When to use the waiver wire

With draft season over and many owners not looking to trade yet, the waiver wire currently dominates the attention of most fantasy owners. However, there are different, conflicting philosophies regarding proper use of the waiver wire, even among intelligent fantasy players and experts. Today, I thought I’d give you my thoughts on the matter.

Different philosophies

Some feel that it is best to be aggressive on the waiver wire throughout April, taking a “shoot first, ask questions later” approach. If a player gets off to a hot start with the peripheral skills to back it up, pick him up and see if he continues. If he doesn’t, drop him later.

The problem with this line of thought is that in April (and through much of the first half of the year, honestly), sample sizes are far too small to make any kind of reasonable judgment based off of them. Surely we’d be better off judging a player by the hundreds of previous at-bats he’s taken as opposed to a handful in April, no?

Still, every year at least a couple of these April surgers outperform projections and turn in great fantasy seasons. Just one year ago, guys like Carlos Quentin, Ryan Ludwick, Cliff Lee, and Edinson Volquez led many owners to a fantasy championship. Had we waited until their numbers stabilized, they would be long gone from the wire and leading the charge for someone else’s team.

In a recent conversation I had with an esteemed fantasy analyst, this analyst expressed the idea that many fantasy players get too caught up in small sample sizes and anecdotal evidence, stating that it is the job of the good fantasy analyst to “drive perceptions back to long-term trends.” He said that he didn’t understand people who claim “if he regresses, I’ll drop him later,” because at that point you’ve eaten three rotten weeks. You would never listen to a financial adviser who says, “If the stock falls after you buy it, you can sell!” Put that way, it seems a little silly that we would even consider such a notion.

My take for ‘first-come-first-served’ leagues

Here’s why I feel differently, though. Baseball players, as much as we may treat them like stocks, are not stocks. They are different and follow a separate set of rules.

The primary cost of a stock is the acquisition cost. You pay a certain amount of money, you acquire the stock, and your return or loss is fully dependent upon its future performance.

In fantasy, however, you can hedge against (or eliminate) the risk of uncertain future performance in your waiver wire pickups. In fantasy, the primary cost is not the acquisition cost (unless you’re in a FAAB league or must use a high waiver claim—these are leagues I’ll discuss next time) but, rather, the activation cost.

In fantasy, the only acquisition cost is the worth of a single roster spot (most often the last spot on your bench). You can bench your pickup for three weeks to see how he does without incurring a single additional cost. Only once you activate him do you start to incur additional risk and additional costs (should he fail).

In the stock market, you can’t pick a stock and say that you’re going to watch it for three weeks before deciding whether or not you wanted to buy it at the original price. In fantasy, however, we can do just this by sitting the player on our bench.

In addition, the last player on your bench will often be an unspectacular veteran player worth only a couple dollars. The cost of losing this player is small because 1) there are likely similar, interchangeable players available on the wire to grab in case your high-upside pickup flops and 2) the high-upside pickup’s value might not be much worse than this player’s value anyway.

Potential scenario

While picking up a player for your bench sounds like a great idea, in some leagues you don’t have a bench or simply don’t have room on it. If, for whatever reason, you must start the pickup, it still might be worth it to take the gamble. Take a look at the following scenario.

Let’s say we pick up Cameron Maybin, who THT has projected for -$7 in value over the entire season. We decide to keep him for three weeks and then drop him if he isn’t tearing it up. If the player he will be replacing is valued at $3 over the entire season (a reasonable estimate for the last player on your roster), that amounts to a $1.20 net loss in value.

Now let’s set up a hypothetical scenario in which Maybin has just two distinct possibilities: playing to his -$7 projection or becoming a $25 player (value dispersed evenly throughout the season). In this case, Maybin would need to play up to his $25 value five times out of 100 in order to break even on the pickup.

So the question then becomes, do 5 percent of April pickups post $25 seasons? This question would take a more in-depth study than I have time for now, but I would imagine the answer is a resounding “yes” if we only include players who are showing marked skill increases (after all, if a player isn’t doing this, he’s just getting lucky and shouldn’t even be considered).

Caveats

The scenario I depicted above is just one possible scenario that was simplified a great deal. For your fantasy team, it isn’t enough to simply use that 5 percent figure and make pickups based off of it. There are a number of factors that really need to be considered:

  • Projected value of pickup
  • Projected value of player to be replaced
  • Potential upside (and downside) of pickup (and this will not be a single value — it will be a numerous values, the probability of each usually forming a rough normal curve)
  • Potential upside (and downside) of player to be replaced
  • Active or bench spot
  • How long you will keep the player before giving up and dropping him
  • Availability of solid, unspectacular free agents should your pickup flop
  • Acquisition cost — is it a simple pickup or does your league require FAAB bidding or waiver claims? (more on this next time)
  • Opportunity cost of passing on other free agents (which can be ignored, assuming you’re picking up the one with the highest upside)
  • Potential trades should the pickup post another couple good weeks

To help you decide the necessary probability for your own pickup, I’ve put together a (very) simplified calculator that will run you through my scenario above and give you a rough estimate of the necessary odds needed to make a pickup worthwhile. You can download it by clicking here. Keep in mind this is for leagues that would require the pickup to be active from day one. In leagues where the player can be benched, the needed odds would be much lower.

A Hardball Times Update
Goodbye for now.

Concluding thoughts

Overall, I think it is wise to make frequent use of the waiver wire in the early portions of the year. Of course, every league is different and every team is different, and in some situations this will not be the best strategy. You should examine your own specific situation and then decide upon the best course of action.

In leagues that allow you to bench your pickups, however, it is almost always a good idea to be aggressive early if you have an easily replaceable player that can be dropped. The cost is very small and the reward is huge.

Next time, I’ll talk a little about handling the waiver wire in leagues with FAAB bidding.


4 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Ed Schwehm
14 years ago

Pretty nifty calculator! And good analysis as well. Of course, some leagues have maximum transactions per week or per season, so owners would have to take that into account too.

Also, the other reason fantasy pickups are not like stock is that these waiver-wire acquisitions are likely to sit on your bench most of the time, not affecting your actual stats.

Unless you had a really good draft, the worst guy on your team is almost by definition a replacement level player for your league. At this time of the season, you can use the last draft round as your replacement level gauge. Any player at that level or worse is expendable. I find I really have to fight the endowment effect hard even for my Mr. Irrelevant.

Will
14 years ago

“In the stock market, you can’t pick a stock and say that you’re going to watch it for three weeks before deciding whether or not you wanted to buy it at the original price. In fantasy, however, we can do just this by sitting the player on our bench.”

Sure you can; it’s called a “call option.”

Starting players = stocks
bench players = call options

Therefore, only drop bench players for speculative, high upside guys.

Derek Carty
14 years ago

Thanks, Ameer.  That would certainly be something to look at, although I should note that some (Tom Tango, for one, here: http://www.insidethebook.com/ee/index.php/site/comments/matt_wieters_and_mles/) aren’t convinced that PECOTA percentile projections actually work.  I do believe it is worth a look, though, and Sean Smith’s CHONE system also does percentiles.

Ed,
That’s something I always seem to forget to mention.  I’ve never played in a league with a transactions cap and it always slips my mind.  That is definitely something that needs to be considered if the maximum is a low number.

You’re right on with the other points.  I think I mentioned or alluded to them in the article.

Will,
I’m not much of a stock trader (instead focusing my brain power on fantasy baseball smile ), but looking into call options a bit that does seem to be a solid parallel.  You’re statement of “only drop bench players for speculative, high upside guys” is essentially my stance (unless, of course, that bench player also has a high upside).

Ameer
14 years ago

Useful stuff!  I just wanted to mention that combining the results with pecota’s percentiles might make this even more useful.  (ie in the example, does Maybin have a 5% chance of posting a $25 season?  We can look at his 90th percentile projection in pecota and get a pretty got idea).