Michael Lewis' extraordinary writing career began with two depictions of life on Wall Street. Chapter 1 of his debut memoir chronicled gambling on the trading floor and provided him with the title of his book, Liar's Poker. The second chapter detailed his experiences during the Wall Street interview process. Describing (and often mocking) the personalities and actions of those on Wall Street who conduct interviews is a staple of the genre. From Lewis’ work, to Jim Cramer’s memoirs, to most recently my former colleague Jared Dillian's book, Street Freak, it's clear that what goes on during interviews at an investment bank provides fodder for commentary.
The thinking in some circles is that by creating an uncomfortable environment for would-be traders, the interviewer can assess a candidate’s ability to function under stress. That’s the thinking anyway. The problem with that logic is this: I’m pretty certain I’d be unable to crack the ultra-confident demeanor of say, Mike “The Situation” Sorrentino by asking him to open a window that’s nailed shut. I’m certainly not going to be able to physically intimidate him. Despite that, I don’t want him anywhere near, let alone in control of, any of my money -- right down to my loose change.
I’d rather try to assess a candidate’s capacity for critical reasoning. One of my favorite questions, especially if the candidate has listed poker as an interest on his resume is this: “You are taking a trip with Phil Ivey, widely regarded as the world’s best poker player, and for the duration of the trip you are playing him heads-up for $50,000 in a poker game of your choice. Whoever has the most chips at the end of the trip takes the pot. What is the probability you win $50,000 off of Phil Ivey?”
The correct response is some variation of this: “How long is the trip? Because if we’re taking a boat ride to Europe there is no chance at all. An Acela train ride from NYC to DC? Also remote, but at least my chances of winning are non-zero. If however, it’s a shuttle flight from LaGuardia to Boston, it might be as high as 10%.”
In this context, the baseball playoffs are a subway ride – on an express train – from City Hall to the Upper East Side.
At the end of each forthcoming series preview, and then before each game itself (usually on Twitter) I’ll make a recommendation, based on the stated odds from Vegas, as to which side is the better investment. But consider this: If every playoff series goes the maximum number of games, there will be 41 games of baseball played between now and the end of the World Series. That’s four less games than are played every single weekend during the baseball season. Over an entire season, the results of my logic stand as a proxy for my model. That’s what the book will be about. Here however, judge the logic more than the results and remember the post-season represents a very small sample size.
Or as Billy Beane is so colorfully quoted by Michael Lewis in Moneyball, “My sh** doesn’t work in the playoffs.”
ALDS – Detroit Tigers vs. New York Yankees
In one of the chapters in my book I raise the hypothetical question, how many games would the 2011 Tigers win if Justin Verlander started every game? (Or if you want to rephrase it in a slightly less unrealistic manner: If the Tigers had five starting pitchers with exactly the same skill set as Verlander possesses, how many games would they have won?) There is no “right” answer of course, but there are some unrealistic ones, which include those which simply take the 25 games the Tigers won when Verlander started, and multiply it by 5. (Incredibly, Verlander was credited with the win in 24 of those 25 victories, an unheard of ratio in this era of six inning starters.)
On the other hand, a lot of the answers that overestimate the impact of Verlander’s “value” to the Tigers, end up realistic because of an offsetting assumption. The Tigers offense is vastly underrated among baseball fans and commentators. It’s an offense that scored 787 runs, fourth most in all of baseball. In fact, in 2011 an offense that scores 787 runs would have a hard time finishing below .500 no matter whose pitching staff they hit in front of. I’d put it like this: The 2011 Tigers offense would lead any other team in the major leagues – using the existing pitchers on that team – to an over .500 level except for Houston, Baltimore and Minnesota.
The logic is simple. Those are the only three teams that allowed more than 787 runs in 2011 and teams that score more runs than they allow almost always finish with a winning record. (Cincinnati is the only exception this year.) Whether or not you buy this logic, at least it should dispel the chatter that existed in August that Detroit had under .500 talent when Verlander didn’t pitch. It’s not true; their offense is simply too good
On the surface it looks unfortunate that one of the few teams in baseball that does have a better offense than Detroit is their first round opponent, the New York Yankees. However, the Yankees were a beneficiary of something I call “cluster luck” in the book. The details aren’t for here, but here’s a brief way to illustrate it: The Boston Red Sox had more hits than the Yankees (by 146!), a higher on-base percentage, and a much higher slugging percentage. They finished tied in Isolated Slugging. Despite that, the Red Sox only outscored the Yankees by 8 runs this season. (When regressed, those factors, which largely account for a team’s ultimate runs scored, reveal the difference should have been closer to 50 runs.) So yes, the Yankees offense is better, but it isn’t that much better than Detroit’s.
The biggest advantage the Yankees had over the Tigers over the course of the season – the total performance they got from all the different pitchers who started for them vs. those that took the mound for the Tigers – is entirely negated in a 5-game series format. Using four man staffs for this series, the Tigers will actually have the starting pitching edge in every head-to-head match-up. If the Yankees were to insist on starting Jorge Posada at DH, or if Eric Chavez is forced into the line-up to replace A-Rod due to injury, the Tigers are an enticing underdog for Game 1 and the Series. A Yankees series win is logical with an expected value somewhere north of 50% per my model but it looks a lot closer to a toss-up to me than the oddsmakers are projecting. The call here is Tigers in 5, accompanied with moderate conviction (in terms of value) on the series odds of +130 on Detroit. The Tigers have a better than 43% chance of winning this series.
ALDS – Tampa Bay Rays vs. Texas Rangers
If you look at the final standings and see Texas with 96 wins and Tampa with 91, consider that Tampa played in a much tougher division, and then factor in that ineffable element of momentum, you might come to the conclusion that Tampa and Texas are evenly matched. You’d have a certain amount of logic on your side and, based on the analysts I heard on TV yesterday, you’d also have company.
I think the Rays are in for a rude awakening.
Texas is a 105 win team in 96 win clothing. As mentioned above in the Yankees/Tigers preview, once you strip out the “cluster luck” elements, Texas actually has best offense of any of the eight teams in the post-season. Even that underestimates the talent level they’re bringing to the first round of the playoffs though, because only three of their starters played in more than 135 games. (Much more about this in the preview of the Phillies/Cardinals series.) Because Texas has a fairly weak bench, that’s significant because it means Texas starters are even better than the collective efforts of the players who won 96 wins. Factor it all together and I have Texas as the World Series favorite.
Tampa’s starting pitching staff as a whole ranks in the middle of the pack of the remaining playoff teams, on a par with Texas and the Yankees. However, it’s extremely top heavy; Tampa’s top two starters, James Shields and David Price, make up a higher percentage of their entire staff’s value than the top two of any other team does. That can be an asset in the playoffs (think Curt Schilling and Randy Johnson on the 2001 Diamondbacks) but the drama of this week has changed that equation. David Price, Tampa’s ace by reputation, is not available until Game 3 after starting Wednesday night – in a five game series that will be his only start. James Shields will start Game 2 and then will be available to start a deciding Game 5. That’s important to their chances because according to my model, Shields’ is actually Tampa’s best pitcher. Still it means on the margin, a David Price start needs to be replaced. If it’s taken by Jeremy Hellickson there’s roughly a 10% drop-off in win probability; if Jeff Niemann gets the ball it’s about half that. Neither prohibits Tampa from winning but in a series where they’re an overall underdog, they can little afford to have their one advantage, a top heavy rotation, negated.
You know who else understands the nuances of those shifting probabilities? Tampa Bay manager Joe Maddon. I wrote the paragraph above yesterday afternoon when Niemann was scheduled to start Game 1. Last night, the Rays announced rookie Matt Moore would start Game 1. Matt Moore has thrown just nine innings in the major leagues, and while he looked fantastic, any guess on how he’ll do this afternoon, is simply a projection based on his minor league performance. In fact, at the beginning of the year, Moore, even then a very highly regarded prospect, projected to perform in the majors this year at about the level Jeff Niemann did perform this year. Moore has something Niemann doesn’t have though, and that’s upside. When Jeff Niemann pitches Tampa needs its bats to come alive to win potential 6-5 games. You are not going to win a 2-1 pitcher’s duel, not against Texas at least. There is a very decent chance Moore gets tagged this afternoon too because he has the same fly ball tendencies Niemann has. The difference is that Moore can miss bats, get hot, and strike out a lot more batters. He’s the right choice to maximize the odds of winning, even if the odds of losing in a blowout are the same.
And do you know what that makes Joe Maddon? An options trader. He’s got two out-of-the-money calls to choose from, in the sense that the Rays are underdogs either way. They are the same cost so he’s choosing the one with more volatility, just like an options trader would.
I started writing a book because I thought that the sabermetric advances in baseball had a lot of overlap with model building on Wall Street. It appears Joe Maddon inherently understands this, even if he would never describe himself as “long vol” this afternoon.
Interestingly the odds for Game 1 didn’t move with the pitching change. I think that’s a mistake as it certainly changed my call from a Texas play to a pass. I’m calling for a Rangers sweep so I can’t take Tampa for Game 1 or for the series where Texas is a -170 (63%) favorite, which seems right to me. However, if you are a Tampa fan, you should take comfort in the fact that you’ve got the smarter manager and, on the margin, there is more value now in backing Tampa for Game 1 (or the series) than there was yesterday.