype is often a very real thing.
Answer me this: how much has materially changed about Jhye Richardson from 2021? How much has materially changed about Kyle Jamieson? How much has materially changed about Odean Smith since last year? If their prices significantly differ from 14 Cr, 15 Cr, and 6 Cr this year, despite teams still needing bowling all-rounders who bowl pace, will you be surprised? What has changed, IMO, is that these players are no longer thought of as unsubstitutable by other players to the extent that they were.
The point is that perceptions are often collective, and they often lead to groupthink in the room: however smart you think you are, when everyone is buying tulips, you just kind of feel like buying tulips. If you think one player will fit your team like a jigsaw and be a complete game-changer, and you think that player is not substitutable with any other player in the market, and you think your team doesn’t need anything else, you can convince yourself the best move is to spend as much money as necessary to get that player. That’s a human thing.
So… what will happen in the auction tomorrow?
Coverage of the IPL auction mostly tends to be in the style of “5 players RCB could bid for” or “CSK source indicates they could look for THIS player” or “These 3 players could be the most expensive in this year’s auction!!!!”. And while that is… cool for everyone who enjoys it, I think there’s a better framework to talk about it.
The truth is, some things that happen in the auction are fairly likely events that you’d expect to happen, and some things are batshit insane, unlikely events that you wouldn’t expect to be happening. There’s an element of subjectivity to deciding which is which, but it’s subjectivity you can put some numbers to.
In the last couple of days, my friend Omkar and I threw together a quick Monte Carlo simulation model to simulate auctions with different prices for players. We thought it would be a fun exercise to put those numbers out there, and see how they turn out after the auction. So our numbers for likely prices for the first 30-odd players can be found here, in a public Google sheet. This is a rough, work-in-progress version that we would’ve loved to work on if we had a few more days, but whatever, this is the current numbers.
Here’s how to read this sheet: in an ideal universe, the simulations of our model reflect potential real-world outcomes with their actual likelihood. So, if a player is currently being sold for what is higher than our 75th percentile price, we would expect that to not happen very often (not more than 25% of the time). And if we’re saying there’s a 30% chance of a player being unsold, we would expect such players to go unsold 30% of the time. That’s a pretty tough gig, and I’m guessing we’ll be pretty far off a few times, but this is an encapsulation of our beliefs in a model.
What we’re really hoping to do with this is to have a conversation about the auction in a slightly different frame of reference from the usual. If you think this is dumb, because <your favourite player> is definitely going to be sold for more, then think about why that could be. If you think a number looks too high or too low there, we’d love to talk about why.
And of course, if you were at the auction table, you could use this differently: you could use it to know when you’re participating in an unlikely event, and when things are going business-as-usual. And you would know the exact implications of both for your own strategy as the auction happened.
Please reach out to Omkar or me if you have any comments, because we love talking about this baby. I hope watching the IPL auction - or as I call it, Reverend Bayes’ Fun-tastic Poker Orgy - is more fun for you after reading this. And I hope that you stay toasty and chill at the same time.