Head-to-Head Strategy

Look, I totally understand why some of you avoid head-to-head leagues. They aren't for everyone. Like wusses, for example. Head-to-head leagues aren't for them. Or people who pretend they're way too busy to maintain a lineup every day.

If you're like that, I get it. Don't play head-to-head. Seriously. You'll just lose. Standard 12-team head-to-head leagues require your attention. There's winning and losing every week. It's a little stressful, and sometimes it feels a little arbitrary, too. In that respect it's like actual baseball. Maybe that's why I prefer it. Head-to-head offers both the season-long competition that exists in roto leagues, and weekly competition with a specific opponent. And if that opponent happens to be an insolent punk who makes ludicrous demands of the commissioner, arrogantly smack-talks the league and is perhaps a bit too proud of his Alex Gordon pick, then all the better – I'm looking at you, Supreme Clientele. Week One, chump. There will be a reckoning. Oh yes, there most certainly will.

For the most part, you don't need to radically alter your pre-draft rankings for a head-to-head league. There are a few unique strategies that come into play, though, and there's at least one essential concept you need to understand:

The Head-to-Head Uncertainty Principle

In quantum theory, the Heisenberg Uncertainty Principle tells us that the more precisely we know a particle's position the less precisely we know its velocity, and vice versa. This has something to do with photons and wavelengths and the characteristics of very tiny things. But you probably didn't come here to get tips for your fantasy quantum physics league. So let's get to baseball.

The head-to-head uncertainty principle tells us that when the season is parsed into week-long slices, we can't accurately predict any individual player's performance. It's only seven days. Baseball is a game in which it literally takes years for a player's true ability to express itself. Over a full season, we can confidently say that Mariano Rivera will get 35 saves, Juan Pierre will steal 45 bases, Adam Dunn will hit 40 homers, and a healthy Chris Carpenter will win at least 15 games. But what will those guys do next week? Maybe nothing. Maybe a lot. None of them can guarantee you a win in any stat in a given week. If you want to regularly win a category, get multiple players who excel in that category. Redundancy is key. Don't simply rely on one player's proficiency in any given area. They won't distribute their stats evenly throughout the season. Nobody does that. Players tend to binge, then fast.

To better understand the uncertainty principle, let's take a look at an elite base stealer. Carl Crawford led the American League in thefts last year with 58. He's a top-tier fantasy outfielder and a clear first-round pick. Basically, Crawford is to fantasy experts what the ACC is to Dick Vitale: an institution with mythic talent that cannot be overhyped. In roto leagues you can draft Crawford and feel confident that you've addressed stolen bases to a reasonable extent. No need to think much about them in your later picks, really. If you accidentally get a couple more league-average base stealers, you're no worse than middle of the pack. But head-to-head leagues are trickier. You can't lean on any individual player too heavily. Take a look at Crawford's stolen base totals from April 2006:

Week One: 1 SB
Week Two: 2 SB
Week Three: 1 SB
Week Four: 2 SB

Useful, sure. But two stolen bases are never enough to win the category. If you don't have other sources for steals, you lose. Crawford did have an eight-steal week in May, though. That was probably more than enough to win the week. Later in the year he posted a six-steal week and a five-steal week – that's a third of his total stolen bases in only three weeks, for those keeping score. After a three-steal game against Boston on July 5, Crawford stole exactly one base over his next 26 games. That's four weeks with only one steal. It happens. It's not unique to Crawford. He can certainly help you in head-to-head leagues, but don't expect him or anyone else to single-handedly deliver a bunch of category wins. That takes redundancy. If you're determined to control steals every week (which is kind of silly, because steals aren't strongly correlated with other categories), you need to draft multiple guys who are better than league average.

Even though one guy by himself can't win a category very often, one guy is still better than no guy. Don't use the head-to-head uncertainty principle as justification to totally ignore a category in your league. That's never advisable.

Don't punt categories. But if you do, which you shouldn't, at least apply token pressure.

If you punt a category with me, I'll make you regret it – Didn't that sound especially tough for a fantasy writer? You need to apply at least modest pressure everywhere.

In one of my head-to-head leagues there's an owner who likes to punt saves. He's done it before, and he's doing it again this year. We count holds in that league, so he usually starts five quality middle relievers and expects to win holds and maintain good ratios – In public 5X5 leagues, sometimes you'll encounter an owner who loads up on low-ratio relievers, then streams starting pitchers. We'll discuss streamers soon. You can manage around them. This owner's strategy works against opponents who don't pay strict attention, but it's a disadvantage against anyone who realizes that it just takes one save to win the category against him. If you get three or four saves, they're wasted. So I'm just looking to pick up a single save on Monday or Tuesday, then I'll bench my closers for the rest of the week and compete everywhere else. If that guy simply had one closer, I'd have to approach the week differently. But he never does.

Token pressure in fantasy baseball is like token pressure in real basketball: you're just trying to make the other guy work a little. Wear him down. Make him start Todd Jones or Ryan Dempster or some other ratio albatross. That's all you'd like to accomplish. In rotisserie leagues, one of the worst things you can do is punt a category yet still keep one dude around to accrue useless stats. You'll get one point if you have zero saves; you'll probably still get one point if you have 25 saves. But in head-to-head, that one guy can keep your opponents honest while you maintain control of your core categories.

Still, I don't recommend premeditated punting, ever. If your league is truly competitive, it's difficult to overcome a guaranteed category loss. However, you might find yourself hopelessly behind in a category on Thursday and willing to give up on it to focus on winnable stats. That's completely reasonable.

Keep an open roster spot (or two) and think of the waiver wire as a really big bench.

In a 12-team mixed league there will always be lots of un-owned talent. The last guy on your bench usually isn't any better than the top ten available free agents. You can either ignore that fact, which many owners do, or you can use it to your advantage. I'm a big fan of the single-serving offensive add on Mondays and Thursdays – those are the MLB travel days, when only a handful of teams are in action. My last two bench spots are usually revolving doors, and I'll gladly drop Garret Anderson for Jose Vidro if the Angels have an off-day and the Mariners face a user-friendly pitcher. My other bench spots are typically reserved for great players who happen to not be playing on a given day.

A few seasons ago I picked up Javy Lopez on a Sunday, because the Braves were among the few teams in action on Monday. Lopez hit a grand slam, and I promptly dropped him and added someone else. It was just awesome. The last thing I needed was a second catcher occupying a bench spot. That move led to a category win, too. It remains the greatest single-serving add of my fantasy career.

You might reasonably ask, "Isn't that streaming and isn't streaming unethical?" The answers are "Sort of" and "Hell no." If you're playing in a public league without transaction or games-played limits, people are going to stream. That is, they're going to add and drop players every day. Typically they stream starting pitchers. We make it easy to do with that nifty little drop-down "P (Probable)" search option, too.

I don't normally go into a head-to-head league intending to stream pitching, but I'll do it in earnest mid-week if I think it can help me salvage a few category wins. Why wouldn't you? It's not illegal, and it only takes seconds to find a useful add. Spare me the sportsmanship nonsense. If something is within the rules of the game, then you shouldn't get too mad about it. If you're militantly anti-streaming, design a custom league with limits or add a few stats, like Ks for hitters or BB for pitchers, to discourage the practice. That's simple enough. In any case, streamers are not unconquerable opponents.

Yes, you really can thrash those streaming weasels.

I don't really think they're weasels, of course. I've already admitted to situational streaming. The real advantage that dedicated streamers should have is in the hitting categories, not in pitching. You can tell during a draft when someone plans to stream if their first nine picks are hitters, and their next three are closers. They're usually just playing for a 3-2 split in the pitching stats, but they plan to dominate in the offensive categories. So if you want to match-up well with them, you'll need to have a formidable group of hitters, too.

The best way to topple a pitcher-streaming opponent in wins and Ks is to stream better than they do. Get the best of tomorrow's starters before they can. However, that might require you to be awake at the exact moment the Yahoo! day flips, and that might be impractical. If so, here's an idea you might want to save for the playoffs: stream a day in advance. This is one of the better fantasy baseball strategies I've ever concocted, and I've used it to great effect in meaningful match-ups with streaming-dependant owners.

Streaming a day in advance means that you're grabbing the best available starter not for tomorrow's games, but for the day after tomorrow. If you do it for a full week, you'll eliminate seven of your opponent's best starts, which will cost them wins, Ks, and very likely hurt their ratios while helping yours. You'll need those two available roster spots to pull it off. Here's how it works: on the Saturday before you face the streamer go to the MLB "Scores & Schedule" page, figure out who the best un-owned starter is for Monday, and add him. Then on Sunday, find the best of Tuesday's starters and add him. Do this every day of the week and you've achieved a significant advantage. Your opponent probably won't recognize the tactic for several days, and by then it's often too late for them to recover.

Yeah, I know. I truly am a formidable fantasy theorist, eh?

Don't forget to employ the strategic Sunday benching.

Remember that league where holds matter? Well, two seasons ago in the playoff semi-finals, I enjoyed a comfortable lead in holds and a slight lead in ERA on Sunday afternoon. I was starting the hugely undervalued Scot Shields in a RP slot. I was ahead 6-5-1 and it appeared that I was moving on to the championships to defend my league title. And then Mike Scioscia intervened.

For reasons that have been lost to history, the Anaheim manager brought in Shields – one of his best relievers – to pitch the ninth inning of a game his team led 5-1. Shields promptly gave up a walk, a single, a wild pitch, another single, and was pulled from the game. Two runs eventually scored. But thanks to Francisco Rodriguez, the Angels still beat the Tigers 5-3. My fantasy team, however, lost 5-6-1. Those two earned runs were enough to cost me ERA for the week. The date was September 18, 2005. Last year I commemorated it by doing horrible things to a Mike Scioscia rookie card – Not the one with Fernando, though. I have too much respect for him.

While I still place most of the blame for that loss on the Angels' skipper, I'm willing to acknowledge that I could have managed around his unimaginable, spirit-crushing foolishness. I could have simply benched Shields to begin the day. Quite often, a player has a greater chance to hurt you on Sunday than help you. I've seen many owners lose a category because a hitter goes 0-for-5 or a pitcher gets rocked in the Sunday night game – That game is really just a cruel joke, for reasons that have nothing to do with Joe Morgan. So don't be reluctant to bench a player on the weekend if he can't help you. That's an element of head-to-head strategy that can't be ignored. Failure to bench has cost me at least one title.

I intend to reclaim it this year, though. The first skirmish begins next week.

What to Read Next