Sunday, May 31, 2009

Depth Perception

A friend of mine recently said to me, “If you were in a deep league you could pick up Gordon Beckham and stash him on your bench, but I wouldn’t do it in a shallow league.” While I agreed with him at the time, I got to thinking about exactly how deep the league would have to be in order for this transaction to make sense. We have statistics for everything else; surely there is a statistic that measures exactly how deep my fantasy league is relative to other leagues.
How much more accurate would it have sounded if that same friend had said “In a mixed league that has a 65% League Depth or higher you could get away with stashing Gordon Beckham on your bench, but your league is only 45% deep.” That would be awesome, right? I thought so too, and those numbers above accurately represent a new statistic I’m referring to as League Depth Rating (LDR).

The vagueness surrounding how we discuss depth in regards to fantasy baseball leagues may seem somewhat arbitrary, but it often inhibits reasoned discussions from taking place. After all, nobody is going to seek advice on whether or not to drop Jose Reyes, but if the conversation turned to someone of lesser value, Melky Cabrera for instance, anyone giving advice would have to immediately comprehend the league depth in order to give a rational suggestion. Generally this means the league participant saying something along the lines of: “12 team mixed, 5x5 roto, 23 man rosters with one DL slot.” That is the absolute shortest description available to us now, but not only can we shorten it even further, we can make the new version even more informative than the current, longer version.

Let’s break down exactly how to calculate LDR. The first component is establishing a bottom, or 0%, for our scale. In all types of leagues, the bottom will be 150 players. This is based on a projected 10 team, 15 man rosters, and no DL slot league. Could there be a shallower league than this? Probably not in terms of total players, but supposing the answer is yes, the LDR would simply become a negative number.

On the high end of the scale, the deepest league will be a 15 team, 28 man roster, and two DL slot league. This equals 450 players. As discussed above, if a league is deeper than this the LDR would be over 100%. There are two reasons I want to stay at these numbers:

1. By limiting the player pool to a number inside of the top 500 ranked players, the possibilities still exist for us to easily quantify value based on depth. A pattern will develop in each individual league that will allow a fantasy manager to easily identify positional depth both preseason and in-season. Once you get outside of the top 500, projections become less accurate, thereby creating certain illusions of positional depth that may or may not exist.

2. If the formula is to work with both a mixed league and a league-only league, the two ends of the spectrum must be manageable, particularly the high end. In a league-only league, it would be virtually impossible to have a league deeper than 450 players. Especially when you consider that in an AL-only league, if you add up all nine positional starters (including DH), all five starters in each rotation, and four relievers from each team you would only be counting 252 players. That being said, the use of LDR in league-only leagues will be somewhat limited, as these leagues are already deep by definition.

Now that we have the bottom at 150, the top at 450, with a span of 300 in between, what do we do with these numbers? Let’s start by using an example. I’m in a mixed league that has 13 teams, 25 man rosters, and two DL slots. Add it up and this equals 351 players. To calculate the LDR of this league we take the number of players in the league (351), subtract the bottom (150), and then divide the new number by the span (300). The LDR of this league is 67%. Here again, the formula:

(Number of Players rostered in League – 150) / (300) = LDR

Now that we have established LDR in terms of how to define it, the real question becomes how can we use it to our advantage? If I know that my league has a 67% LDR, what does that tell me that my competitor may be unaware of? Individually, if you are prone to making your own preseason projections, you can use it to your advantage at least a couple of ways:

1. Pre-draft, if you eliminate positional scarcity when making your projections and go solely off the actual production, you can determine positional scarcity later on by counting up the number of players at each position inside the total of all players that will be drafted. Granted, we do this now to a degree, but the difference is I now know that if I have a few different leagues that I’m draft-prepping, I can quickly say to myself: “In this 38% LDR league, OF’s will be more scarce than 1B, but in this 59% LDR league, I can wait a little longer on OF and take an elite 1B early.”

2. Regardless of what type of league you are in, knowing how to project the production of first-year players and prospects is both extremely difficult and highly valuable. Predicting the unpredictable is often the difference between winning and losing in fantasy. Where the value of having LDR comes in is that you will be able to determine the amount of available roster space you can use for upside players. Back to the Beckham example, say you want to shelve a prospect like him on your bench, you will first need to know if you if you will need his value at some point. If Beckham is SS eligible, how many other SS’s are owned in your league? Beckham’s value will be high if the best available SS via the free agent pool is Tyler Greene. Conversely, his value will be low if Jason Bartlett is available. LDR, used properly, will work very well in conjunction with a statistic like Value Over Replacement Player (VORP). I see this tool as something along the lines of “Fantasy League-specific VORP”. Simply put, this all has to do with relative value. Obviously, LDR is not going to be a tool to help you determine the actual value of the player, with regards to their actual production, but knowing the relative value is just as important. There are a lot of methods already being utilized in this quest to determine relative value, and VORP is but one of them. LDR can be a new weapon in the fantasy manager’s arsenal.

For those looking to expand the use of LDR more universally, the advantages are even more obvious. Fantasy advice columnists and bloggers are always looking for ways of dispensing information that is relative to everyone. They often fall short, though, and can only reach those managers in the most normalized league formats. Whether this problem is of their own making or simply an inherent flaw in the system, it is a problem that needs to be resolved. Take this article by The Hardball Times: If, instead of "last player drafted", they used the last player ownable at each position per every ten percentage points of LDR, they would be able to tell you the PRLP before you drafted. LDR is not going to completely fix these issues, but it could be a big first step.

The methodology currently in place to determine positional depth assumes an average player pool of 300 players or so. If you are in a very shallow or a very deep league, you would not be facing the same situations as those outlined by someone writing an article based on that number. Conventional wisdom tells us that the common shallow positions are generally 2B, 3B, and C. But in a shallow league, where you need a star at every position, the position of least depth is probably OF. This may seem obvious for the extremely shallow leagues, but where is the line? Think about all fantasy leagues as a big swimming pool for a second. You have a shallow end, a deep end, and in the middle there is a long ramp. The ramp is the area of concern here. Where is your league on that ramp? That’s what we can determine with LDR.

Example: Many of the 3rd-tier OF’s are going to be rated somewhere between 150 and 250 on the average mixed league player rater. This means, assuming three OF spots, one-third of the ownable OF’s in a ten-team mixed league will not be among the top 150 total players. Whereas, assuming one 2B spot per roster, nearly all of the ownable 2B will be ranked inside the top 150. In this league, if you decide to draft all of your outfielders before you take a second baseman, you would be justified in doing so.

As a fantasy baseball player, you want to be able to utilize all of the advice out there from columnists, bloggers, stat sites, etc. However, you also have to decide for yourself what information is relative to your league. Often, this is just a guess. Over the last few years, though, there have been many tools that have been developed that have taken the guesswork out of fantasy baseball. If this is the objective, then LDR will be another example of one of these tools.

No comments:

Post a Comment