March Madness and the Middle East

It’s that time of year again, where everyone who doesn’t pay attention to basketball all season long pretends they’re an expert and fills out a tournament bracket.

I’ll get to some thoughts on this year’s field a little later, but this March is a little madder than usual given events in Iran, so I thought I’d use what we have learned about how to successfully pick basketball upsets to explain Iran’s military strategy.

You may think they have nothing in common, but by the end of the article, you may be surprised how convincing my argument is!

Giant Killers

There is a term in basketball analytics circles for teams that have better than average chances of pulling off an upset – Giant Killers. In case that’s not obvious, it means looking for underdogs that have a playing style that is conducive to upsets.

What does that mean? It means they embrace volatility. It is better to risk losing by 20 to have a chance to win by 2 than to know you’re going to lose by 10.

Or, alternatively, it is preferable to have a 50% chance to win by 2 and 50% to lose by 20, then a 10% to win by 2 and a 90% chance to lose by 10. The goal is to win, so you select the strategy with the best odds of winning.

How does a Giant Killer increase its odds of winning by 2? It does things that are more likely to result in losing by 20! Giant Killers tend to take a lot of three pointers and gamble a lot on defense to attempt to create turnovers.

These are low percentage, high reward strategies. The odds are your team will fail and get blown out. But, on a given night, if the percentages land in your favor, you are capable of a major upset.

To give a simplified example, if the expected score of an underdog vs. a top 10 team is say a 65-75 loss and that assumes the underdog make 10 of 28 three point shots, then if the underdog gets hot and goes 14 for 28, that’s 12 extra points and they pull off the upset – even if the favorite plays a “typical” game.

If that underdog mostly takes easier two point shots, they have a smaller chance at an upset because they don’t have as much upside when they have a hot shooting night. A savvy coach of a Giant Killer would even say, let’s aim for 38 three point attempts in the tournament instead of 28 to increase our variance.

What in the world does this have to do with Iran? Let me explain…

High Variance War

Iran understands it is outmatched and can’t possibly win a long war with the US. Their only chance to win is to bluff in such a way to make the US think it’s stronger than it is so that they will back down.

Thus, we saw their initial response was to fire as many weapons as possible early to project strength. This may have seemed irrational as the US likely understood it was unsustainable and it was unlikely to be enough to work (it didn’t) and would thus leave it easier for the US to defeat them afterwards.

But, from a game theory standpoint, trying to draw out the conflict by holding back missiles meant the only way to win was if Trump got bored and walked away.

By, instead, throwing the kitchen sink in on day one, it created the potential for variance. There was at least the hope that they could score some devastating blows that would have caused backlash both within the US and with our allies.

Thus, while it seemed foolish to attack their Arab allies, it raised the odds of creating political pressure on the US to stop. It was still a negative expected value strategy but Iran only had negative expected value strategies. That’s what it means to be an underdog.

Their choice was to pursue a negative value strategy that had the widest dispersion and hope they got lucky with a favorable “dice roll”. It didn’t work out that way, but the strategy was understandable given the constraints.

Note, Ukraine, by contrast, was able to pursue a more conventional, long game strategy because, while at a disadvantage to the Russian army, the magnitude wasn’t nearly as severe as that between the US and Iran. Ukraine, thus, could try to “win” by buying time and wearing down Russian support at home.

Iran’s first Giant Killer strategy of a “shock and awe” military showing didn’t work, so now they are trying an even more negative expected value tactic – economic destruction.

Again, the aim is similar – cause enough short term pain from higher oil prices to break American will. It’s a longshot strategy and it may eventually work, but one important fact is widely overlooked. Closing the Strait of Hormuz is worse, in the long term, for Iran than the US.

If Iran can’t export its oil, it can’t pay its bills. Their economy was already teetering from the prior US strikes (the protests earlier this year were related to economic pain and Iran choosing to spend on more missiles rather than economic relief).

For now, they can let their own ships through while blocking others, but that isn’t a sustainable approach. It’s arguably easier for the US to block Iranian ships from leaving than to provide convoys for non-Iranian ships.

So the present approach isn’t sustainable and only produces short–term pain for the West. Eventually, it pinches Iran more than the US, and Iran is in no position to endure more economic pain.

As a long term strategy, it’s pretty dumb to fight an economic war when you are the one with the impaired balance sheet. It would be like if Germany 100 years ago decided to start a trade war while it was still trying to pay reparations.

Could they come up with even crazier acts of desperation? Absolutely. I mean if you want a wild card how about an invasion of Dubai to dare the US to send ground troops to defend them?

They are not out of ways to create variance, but each incremental attempt has lower odds of succeeding. Still, they only need to get lucky once to come out ahead.

Bracket Advice

OK, on to filling out your brackets. Long time readers will recall my theory on what matters when filling out a bracket (see here if you missed it). The short version is you need to pick the best team that nobody else is picking.

You also need, like Iran, to take risks. It doesn’t matter if you finish 2nd or 402nd. There are no runner up prizes or huzzahs for “looking smart”. You maximize your chance to win by being willing to look dumb if you end up being the only person in your contest with Purdue as the champion.

However, you don’t want to be crazy different. There is a very simple rule to pick your champion. The winner is nearly always in the top 20 in the country in both offensive and defensive stats. This year only six teams are likely to meet that criteria – the four #1 seeds (Duke, Arizona, Michigan, and Florida) plus Houston and Iowa State.

If you pick one of the four one seeds, you are unlikely to win your pool though because that is who all your friends picked. For reference, here are the most picked teams to win on ESPN.

Now, you could conclude from that chart that you should pick Florida as they are the defending champ and the least chosen of the one seeds. I have no problem with that approach.

You could also say, “hey, Iowa State is under 3%. Nobody in my pool will take them!”. I would consider that reckless (though they are a good strategic choice to make the Final Four). Yes, if Iowa State wins, you will win the pool, but there is likely an easier way.

The next least picked team of the six is Houston. Why? Because they are a #2 seed. But they are on par with all the #1 seeds analytically. People just feel better picking a 1 seed than a 2.

Also, Houston gets to play its third and fourth games at home (nearly every other game is on a neutral court). This is a huge advantage, as it means they would get to play Florida in a home game to make the Final Four.

PS: Houston lost to Florida in last year’s final. So, there is a reason Florida has the worst odds of the one seeds. People see that potential Houston rematch as a large obstacle.

Thus, my championship pick is the winner of the Florida-Houston game. I favor Houston in that game so they are my pick, but I think Florida is a perfectly valid way to go as well.

I do expect a lot of favorites to go far this year, so I think one should choose either Arizona or Michigan to lose to Houston in the final. If you can get the two finalists correct and Houston (or Florida) wins, you will likely win your pool. Normally, you have to be willing to go out further on a limb to win, but this year, I think just not picking Michigan, Arizona, or Duke will be enough.