Positional Scarcity

POSITIONAL SCARCITY

A concept that has been put forth in recent years is the concept of positional scarcity. However, we have never seen a calculation of it that works for "standard" leagues (i.e. 12 team AL or 14 team NL leagues).

This concept is rooted in superstition, not fact. The systems that we have seen generally say that you take a scarce position and increase the value of players there because they are hard to come by. However, in order to make this work you must also "devalue" stats at "cheap" positions, like outfielders. If you do not adjust the players at the "cheap" positions, your dollar values will not add up.

Two things work against this theory. First, most leagues award points for most home runs, so why should I care which position produces them? Why should I pay more for home runs hit by a catcher than an outfielder when they are worth the same amount in the final standings? You shouldn’t.

The second thing that works against this theory is the very gyrations that some "experts" go through to factor the theory into their systems. Some "experts" espouse the best/worst theory. Since the worst catcher is so bad (so the theory goes) and the worst outfielder is so much better than the worst catcher you will do better to take the worst outfielder and the best catcher.

In theory this sounds great, however, if the valuation system is worth anything the worst catcher and worst outfielder are about equal. This is due mainly to the positions. On most teams the back-up catcher will get at least 150 at-bats. This is primarily due to the strain of catching where the starter is given at least one game off a week. Factor in double headers, late night extra inning affairs and just nagging injuries and it is easy to see how most back-ups end up in the 150 AB range. For this reason it is fairly rare to allocate over 500 at-bats to any starting catcher.

However, in most AL leagues you need at least 60 outfielders. That is all 4th outfielders and 4 of the fifth outfielders (assuming that no outfielders are used as DH or utility). Thus, the final outfielders taken will generally have fewer than 100 at-bats. For 2003 in the NL (assuming a 14 team league) you will need all 4th outfielders and 6 of the fifth outfielders.

To show you this using concrete numbers we took all the NL catchers on opening day rosters in 1996 and all the players who qualify as outfielder only. Using a "standard" league in which you would draft 24 catchers and 60 outfielders (assuming you use no catchers or outfielders in your utility spots). The results may shock you so make sure you are sitting down.

For 1996, the 24th to 26th catchers based on at-bats were:
24 - Steve Decker SF/Col (147 AB .245-2-20-1)
25 - Tyler Houston Atl.ChC (142 AB .317-3-27-3)
26 - Keith Osik Pit (140 AB .293-1-14-1)

For 1996 the 59th to 63rd outfielders based on at-bats were:
59 - Chris Gwynn SD (90 AB .178-1-10-0)
60 - Trinidad Hubbard Col/SF (89 AB .213-2-14-2)
61 - Vince Coleman Cin (84 AB .155-1-4-12)
62 - Jacob Brumfield Pit (80 AB .250-2-8-3)
63 - Milt Thompson LA/Col (66 AB .106-0-3-1)

Now, do you really think there is inherently more value in the last group of outfielders than the last group of catchers? Of course there isn't and that is why positional scarcity does not work.

Aha! you say. We obviously used an NL example because with the DH in the AL our example does not hold water. Aha! we say back, let's look at the numbers. Remember, in the AL you have the DH but that only adds about 350 at-bats to most teams. In the NL, the pitchers are pinch hit for about 200 to 250 times per team. Those at-bats generally go to the 4th and 5th outfielders. In the AL the pitcher is replaced by the DH. On some teams the DH at-bats are shared by the outfielders. On other teams where there is a full time DH (eg. Brad Fullmer, Josh Phelps, Ellis Burks, Edgar Martinez, etc,) the only at-bats available to the outfielders are the at-bats from the 3 outfield positions. There are relatively few pinch hit opportunities in the AL. Thus, using the same analysis from 1996 as we used in the NL, we find that the 25th and 26th AL catchers (in terms of at-bats) were Chad Krueter (114 AB .219-3-18-0) and Greg Zaun (108-.231-1-13-0). For 1996, there were only 70 players who qualified at outfield only on opening day rosters in the AL. (Remember, if the positional scarcity theory holds true, players who qualify elsewhere like Surhoff and Bonilla will be assigned to positions other than the outfield.) Thus, there will only be 10 left over after drafting the 60 outfielders (again assuming that no outfielders are used at DH). Of those 10, five had 40 or fewer at-bats! The 60th outfielder (again, in terms of playing time) was Alex Diaz (79 AB .241-1-5-6) The 61st and 62nd outfielders were Dwayne Hosey (78 AB .218-1-3-6) and Turner Ward (67 AB .179-2-10-3). I bet you feel good if you used a system that year that adjusted the value of those outfielders down!

To roll these examples forward here are the results for 1998-2002: (of course, after expansion, we are looking at catchers 27-30 and outfielders 71-74 in the NL).

To summarize the average results over those 5 years we look to the average of the 4 players in each group.

NL Catchers: 119 .233 2-12-1

NL Outfielders: 102 .220 2-10-3

AL Catchers: 127 .259 5-18-1

AL Outfielders: 104 .249 3-11-3

In both groups the catchers hit for a higher average with more the same or more HRs and more RBI.

Remember, these examples assume that no outfielders are used at either DH or utility.

Again, we look at what happened in 2003. Catchers 23-26 in the AL and 27-29 in the NL (now assuming a 14 team NL league). For the outfielders we used numbers 59-62 for the AL and 69-72 for the NL.

NL Catchers: 128 .231 3-16-1

NL Outfielders: 108 .235 2-12-0

AL Catchers: 123 .209 2-10-0

AL Outfielders: 62 .211 2-7-1

As you can see the stats are remarkably consistent. As they should be! And, any good valuation system will have the last outfielders taken and the last catchers taken at the same values if they have properly allocated the ABs.

Now, back to how the theory works. In implementing the best/worst theory some "experts" twist themselves into pretzels to make the numbers work. To factor in positional scarcity they increase the value of the output of the top catchers. However, in order to keep the system in balance, they have to "devalue" similar performances in outfielders. What happens under this system is that an outfielder who sells for the same price as the best catcher will put up better stats. He has to since the system was designed that way!! If that sounds like a great exercise in tail chasing you are right, it is.

Another reason some prognosticators believe in positional scarcity is because their system tells them it exists. The problem here is with the system, not positional scarcity. If you have a system that does not tie the playing time of players to the playing times that teams really generate you end up believing that positional scarcity exists. Thus, a system that allocates 200 at bats to the 60th outfielder and 200 at-bats to the 24th catcher is not mirroring reality. However, the prognosticator looks at the system and sees the values out of whack. The 60th outfielder is worth 4 or 5 units and the last catcher worth drafting is the 18th. Rather than correcting the at-bats, he fiddles with the values. Thus, he corrects his mistake by foisting on his customers yet another erroneous theory. Our customers thank them all the way to the bank.

If you see any book or article which puts forth this theory, do your own comparison. First, make sure the dollar values add up in total. Then compare the stats of the top catcher to an equally valuable outfielder. If the author is true to his theory you will see better stats from the outfielder, you must!

Remember, a homer is a homer is a homer...

The moral here? Whenever anyone in your league espouses positional scarcity, praise them for their insight and knowledge and don’t laugh until you are out of earshot.

CAUTION: You do need to look at league size. For example, using the 2003 AL example above, what happens in a 9 team AL league? The 18th catcher was Ben Davis of Seattle (246 AB .236-6-42-0) who was decent, but not starting to get a little better than the 24th to 26th group. However, with only teams you move up to the 45th outfielder. This was Eric Owens of Anaheim (241 AB .270-1-20-11). While you can argue the merits of the particular value of Owen's numbers versus Davis' numbers, the inescapable fact is that the last outfielders are now generating as many at-bats as the last catchers. As a rule the outfielders are better hitters on a per at-bat basis. Thus, with smaller leagues and small combined leagues there can be positional scarcity. However, for the vast majority of "standard" leagues it is a red herring.

One last cautionary note. This essay pertains to auction leagues. In leagues that use a straight draft you are looking at the relative drop in value at a position. Thus the good middle infielders and catchers are draft first because there is such a drop in talent after the top tier. However, at the bottom it really shouldn’t make a difference who you take.

Return to Strategic Essay List
Return to PPI Home Page
Email questions to PPI or email support@stallvalue.com