Wednesday, August 22, 2012

Evaluation or Strategy


The web site Fantasy Baseball Crackerjacks asked a few "experts" (including me) to weigh in on the following question:
...a lot of strategy guides, fantasy baseball blogs, and experts lean in one of two directions as to the secret for fantasy baseball success: player evaluation or strategy. So is it top-notch player evaluation or cunning strategy the secret?
This is their take on the question.

My belief is that while evaluation is important, strategy is the difference between doing well and standing out from the rest of the pack.

Evaluation
The first thing I do every year is sit down with an Excel spreadsheet in mid-January and come up with raw bid values for the top 276 players in the American League and the top 299 players in the National League. I might miss a few back-ups or middle relievers on my first pass, but the point of the exercise isn't to be precise but rather to get a rough idea of what I think players are going to be worth before the CBS expert auctions in take place in early to mid-February. Years ago, I used to "cheat" by looking at a magazine, but have now developed enough confidence in my prowess that I don't look at the magazines or web sites of the other experts until afterward. 

Every year I find the same thing. While there might be some differences of opinion in what we all think a player is worth, there isn't as much variance as you might think. This isn't just true for the no-brainer superstars like Ryan Braun or Albert Pujols, but also across the board for nearly every player in the pool. Over the years, I've documented several cases of players where you'd expect there to be a great deal of variance due to various factors (rookies, injury, uncertain playing time) but where the salary for the player is nearly identical across all three expert leagues that I track (CBS, LABR, and Tout Wars).

Is this because experts have the same vanilla opinion for nearly every player that comes up in auction? No, not at all. But when experts sit down at an auction, they aren't making predictions but rather placing bets. We're not bidding based on what we think a player will do. Instead, we consider what a player might do, consider best case and worst case scenarios and decide what to pay with all of the information that we have at our disposal. Most experts devise their bids based on middle-of-the-road expectations, not pie-in-the-sky or doom-and-gloom forecasts.

The result of this is that while Scott Pianowski and I might have had a completely different take on what Alejandro De Aza was going to do coming into 2012, our bid limits probably weren't more than $3-4 apart. Even if I thought coming into this year that de Aza was going to earn $20, I never would have bet $20 for him. And even if Pianowski thought de Aza was a bum whose 2011 was fueled by a ridiculous BABIP, I doubt Pianowski would have let de Aza go for $8. In this hypothetical example, while my raw bid might have been high and Scott's might have been low, our adjusted bids would take into the account the idea that there was a lot of room for variability in what de Aza might do and meet somewhere in the middle.

Evaluation is important, but not nearly as important as it was when I started playing this game in the late 1980s and there was no Baseball Weekly, no Internet, and USA Today/ESPN were the only games in town for baseball news. Back then, you had to dig, dig, and then dig some more to glean every nugget of information about a player, and it was quite possible that some of your opponents had no idea who certain players even were.

Those days are long gone. We all have the same knowledge base, and while we might have differences of opinion, they're not so great that player evaluation is going to win you your league...unless you're in a very weak league. You need something else.

Strategy


In Rotisserie, beware of "The-Single-Theory-To-Explain-Everything-Maniac" (shown in this awesome Matt Groening cartoon in the bottom left). In Roto this is characterized by experts that tell you there is always a right or a wrong way to play the game. "Always spend $200 on hitting." "Never spend over $30 on an ace." "Don't dump categories." "Don't make trades until May 15." All of these theories have some general validity. In practice, though, you might find yourself in a situation where you want to spend $160 on your hitters, where spending $35 on an ace starting pitcher isn't a bad idea, or where dumping steals is going to push you to a title.

My first foray into serious strategy came in a home league that I joined way back in 1996. I inherited a terrible freeze list and everyone assumed that I was going to do what all new owners do in keeper leagues: buy a handful of expensive players and dump in early May. Instead, I embarked on a no power strategy, making what appeared to be unfavorable trades before we even submitted our freezes. It didn't quite work (I came in third and lost by 2 1/2 points), but instead of spending five months in the second division waiting for football season to start, I spent the entire season in a fun, competitive race. While I didn't win, that season set the stage for turning me from an above average Rotisserie player into a great player. 

Can you afford to add a poor ERA/WHIP starting pitcher because you need wins at any cost? Should you trade away Rajai Davis - even though he's your only speed guy - because you have a better shot at gaining 6-8 points if you do so? Should you waive a good player in the hopes that someone at the bottom of the standings claims that player, allowing you to pick up points by indirectly taking points away from a competitor? A strong owner will consider all of these options under the right circumstances...and make the correct decision based on his ability to properly analyze his league.

I can't tell you what you can or should do in your own league. Becoming a great player means honing your own instincts and developing your own sixth sense for what is right or wrong for your team. What I can tell you is that you should never get caught in what I'll call The Culture of Can't. If you tell yourself that you can't ever dump saves, you can't ever allocate your FAAB a certain way, or you can't ever carry a dead spot at third base because it just isn't done, you are painting yourself into a corner. Even worse, your savvier opponents will quickly figure out all of the things that you "can" or "can't" do and will act accordingly. This provides the double-edged sword of putting you at a disadvantage and giving your opponents an advantage.

If you want to be a perennial winner, you want to develop a reputation as the smartest guy in the room. You can do this by:

1) Having strong valuation/evaluation principles that build you a strong core every single season.
2) Working all season long to identify opportunities - whether through the free agent pool or trades - to improve your team.
3) Developing a picture of your league's categories that allow you to see opportunities that your opponents cannot see or refuse to take advantage of.

If your Rotisserie squad is a house, evaluation/valuation is the foundation on which you build your dwelling. Strategy is how you make sure you get the most value out of your investment. In some leagues, you can make these adjustments at the auction. In more competitive leagues (like Tout Wars), you have to be willing to make these adjustments during the season as you go along.

4 comments:

Rob said...

Good post...
I finally, this season,figured out that strategy was important... My league goes crazy for closers in-season. So I loaded up at the draft and traded them away all year to get myself ready for 2013 and beyond... I'm middle of the pack now, but have a great chance to be a favorite heading into next year. (Pat myself on the back - I have trout, matt moore and jason heyward for less than 8 bucks each thanks to this strategy)

clavejones said...

Excellently said!

Both important, but you can't win without employing strong strategy, especially since everyone is using the same raw data to pump out their projections.

George said...

For me, there is one fundamental rule about the draft (which can also be translated to auctions):
Don't waste your top picks.

It is far, far more important to avoid mistakes than to try and catch lightning in a bottle. The key to player evaluation is to try and limit the chance you'll blow a bunch of resources on a non-contributor.

For strategy, the big thing I try to do is evaluate which guys I drafted really just represent roster slots for churning. As the first month or two of the season progresses there will be lots of opportunity to improve. You need to show no love for the roster filled you either bought cheaply or burned low draft picks on.

mike fenger said...

Nice piece. To amplify Rob's point, one aspect of strategy is to know your league. That's the market that will ultimately set prices in your auction. As Mike says, different people will have different opinions on what a given player will do going forward, but that isn't likely to mean a difference of more that a few bucks in their actual bid values.

So, in making your price list that adds up to the total amount to be spent, you do need to put at least reasonable values on players you don't like much. You want a reasonable amount of fudge factor for the guys you do like, and you want to price enforce to at least some degree (as in Mike's example of Pianowski and De Aza).

(BTW, that's even weirder trying to come up with price lists for publication, that is, outside of a specific league's context. Just another variable -- it's even more important to be a bit conservative, but also to not be unrealistic. At the pattonandco website last year, we had some discussions about Juan Pierre, for example; I knew I wouldn't buy him (or recommend him), but I found it hard to come up with a price under those circumstances. High enough not to be silly, but still low enough that it was hugely unlikely that the bid would get him. Doesn't help that he has out-earned my bid (and everyone else's!). I guess this is also an admonition to take expert price lists as a starting point, to be adjusted based on your league's market.