A fan of the Yankees, Red Sox, and large sample sizes.

 
CONTEXT AND SCARCITY

If a baseball team wins 80 games, is that good?

Depends. Is this a major league baseball team during a 162 game season? Is this a major league baseball team during a strike-shortened season? Is this a major league baseball team over an entire decade? Is this a little league squad that only played 80 games over one summer?

If a baseball player (in MLB) hits 30 homers, is that good? Is it as good if he plays in Colorado? Is it as good if he plays in Detroit? Is it good if he only batted 200 times?

I think you get the idea. Whatever happens on the baseball field is context dependant. During a MLB season, many things tend to cancel each other out. Nobody bats against Pedro every game. Nobody hits in Colorado every game (although half the time is pretty significant.)

There are many things in "real" Major League Baseball that are consistant year to year and aren't given second thought. The number of players per team, the numbers of teams in the league, and the way the game is scored (more runs win) are just a few examples.

In fantasy baseball, however, these things can change. And most people don't adequately adjust their strategy for different situations. Is a fantasy league head-to-head or rotisserie? Is it 5x5, 4x4, or some crazy 8x6 configuration? Are you playing against 8 other people, 12 other people, or 20 other people? Do you have to draft five outfielders and two catchers, or two outfielders and five catchers?

When you start changing around the number of teams, and the number of players per team, the amount of players that will end up on someone's team changes. The total number of players drafted and the number of players at each position determine how you value those players. And that, my friend, is the most misunderstood (or perhaps the most ignored) aspect of fantasy baseball.

Eventually I may post something longer, but here's the quick and dirty explanation on how to value players: Find the replacement level at each position. By replacement level, I mean the amount of HRs, AVG, RBIs, RUNs, whatever, that you'd get at each position if you had to settle for the worst player picked. For each player above replacement at each position, add up their stats above replacement. All of the stats combined from every player slated to be drafted make up the draftable pool of stats. Take the amount of money to be spent per category ($260*hit-split*#teams/#hiting-categories) and divide it by the draftable stats in each category. This gives you a category $/stat ratio. For each player, multiply his stats above replacement by the category $/stat ratio to find the amount of money he earns in that category. Add up the values in each category to get each player's total value.

Don't pay for what you can get for free, and pay for stats according to what percentage of the overall pool each player provides. 10 homeruns is awful from a player when you can get a 15 homerun player from the free agent pool and there are 1500 total homeruns to be drafted. But if you can't get anything more than 5 homeruns from the free agent pool and there are only 500 homeruns to be drafted (circa 50 years ago), then that player has a lot of value.

Are things really that different between a 12 team mixed league and a 12 team "only" league? What's this scarcity thing all about? I mean, isn't it harder to find a quality SS in a deep league than a shallow league? You can't draft a bad player when picking from both leagues. This goes back to the idea of context. Carl Everett was pretty good in an AL-only league last year. Pretty studly in fact. But in a mixed league, he actually loses a lot of value. Why? Because he's not that much better than replacement level for that size league. If it's true that anyone can find a pretty good outfielder, then in order to win, you need to have players better than "pretty good."

Ok, so let's talk actual numbers. Here are replacement levels for four different groups of players, based on 2003 stats in a 12 team 5x5 mixed league. Traditional 14 hitter lineups are used.

12 Teams	Outfield	First/Third	Second/Short	Catcher
repxH	-4	-6	-5	-5
rephr	15	18	10	7
reprun	65	74	65	34
reprbi	55	76	48	36
repsb	8	5	7	2

What does that table mean? Well, for example, a catcher will have, on average, 40 fewer RUNs, 40 fewer RBIs, and 11 fewer HRs than a firstbasemen with equal value. That's a fact. How much more would you pay for a player with eight extra homeruns? That's how much more you should pay for the privelege of moving a firstbasemen over to shortstop. Throw in the thirty extra RBIs and ten extra RUNs you'd get from the move, and you've got yourself a significant position scarcity effect.

Let's look at the same table, this time for a 24 team mixed league. This simulates a 12 team "only" league, but is easier on me because I don't have to remove half the players from my list, and also removes the flukish differences between the AL and NL player pools.

12 Teams	Outfield	First/Third	Second/Short	Catcher
repxH	-8	-6	-8	-8
rephr	6	8	5	6
reprun	23	30	34	21
reprbi	25	31	30	24
repsb	2	2	3	1

What do you know... each position has replacement levels that are almost exactly the same. Only a difference of three HRs, thirteen RUNS, and fourteen RBIs. Position eligibilty is not as significant in deeper leagues, although it still affects value to some extent.

How do the top players compare between the two size leagues - do their rankings stay the same?

12 Team League:
Soriano
ARod
Sheffield
Pujols
Beltran
Lopez
Renteria
Pierre
Boone
Garciaparra
Helton
Bonds
Podsednik
Ichiro
Wilson
Wells
Lee
Ramirez
Delgado
Furcal
Crawford
Posada
IRod

24 Team League:
Pujols
Pierre
Sheffield
Beltran
Soriano
ARod
Renteria
Podsednik
Ichiro
Helton
Crawford
Bonds
Nomar
Boone
Wells
Ramirez
Lee
Lopez
Abreu
Delgado

Those are similar lists, but in very different orders. As was shown with the replacement levels, the middle infielders and catchers lose ground to the outfielders and corner men going from a 12 team to 24 team league. It's also interesting to note that players at the same position can flip-flop rank order. Why is that? Because the player pool changes. Adding in more players alters the total number of each stat available to be drafted. Flooding the draftable player pool with more SBs relative to HRs decreases the value of each SB relative to each HR.

Finally, let's take a look at the SB versus HR issue. So many people out there don't believe Juan Pierre, Carl Crawford, and others like them are worth so much. Listen, there aren't many SBs available. When a player has such a huge chunk of the pool, they're worth a lot of money. To paraphrase Lawr Michaels from the MLB experts draft, a category is a category. It doesn't matter if you get your points from the ERA category, the HR category, or the SB category.

In a 24 team mixed league, there are 2700 draftable HRs versus only 1600 draftable SBs. For every SB drafted, there are 1.6 HRs drafted. So a 60 SB freak like Pierre is like having a 96 HR power hitter. Can you imagine having a hitter on your team with 96 HRs? Yikes. Sure, the power hitters usually also huge RUN and RBI numbers, adding to their value, but the value advantage in those categories is nothing compared to the SB (50 cents each) versus HR (30 cents each) advantage. Add in a solid AVG and RUNS, and you've got yourself a speedy fantasy stud.

As always, if you've got questions, want more info, or just want to invite me into your fantasy league, email me.

Posted by Sky at 8:10 PM | 0 comments