"Measuring" a Player's Value

Yesterday, I revitalized the discussion about valuation, particularly pertaining to how much a "winning" Rotisserie team is worth.

sas4 continued to question my logic, which is good. This means that either 1) I'm not explaining myself very clearly or 2) I'm wrong and sas4 is right.

Let's take two leagues, A and B. They both have identical auctions and starting rosters. The entire season the owners in both leagues make no roster changes except that Team #6 in League A aquires and activates R. Braun and releases Counsell. Team #6 stands pat.

It seems to me that as Counsell and Braun proceed through the season to have a different number of AB, runs, RBIs. etc. They each have a different value AND they change the value of every player in their respective leagues. If the rest of the season Braun hits 9 more HRs than Counsel, that changes the number of HRs produced and that in turn changes the value of every offensive players' value.

The key to Alex Patton's pricing system is that the system is measuring the value of the average league's auction roster. In past years, Alex would take a handful of leagues, mash them together, and calculate the average number of HR, RBI and SB produced, along with the league average BA, ERA and WHIP.

So the formulas Alex uses are attempting to take into account an average league. However, you can build the formulas out for your own league if you wanted to do so. Instead of taking Alex's denominators, you can build out formulas for your own league.

In the example sas4 presents above, it actually is correct that League A and League B would have identical formulas.

Why is this? I think this is the question that I am failing to answer, which is why sas4 keeps coming back dissatisfied.

The answer is because the formulas are used to measure each player's value at the auction. And, because there is $3120 to spend in the auction the formulas demand that, when all is said and done, that the average league's auction rosters are worth about $3120.

For the purposes of this exercise, then, Ryan Braun's value for the rest of the season is useless. The Patton $ measure the auction value of every player, not the player's value compared to either the Major League or Rotisserie League universe.

Someone like Alex can calculate the value of all players based on an assumption of total ouptut by all players, active and fantasy. That projection can be fairly accurate. But the real value actually changes every AB in every fantasy league as different owners have different players up and different players down.What does this have to do with anything? I am not sure - I am not enough of a mathematician but I think it has somethimg to do with roster spots.

Alex's values are actually an assumption based on the values of all players available on Auction Day. Different leagues can indeed have different values and - if you like - you can use Alex' denominators to figure out the exact value of your league. However, the value has nothing to do with who is called up and who is sent down by your league once the opening bell sounds. It has everything to do with your auction. Alex's dollar values are based on the draft universe only.

Why is this? Why not measure the value of every player in the Rotisserie universe instead, or at least the ones who make it onto our rosters?

The masochists amongst us certainly could attempt to do this. But what would the baseline be?

$3120 (for a 12-team league at $260 per team)?

We could assign that dollar value to the formulas and simply ramp up the HR, RBI and SB totals to match our league's totals. As a result, the dollar value of every player would drop, since each star would be grabbing a smaller slice of each league's pie. In theory, yes, every player would be worth less.

In practice, though, we don't change the formulas because we know, intuitively, that $3120 isn't an arbitrary number. It's the number. It's what each league uses to buy (or freeze) 276 players at auction. The dollar values are meant to be used as a yardstick for what a player is worth compared to all of his peers that were purchased at auction.

If you did want to take on sas4's challenge and come up with a formula for the Rotisserie universe, you can use any dollar value you want. I would suggest using $4320. That's the total of $3120 spent by the league in auction plus $1200 spent by each league in FAAB, assuming a $100 budget per team.

You could do this. But a FAAB dollar is not equal to an auction dollar. The owner who spent $98 of his FAAB on Carlos Lee in my A.L. last year wasn't disappointed when Lee only earned $13, compared to the $34 worth of return Vernon Wells' owner got for his $34.

My point is that $3120 is used to measure the auction value of players because that's the amount of money that we use at the auction. No yardstick has yet been established for measuring the value of a Rotisserie universe. Any total dollar value that was assigned would be arbitrary.

Players A and B each get 300 AB, score 50 runs, drive in 60, hit 20 HR, and steal 10 bases. Player C gets 600 AB, scores 100 runs, drives in 120, hots 40 HR, and steals 20 bases. Player C is much more than twice as valuable as the total of players A and B. Why? -- because he takes only one roster spot to accomplish that output.

I know where sas4 is going here. Let's look at his point through a less elegant, but more tangible, example.

Jermaine Dye hit 44 HR, 120 RBI, 7 SB, .315 across 539 AB in 2006. He earned $36.

Nick Swisher hit 35 HR, 95 RBI, 1 SB, .254 in 556 AB. He earned $18. Reed Johnson went 12/49/8/.317 in 461 AB. He also earned $18.

sas4's argument is an argument in favor of the "Stars & Scrubs" strategy. Throw a stone at a $1 OF at your auction and you have an excellent chance of getting an OF who earns more than $1.

That is absolutely correct. However, that is also absolutely using the benefit of hindsight.

For one thing, I would hope that the major league team that owned Players A & B turned them into every day players. With those statistics, it's a shame to see Players A & B on the bench.

I know; I'm being funny. But I am also trying to prove a point, which is that there is variability not only during the regular season but at the auction as well. Anyone who bought Jermaine Dye for $22 last year who believed he was going to be the 5th best position player in the American League raise your hand.

You can hear the crickets.

Furthermore, you can see in October who the top free agent acquisitions were. In April, you cannot. The owner who bought Brian Anderson at $1 wasn't happy with Anderson's 2-for-17, 0/0/0. So who are the outfielders available in the free agent pool for an owner who needs an OF this week?

Jason Tyner, Fred Bynum, Lew Ford, Shane Costa, Tommy Murphy, Willie Bloomquist, Luis Terrero, Marlon Byrd, Hiram Bocachica, and Miguel Cairo.

Which one of these guys is an absolute sure bet to earn even a mere $1? Bloomquist and Cairo are probably the safest bets because of their speed, but if you need power more than speed you're out of luck.

It takes me back to the notion that we really do not get full value from players in that some are not active and the output is not just a function of the raw numbers but also taking up roster spots.

Most leagues do attempt, though, to squeeze the best possible stats out of the available player pool. Jeremy Sowers was waived by his owner last week; he finally decided that he couldn't risk another lousy week of Sowers. Reggie Willits was FAABed early and then waived, but was claimed on waivers the following week. Even though not one of the 12 owners in my league can see into the future, we all make very similar educated guesses about who the best 276 players are out of the 350 player American League universe. Since this represents nearly 80% (78.9%) of the A.L. player universe, opinions generally do not differ much from league to league. We are always attempting to maximize our value.

This discussion began, though, with the question of measuring value. And value is measured using the yardstick of the auction. We could attempt to measure value using the Rotisserie universe. But then the $3120 becomes arbitrary. And an arbitrary valuation system is just that: an educated guess of what players are worth.