September 3, 2003
Lies, Damned Lies
Moneymaker (or, Everything I Need to Know about Baseball I Learned From Watching the World Series of Poker)As I sat in the upper deck at Jacobs Field last Saturday, taking in the Indians-Blue Jays tilt and shivering in the Lake Erie breeze with our Cleveland Pizza Feeders, the conversation turned to Texas Hold 'Em. Poker is a natural fit for baseball fans, especially the sort that are likely to attend our events. Like baseball (or at least the 'game' of baseball management), poker is a game that's grounded in mathematics, and in optimizing the use of limited information. Like baseball, it's also a lot of fun--at least when you're winning.
Just a bit of background here, which will be unnecessary for some readers and inadequate for others (if you've never played poker at all, this probably isn't your column). Texas Hold 'Em is a variant of poker in which each player is dealt two 'hole' cards face down, and makes the best five-card hand he can between his own cards and five common cards that are dealt to the entire table. The 'face down' part is the key: a player's hole cards are never revealed until the last round of betting has been completed. In fact, in a tight game, the hands are often not revealed at all--every player but one will have folded before the showdown occurs.
I've always found that last bit fascinating: players are willing to risk (sometimes large) sums of money on hands that they're never able to see. While a good player can pick up plenty of information between observing the table's betting patterns, running and rerunning the odds of particular hands occurring, and observing the other players' "tells," there's always the lingering possibility of a bluff, which as a game theorist can tell you, will occur just often enough to keep a bettor on his toes.
Lest you think this is a Bill Simmons-style off topic diversion, there are lessons that can be drawn from Hold 'Em and applied to baseball. Let's take a break from the usual dose of number crunching and look at those this week.
Lesson #1: Don't Go Chasing Against Unfavorable Odds
'Chasing' is the practice of continuing to call bets in the hopes of catching cards that are unlikely to come. Say, for example, you've got a pair of twos with just one more common card to be revealed--the 'river' in poker parlance--and would need another two to win the hand. The odds of that happening are better than 20-to-1 against; not a good bet. While there are high-reward situations in which paying for an extra card is warranted, over the long run, chasing is a surefire way to lose a bunch of money.
You'll see baseball teams chasing all the time. The Pirates will sign a couple of mid-level free agents to make their remote chances to win the Wild Card slightly less remote, then cite financial difficulties when the playoff run doesn't materialize, and trade their best players away. The Devil Rays will put together an entire roster full of washed-up vets, delaying their growth process by years. Knowing when to fold 'em is the toughest thing in baseball and poker alike.
Lesson #2: Beware Good Hands Gone Bad
A particularly troublesome variant of 'chasing'--something that even good players are prone to do--occurs when hands that are good before the flop (the 'flop' is the first three common cards, which are revealed simultaneously), such as Ace-King of the same suit, go unimproved and are surpassed by other hands at the table. It may be perfectly obvious that your hand is going to be beat; the other players know it too, which is why they're making bets in your direction. But stubbornly, you'll call their bets--maybe even raising, if you're really aggressive--in the hopes of catching a miracle card or uncovering an unlikely bluff.
The Mets have long been the poster children for this sort of behavior. They were slow to recognize when their great teams of 1986-1988 went down in a cocaine-induced haze; thus were begat the Vince Coleman Years. They were equally slow to recognize that their near miss in the 2000 Subway Series was less a harbinger of future glory than the last desperate gasp from an old team playing above its level of ability. Even their last place finish of a year ago didn't deter the Mets from trying to make a big splash in the free agent market by overpaying for Tom Glavine. You keep betting a weak hand, and the hole grows deeper.
Lesson #3: Bet When You Have The Best Hand
There's a flip side to this, too: sometimes, you get exactly the flop that you're looking for, and an average hand suddenly becomes much better than that. While there's a tendency among some players to 'slowplay' their hands--deliberately underbetting their cards in an effort induce more bets from their opponents--there's danger in that approach, too: if the other players can see another card cheaply, they may well catch exactly the card that they need to complete their hand. Just when you're preparing to make that big bet on the river, you'll find that your hand has already been beaten.
This is a lesson that the A's understand well. Given the team's permanently precarious financial position, its soon-to-expire contracts, and the amount of talent already accumulated on the roster, Billy Beane isn't afraid to make a move that will push the A's over the top, even if it means trading away a prospect or three. For a team that has a low revenue base, opportunities to reach the postseason are few, and it's critical to take advantage of them before the window has been shut.
Lesson #4: Don't Bluff Against Players That Are Going To Call It
An equally important skill to reading one's hands right is reading the other players at the table: poker players run from risk-averse softies who will fold at the first side of resistance, to stubborn 'calling stations' that will see your bet even when they have no chance of winning. Against the former group, a well-timed bluff may be highly profitable; against the latter, it's money thrown away.
A particularly bad player to bluff against is someone who you've burned before. Brewers fans have heard for years that success is just around the corner, only to discover that the rebuilding will be followed by...more rebuilding. A new stadium is built, requiring $400 million plus in taxpayer money, to dreams of autumnal tailgating and sausage racing. But without a change in organizational philosophy, without discernable improvement on the field, the fans, once-fooled, have stayed away. To his credit, Doug Melvin has the beginnings of a decent farm system on his hands, but there's little to be done to finesse the indifference that Milwaukee fans are feeling.
Lesson #5: Luck Is For Losers
This is the most important lesson of all. Players new to Texas Hold' Em are sure to confuse it with a game of luck, comparing it to other casino games like blackjack, roulette, or--blasphemy!--slot machines. Certainly, in any given hand, there's a significant amount of luck involved. Even the best starting hand (a pair of aces) has only about a 30% chance to emerge victorious on a full table of 10 players.
But in the long run, a skilled player will win money, and plenty of it. He'll win money, in fact, despite the casino's 'rake' of as much as 10% of the pot. And that's really the key to a lot of the analysis we conduct on baseball decisions.
Take a look, for example, at the PECOTA breakout rates entering this season for a select group of 24-year-old hitters:
Adrian Beltre 16% Nick Johnson 25% Miguel Olivo 24% Juan Rivera 11% Jimmy Rollins 17% Vernon Wells 22% Jayson Werth 17%
If you're trying to identify attractive targets in the trade market, or prepare for a fantasy draft, what good does that do for you? PECOTA thinks that none of the players have better than a one-in-four chance to have a breakout, and none except for Juan Rivera have worse odds than one-in-six. As I found out at the Cleveland feed, this is a source of frustration for a lot of people who are otherwise fond of PECOTA.
While PECOTA is pretty good at picking up on 'tells'--a high batting average accompanied by poor plate discipline is often a sign of weakness; a high doubles rate can be an indicator of burgeoning power--there's only so far that a projection system can take you, and we're not shy about admitting it. Certainly, taking into account other sorts of information, like scouting reports and medical records, can help to paint a fuller picture, but even so, no prediction system will be determinative. Two players that appear to be all but identical can see their careers veer off in wildly different directions.
Even so, small advantages, when repeated enough times, can soon become large ones. Say, for example, you're flipping a doctored coin that has been engineered to land heads exactly 55% of the time (just enough to provide an advantage, but not so much as to be obvious). In a best-two-out-of-three Flip-Off, the bettor calling heads will prevail 57% of the time. Best five-out-of-seven, his advantage has improved to 61%.
But make it a best 50-out-of-99 contest, and the player calling heads will prove victorious around 89% of the time; the small advantage has become overwhelmingly significant. Baseball teams make hundreds of player personnel decisions every year between trades, draft picks, salary negotiations, and so forth, and relatively small differences in organizational philosophy can come to produce huge gaps in the standings.
Lesson #6: Don't Go All-In With Unimproved Pocket 3s.
Hell, even the good ones can use a bit of luck sometimes.