Advanced supply and demand

Anyone who’s had some basic economics, or even just basic cultural exposure to the ideas therein, knows about supply and demand. The gains from bringing it to your fantasy baseball strategy are apparent. The supply of pitchers goes up relative to the demand—well, that means that pitchers’ prices (aka their values) go down. Well, perhaps. You must be careful with what it is that is actually being demanded and supplied. As always, if you’re not fastidious, economics can take with one hand what it can give with the other.

Take the following example that came up as I wrote about my home league: In my league, we have holds as a scoring stat. So middle relievers are valuable and there are fewer starting pitchers that end up on starting rosters. What should holds do to the value of starting pitchers?

At first blush the answer is: Holds reduce the demand for starting pitching but don’t change the supply, so the price (value) of starters should go down. So, the value of Tim Lincecum in my league, according to this theory, would be lower than his value in standard leagues. This approach is almost certainly wrong. Why?

First of all, as noted in the comments to last week’s article, it is correct to say that the stats of a replacement starting pitcher go up. With holds, the last undrafted started may be Derek Lowe, rather than Ryan Rowland-Smith.

But in the land of the one-eyed, the two-eyed man is king. What you buy in rotisserie leagues (as opposed to, maybe, points leagues) are better stats, not absolute stats. You don’t care if you win by one home run or 20. In the same respect, if everyone has better starters, then the only way to get value out of your starting pitching is to have the best. The price for Lincecum should probably stay about the same or perhaps even go up. What the addition of holds does is skew the value distribution: Guys like Rowland-Smith give negative value now, and Lowe gives zero (since he’s replacement level), but Lincecum is worth around the same. You can blow your entire starting pitcher budget on fewer pitchers now—so even if the budget is lower (since you have to spend some on setup men), you have fewer pitchers that you need to spend it on.

Now there is a slight caveat here: To the extent that you can get stats from different positions, there will be a decrease in the high-end starters’ price. If those holds guys really could help you in WHIP or strikeouts, then the value of a high-strikeout starter would be a bit lower. But this is second-order compared to the two-eyed-man effect.

However, when it comes to batters, the preeminence of the two-eyed-man effect is not necessarily true. Different positions can heavily cross-contribute—mostly you expect contributions (or hope for them) in most batting stats from most positions. Maybe you hope for a bit more power from your first baseman and more speed from your shortstop. But in general, an increase in demand at one position will affect the price of another position.

Again, though, one must be careful. Certain fantasy analysts at behemoth Sports Network have claimed that speed shouldn’t be valued this season because you can basically throw a dart and hit a speed demon in the outfield. Let’s just assume that there is actually a higher supply of stolen-base threats this year versus past years. Does this mean that you shouldn’t try any more for Michael Bourn because there are many marginal speed guys this year? Well, if everyone is getting 20 stolen bases from the center field spot, everything else equal, I want 40 from mine. Sure, 19 is useless to me—that’s the skew working—but I still think 40 is pretty valuable; wasting two starting roster spots on two 20-steal guys is not the way I want to go, if possible.


8 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Werthless
14 years ago

There are two main aspects of the price one is willing to pay in a draft or auction, and you addressed one of them here. 1) The expected value that player X will achieve over a comparable waiver wire addition. 2) The range of outcomes that a player may achieve around that mean.

For brevity’s sake, this article (perhaps inspired by the discussion about the value of SPs in a holds league) focuses on the first aspect, but I feel that the unstated range of outcomes is an important factor backing up our intuitions. Lincecum retains his value in a league with few SP spots because his his range of outcomes is so low… we know he’s going to be good. And with replacement level so high, that’s the only type of pitcher worth investing in. Otherwise, a team could hope to compile the numbers of a Cole Hamels type (3.70 ERA, 1.20 WHIP) by investing in a handful of replacement level starters (Liriano, Zambrano, Harden, Price, Cueto are replacement level in a 12 team league with each team having 4 starters) for pennies.

Part of the reason that Lincecum retains his value (despite his VORP being lower) is that he has a very high chance of producing positive value.

What I’m trying to say is this: As the league settings reduce VORP by Z% at a certain position, the prices of everybody but the top players at that position decrease by (Z+A)%. The same effect occurs when the number of fantasy teams is reduced from 12 to 6, say. Hanley Ramirez is still very valuable (and would be more pricey in an auction) but Rollins/Jeter/Zobrist etc, who are still top 6, will go for drastically lower in an auction.

Chris
14 years ago

But wait…however..although…yet again…trick ending coming…

Chris
14 years ago

Sorry that was not a constructive comment. The article was just a lil hard to follow because it seemed like you’d say one thing with two caveats attached that possibly negated what was just said.

Derek Ambrosino
14 years ago

Jonathan,

I think that both you and Werthless make good points. Perhaps you didn’t give enough info about the context of your home league in your original article, which led people to react reflexively to your argument. The argument is sound, and I agree with it in theory.

But, Werthless does point out two important caveats. The first is reliability. Pitchers like Hamels only really have that value if they perform on the high side of their reasonable projections. If Hamels has a season in which he performs in the 50th percentile of a reasonable range of outcomes is for him, then his performance could easily be captured by a pitcher a tier or tier and a half lower than Hamels performing at the 90th percentile of his reasonable range of outcomes. What you are really paying for is probability, which is totally fine and it’s something I preach too. You can’t go around assuming that you are going to find sleepers that match your excessive expectations all over the diamond. Some managers are either delusional or just need to join more competitive leagues…

The other point um, werth mentioniing is that price per unit of production is not necessarily linear throughout the production curve, for lack of better term. Perhaps you can get 80% of Lincecum’s production for 40% of the price. That’s great, you made a great pick. But, you still have to make up that production somewhere if you want to compete. Even though you can get 80% of Timmy for 40% of his price, you can’t get 100% of Timmy in one transaction anywhere, for any price! And, that’s what many neglect in analyzing this point.

I wrote about this a few weeks ago regarding A-Rod’s contract, noting how silly it is to measure a player like A-Rod by something like dollar cost per win share – he’s so beyond the margins of a replaceable player that the linear price/production relationship ceases to hold.

Now, some may argue that the way you make up the missing value on the 80/40 Lincecum proposition is in offense. What you are “banking” is more resources (dollars or high draft picks) to spend on hitting, so you hope to draft wise enough in pitching to compete and use your surplus to bolster other categories. That’s a fair holistic argument, but it’s not an argument about Lincecum’s absolute or relative value; it’s a (hardly novel) argument about overall draft strategy.

Derek Ambrosino
14 years ago

BTW, in terms of economics this is a much more complicated question to answer, but just rhetorically – if you have Lincecum, for example, what is the the added value of nobody else being able to have him?

I always thought about this with the Bulls and Michael Jordan. In 96-97, the Bulls paid him 30M, and in 97-98, they paid him 33M. That’s effing insane. But, you know what, but if the Knicks, Jazz, Pacers, Sonic, Blazers, Rockets, Lakers (or probably the Hawks, Hornets, or Pistons) would have been able to land MJ – they’d immediately have become the best team in the league. The team that had Michael Jordan would win the championship, so not only did you need him to that you could win – having him was basically the only way you possibly could win!

(not comparing Lincecum, but I’m just saying)

rwperu34
14 years ago

Derek,

A point is a point, regardless of where it comes from…if you have set the correct baseline.

Most valuation systems set the baseline at worst starter. In any league that would consider itself standard, the waiver wire is going to produce players better than the worst starter, especially on the pitching side and even more so in a holds league. By setting the baseline too low, you need to have an MORP type of valuation (ie big superstar premium) as opposed to a WAR type valuation system. If the baseline is correct, keeping Lincecum’s points off of another team is no different than keeping any points off of another team.

Now, even with the correct baseline, there might be a little bit of a superstar premium in keeper leagues, but that has to do with the secondary trade market for rebuilding teams. In a redraft league, there should be no superstar premium if the baseline is set correctly.

Derek Ambrosino
14 years ago

But that’s because there’s no Michael Jordan among fantasy baseball’s starting pitchers. And, that’s why my question is largely rhetorical. If you play fantasy basketball, for example, you have a different story. Lebron James is the number 1 ranked player, followed by Kevin Durant (Chris Paul has been injured). Hold that for a second, while I jump back to baseball

Now, I would be willing to bet that if you were to look at all the 12 team leagues in fantasy baseball and see where the winning team picked in the first round, you’d find that having the first pick isn’t particularly predictive of a league champion. I’d presume that league champs would emerge with relatively equal frequency from every draft slot.

But, in fantasy basketball, Lebron James is so far and away more valuable than the next best player that being ranked #1 vs #2 doesn’t even begin to encapsulate the advantage of the #1 pick. There’s a major talent drop off, not just within the first round, but after the very first pick! I’d guess that having Lebron James on your roster would be more strongly correlated with winning a fantasy basketball league than having any player would be to winning a baseball league. …And, not just first rounders – because we know that the players who are most strongly correlated to fantasy baseball championships are not the first rounders; they’re the Mark Reynoldses, the guys who come out of nowhere to be superstuds. That speaks to the parity of elite talent within the first round of fantasy baseball drafts.

In fantasy basketball – Lebron is that guy where there is huge value in keeping him from any other team. If I had the option of trading my second round pick for an extra fourth round pick before a draft even started in exchange for getting the #1 pick overall in a fantasy basketball draft, I’d do it in a heartbeat. That is to say that I would rather have Lebron, no 2nd round pick, 3rd round pick, and 2 picks in round 4 (or maybe even round 5 or 6) than have the 2nd pick, and then a pick in each round.

That dynamic exists – but Lincecum hasn’t sufficiently lapped the field to activate it. For an extreme example, wouldn’t Babe Ruth have basically made fantasy baseball an illegitimate exercise for the 1921 season by scoring in 33% more runs, clubbing 150% for homers, and driving in 30% more runs than his closest competitors, while finishing third in hitting and stealing only 18 fewer bases than the league leader?

rwperu34
14 years ago

So how much of a $200 budget would you pay for LeBron? How much should you pay? $100? $120? $150? There is tipping point where spending $x on Lebron gives you less of a chance of winning than spreading that money around over several players. It’s going to come down to (pts)*($/pt). 

Snake drafts are a whole different ball of wax. You take the best available player that fills a need for your team. That’s the same for the first pick of the first round and the last pick of the last round. In that case, it doesn’t matter where you set your baseline. It’s auctions, where you have the choice of how you distribute your money where supply and demand really comes into play.