Aren’t hybrids all the rage?

One of my favorite fantasy offseason pastimes is to think about various potential league structures. An economist will argue that people’s behaviors are largely determined by incentives. So it stands to reason that the way you build incentives into a system will influence both a party’s willingness to enter such a system and the behavior he/she will display once engaged. I like to tinker with these systems and think about ways to promote healthy, competitive fantasy behavior. This week, I present for your consideration a high stakes H2H-roto hybrid with both weekly and end-of-season payouts. I’ll explain how I conceive it working shortly, but first let me briefly mention what the design is meant to achieve.

This design intends to accomplish two things. First, the weekly payout system is an attempt to proactively discourage deadbeating. A team that is statistically “out of it” can still recoup “units” on a weekly basis. Deadbeating is a problem in every league. Some might think that simply making a league high stakes takes care of deadbeating, either because of the size of the prize or the selection bias of entry, but such is not the case. Some may think that high-caliber owners never deadbeat, but my take is that deadbeating is not just an issue of absolutes, but one of degree. I’ve never deadbeated in my life, but I’ve occasionally been non-competitive in leagues, and quite frankly when in this situation, I’m just not as aggressive and attentive as I normally am. Maybe an owner will rotate players regularly, but not really monitor the wire with usual vigilance, for example. Weekly payouts are an attempt to incentivize being as engaged as possible at all times.

The second thing this design attempts to accomplish is to up the willingness of participants to play for higher stakes. In my oldest league, the league is about two-thirds deep with really high-quality owners. The remaining guys are certainly not pushovers, but they rarely win. By most standards, our entry fee (which is tuned in four-year cycles, escalating each year and resetting in the fifth, in correlation with the end-of-keeper cycles and a full redraft) is pretty high. Some of us want to bump it up even further, but we are often unable to push those increases through at draft time given that we need near unanimous agreement on entry fee set-up or amendments. The potential of having additional ways to win money often appeals to those who don’t want to put a big enchilada on the line on a 25-week tournament of merit. Consolation prizes make for good carrots.

So, here’s how I would see this league working. The league is scored in a H2H format with weekly scoring periods. Owners compete against each other directly in the standings, but compete against the whole field for the weekly scoring prizes. The best individual team total in each category each week earns a small payout. In other words, the weekly categorical prizes are awarded on a roto model, but the season standings are calculated on a H2H model. There are no ties; the would-be prize for all pushes gets rolled into the regular-season champion payout. There are no weekly payouts in the playoffs.

Of course, to fund weekly payouts you’d need considerable entry fees from each participant. Here’s how the league might work at a 250 signaro entry fee:

Teams: 12
Entry: 250
Total Pot: 3,000
Categories: 12
Per category weekly payout: 5
Total weeks (non-playoff): 22
Total category payouts: 1320 (minus roll-overs for ties)
Regular season champ and runner up payouts: 450, 230 (plus roll-over ties, split evenly)
Postseason champ and runner up payouts: 650, 350

Basically, I just chose to pay each category out at 5 units per week, and then divided the remaining pot into five shares, dumping two or them into the regular-season payout (since they’ll also get the roll-overs) and the other three into the playoff pool. I then simply split those shares two to one.

Obviously, there are myriad ways to amend this. You could raise the weekly payout to 6 per category. You can play with few categories for higher value or more categories for lesser value. You can award the regular- and post-season finishers with different weights. You can pay out third place in either or both the regular and postseason.

I will issue a word of caution about tweaking though. I did not pull the numbers I used above out of thin air. When you create a new model, you have to think through how people might try to beat that model, where that model is susceptible to exploitation. For example, it is fairly easy to build a team that will win certain categories outright almost, if not every, week. The easiest way to do this would be to build a team that dominates saves and steals. Note that in the above model, even an owner who succeeds absolutely at such an attempt will not even recoup his/her entry fee.

Remember, setting up systems is all about incentives. You want to create a system that incentivizes the league to generally play the same way, for the same thing. Sure, in a given week a team may choose to just chase a weekly payout or two, but that is rational and isn’t corrosive to the system as a whole. Under the proposed system and payout structure, each owner is encouraged to compete the same way they would without the weekly payouts because the real profitability is in winning long term.

One minor problem with this design is that it requires some manual bookkeeping. Really, we’re talking about 5 minutes a week here. Set up a spreadsheet with each of the teams, look through the match-ups at week’s end and record the best overall totals in each category. If you like this idea but don’t want to be the bookkeeper, you could either offer a small credit to the bookkeeper by taking, say, 50 units out of the pot to compensate him/her or you could stipulate that whoever volunteers to do the bookkeeping gets to choose his draft slot (if you are doing a draft league). Another interesting idea would be to declare that the payouts from the in-season ties go to the bookkeeper. I’m not sure how many ties you’d get, so I don’t know what this ties future would actually be worth.

I’m interested in your takes on this model (it should work as an auction or a draft). I’m considering trying turn one of my leagues into this type of design (I doubt I’ll get up to 250 entry fee in its first year, but I am looking for a way to raise the stakes in that league). Also, since I get ideas for unconventional league designs fairly frequently, let me know if you find it worthwhile for me to write them up for discussion.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: BOB:  Twins and Rays stadium news
Next: 2011 Top 10 prospects: New York Yankees and Boston Red Sox »

Comments

  1. Millsy said...

    Derek,

    Sounds eerily familiar to something I have been designing and writing about at FBJ this past year.  Check out my model at my own site here:

    http://princeofslides.blogspot.com/2009/08/dynasty-rules-under-revenue-model.html

    I went into a similar discussion at Fantasy Ball Junkie.

    We’re thinking about putting the league in action, and I’d like to have some solid competition.  Shoot me an email if you’re interested.  I have revamped the rules somewhat, added some fun strategical portions, and I can email you the rules if you really are interested.

  2. Millsy said...

    PS-The roto AND H2H model sounds like an interesting strategical point…might make for some interesting decision making for winning this week or taking the Roto scoring at the end…

  3. Jason said...

    Best league I’ve ever played in is Rotisserie.
    Points are as follows:

    TB: 1 pt per base
    R: 1 pt
    RBI: 1 pt
    SB: 1 pt

    W: 18 pt
    SA: 8 pt
    K: 1 pt
    CG: 5 pt
    SHO: 5 pt
    NOHIT: 10 pt

    Bi-Weekly payouts.
    All points are reset at the All-Star break, making for 2 separate seasons: 1st half and 2nd half.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>