Murthy dropped in last week, the day I’d got a bottle of 12-year-old Irish whiskey. I don’t know how he does it, but he invariably drops in – always without notice, mind you – just when I’ve got a bottle of something special that I want to keep for an occasion. But this time he drank my whiskey with a distracted air, staring out of the window at the grey drizzle falling outside. “What’s up?” I asked.
“I need advice from a statistician,” he said. “You know a lot of professors, and one of them might be good with statistical analysis, besides being able to keep a secret.”
“Sure,” I said, “I know a couple of guys. But why? You’re into wheeling and dealing, not stats.”
He evaded. “It’s just some minor project that came up a few months ago, you know, after the recent state elections that took us all by surprise.”
But I knew it was more than that, and, sure enough, a couple of whiskies later, he got expansive. “Those elections,” he said, “they’re going to change the way elections are managed.”
“What do you mean?” I asked.
He emptied his glass. After I poured him a refill, he continued. “I won’t name any names,” he said, “because all this is speculation. You know, during those last elections up in the east of the country, there was this local state-level party facing a national-level juggernaut. Every analyst predicted that the juggernaut, which had done very poorly thus far in that state, would give the local party a drubbing, or, if not that, at least a great fight. There were lots of people who said that the people hated the local party because it was corrupt and favoured some sections of the populace and so on.
“But when the results came, all that went up in smoke. The local party got a two-thirds majority, the juggernaut got half the seats they’d expected — and a third of what their leaders claimed they’d get — and everyone else got wiped out.”
“Right,” I said. Of course I remembered! “That was when all those MLAs who had deserted the local party for the juggernaut began to queue up to return… I don’t remember seeing such a flood of sleaze ever before.”
“You judge too quickly,” he said. “It’s human nature. You won’t succeed if you sit around judging it instead of understanding it and using it.”
“I’ll do that when I can stomach it,” I said, “but what’s this all about?”
“Well, practically the only one who got it right was a hired advisor to the local party,” he said. “He said that if the juggernaut got more than a hundred seats, he’d give up his career.”
“Yes, I remember,” I said.
“After the results came,” said Murthy, “he quit the company that officially assisted the local party, the company that he founded, despite having called it right.”
“So what?” I asked.
“Another state government that will be facing the juggernaut soon, and has dissent within its ranks, has hired him as an individual at an undisclosed fee… That fee must be stratospheric. So he has all the power and the money he wants but is accountable to no one. Brilliant!”
“So why are you looking for a good statistician?” I asked.
“Because I figured out what he did,” he replied. “He figured out a way to beat all the pollsters and election-watchers. He’s changed the way elections are managed. Permanently.”
“You still haven’t told me how,” I complained.
“Ah, yes,” he said, in his best superior manner. “Whenever there’s an election, parties and candidates… Well, they offer voters money and booze to vote for them.”
“Yes,” I replied. “Part of the sleaze.”
“Well, I think this guy took it a step further,” he said. “He offered his flock the usual money to vote for his party, and a little extra to tell everyone that they were going to vote for the juggernaut.”
“But why?” I asked. I’m a simpleton in these matters.
“Two things,” he said. “It gave the juggernaut’s people the wrong idea about how well they were progressing. Second, the poll analysts, too, thought the juggernaut would do better, and got thrown off.”
“But that still doesn’t explain why you need a statistician!” I said.
“I thought you’d understand,” he said, disappointment in his voice. “This guy created a lie that’s essentially a statistical bias, so I want to know if there’s a statistical method to understand and neutralise it. The other thing is, the local party had limited cash. So the guy had to figure out a way to spend his money in critical constituencies in areas where it would have the most effect. There might be a mathematical way to do that.”
“Well,” I said, reaching for my phone, “I’ll put you on to a couple of profs right away. You can speak to them directly.”
“No thanks,” he said, emptying his glass again. When I reached for the bottle, he waved it away. “No thanks for that, too. I’ve just figured out a possible solution.”
“But you don’t know any mathematics,” I said.
“I don’t,” he said. “But this is politics and people, not mathematics. This guy we were talking about, he told people to lie: they did one thing and said another. Like double agents, sort of.”
“Yes,” I said.
“If you want to beat him and his double agents,” he said, rising to leave, “you pay some of those double agents a lot of money to change who they vote for, but not what they say. You pay them to tell the truth about who they’re voting for. You spread the rumour that you’re willing to pay other double agents as well. It won’t cost much, since you’re working with rumours and confusion. So you create not double but triple agents, see?”...