What the IPL can learn from telecom
In the recently concluded Indian Premier League (IPL) auction, 171 players were sold for a total salary of Rs438.5 crore, an average of Rs2.6 crore per player. The top 10% of the players earned Rs151.8 crore, 34% of the total price paid. The bottom 10% of the players earned Rs3.4 crore, only 0.7% of the total price paid. Do the prices genuinely reflect the differences in market value of the different players?
The IPL auction is a multi-unit auction where bidders bid for multiple units of an object, for example spectrum blocks in different telecom circles or, in this case, players. According to the current format, in the auction for a player, bidders indicate their willingness to continue or step down as an auctioneer raises the price in fixed increments, till only one bidder remains. This is called a Japanese clock auction.
The 578 players were divided into 62 lots, which were auctioned sequentially with marquee players being auctioned first, and players sharing the same specialization being clubbed in the same lot and auctioned later. Internationally capped players were auctioned before others. Within each lot also, the players were auctioned one after the other.
Multi-unit auctions are distinct from standard single object auctions because of the linkages—complementarities and substitutability—between the objects. For instance, the value of acquiring two specialist openers and a one-down batsman may be greater than the value of acquiring three “good” batsmen. In such auctions, sequential sale limits information available to bidders and the manner in which they can respond to information. Bidders must guess what prices will be in the future auctions when determining bids in the current auction. A bidder cannot switch back to an earlier item if prices go too high in a later auction. Bidders may regret having purchased early at high prices, or not having purchased early at low prices. The guesswork about future auction outcomes makes strategies in sequential auctions complex, and the outcomes less efficient.
The Swiss wireless-local-loop auction conducted in March 2000 illustrates the difficulties of sequential sale. Three nationwide licences were sold in a sequence of Japanese auctions. The first licence sold for 121 million francs, the second for 134 million francs, and the third, double the size of the first two blocks, sold for only 55 million francs. This example illustrates that the much higher prices earned by players earlier in the IPL auction may not just be due to their status as “marquee” or international players but could also be influenced by the particular choice of auction method.
In 1993, to address the problems of sequential auctions, game-theorists formulated a novel auction format to auction spectrum licences in the US. They chose to auction all licences simultaneously in a format called the simultaneous multiple round ascending auction (SMRA). In the SMRA, bidders simultaneously submit bids in each auction, round by round, until a round is reached in which no new bids are received in any auction. This gives bidders flexibility to target each unit being auctioned (spectrum blocks or players) based on the progress of bidding in all the individual auctions. If bidding is in progress in even one remaining region with all the other regions closed, and if it goes very high, then some bidders may want to reopen bidding in other regions which are substitutes for the last remaining region.
This auction format has become the workhorse for spectrum auctions the world over, including in India. Specific rules have been incorporated in the SMRA to avoid collusion, a winner’s curse, or even the strategy of “sniping” where a bidder waits until the last moment to bid, thus preventing market information from emerging in an orderly manner.
While the simultaneous auction addresses the issue of complementarity better than the sequential auction, it may be inadequate in case the synergies are very high. A bidder may fail to acquire key pieces of the desired combination, but pay prices expecting the synergistic gain. In this case, game theorists have designed a package or combinatorial auction in which bidders bid for combinations of units rather than for individual units. However, this auction is highly complicated and has not gained much currency.
For an efficient process, players must choose reserve prices carefully—setting it higher than one’s own estimation of one’s value can change the expectations of the bidders and inflate the price of a player, but over-optimistic reserve prices can result in perfectly good players remaining unsold.
While the SMRA with conservative reserve prices promises much in terms of reducing inefficiencies, it needs to be tweaked before applying it to the IPL auction as there are over 500 players in the fray, far too many for a sensible simultaneous process. Therefore, one should divide the players into 10-12 categories (certainly not 62!) with the categorization aimed at minimizing synergies within the lot of players who are being auctioned together. For instance, one could auction opening batsmen, batsmen, fast bowlers, spinners, all-rounders and wicket-keepers separately, with internationally capped players being auctioned before national players. While the lots should be auctioned sequentially, with the order being chosen randomly, within each lot players should be auctioned using an SMRA.
The revised format would bring about greater efficiency in the process of determining the price of players. And of course, it would improve the quality of cricket.
With inputs from Mohit Gaur, student at MDI, Gurgaon.
Rohit Prasad is a professor at MDI, Gurgaon, and author of Blood Red River. Game Sutra is a fortnightly column based on game theory.
Comments are welcome at firstname.lastname@example.org