Wednesday, February 25, 2009

The FSL Contest Hog Issue

The post below is from the FSL Blog. I am also posting it here, because it shows an interesting contrast between online poker and daily fantasy sports.

Over at the FSL Forum there has been a lot of talk about the issue of contests hogs. We have also received a handful of email complaints about the same issue. I thought it would be a good idea to go over our thoughts in depth about this issue in one place. So what exactly is a “Contest Hog”? Basically a contest hog is someone who enters a large number of fantasy leagues every day. Because it just takes a couple of minutes to draft a team at Fantasy Sports Live, it is pretty easy to enter 10s of contests a day. In general, we call these types of users volume players, and they are our best customers. They help to create a lot of action on the site, which helps us to attract new users. There are two things that volume users (contest hogs), sometimes do that can hurt business at FSL.

Blanketing The Contests

This is where one user enters every contest for a given sport and re-enters any empty contests that reform. They basically are not giving anyone a choice at all of who they play against because they are literally in every contest. This has been attempted a few times in the past for brief periods, but we have always asked the user to stop, and they have complied. In general this is a very high variance move any way. If you are risking say ½ of your bankroll to cover every contest, even with an edge you will go bust in a very short amount of time.

Preying On New Players

This is where a user is constantly watching the website, and will fill up any heads up contest with one entry no matter how low the stakes are. The problem with this approach is that you need a lot of time on your hands to constantly be watching the web site, and you can only win so much money anyway at the lowest stakes. New users will feel like every time they enter a contest, it gets filled by the same guy, and he happens to be near the top of the all-time leaderboard. This type of behavior is what the recent complaints are about.

Before I go into what we are already doing to address this issue, I wanted to go over why this will always be an issue for daily fantasy leagues that use a salary cap type draft. It gets down to the concept of leverage. Many people will notice that our site does not look like the most of the other fantasy sports sites out there. Our site it based on what is called a “Sit & Go” at an online poker site. A Sit & Go is a small fixed sized tournament that forms in real time and starts running when it fills up. A good poker player can be profitable playing these long-term, and may chose to move up to higher stakes (where the competition is tougher), or play more than one at a time to leverage their skill advantage. In online poker, the leverage you can apply is greatly limited. If you move up to higher stakes, your win rates will go down because of the tougher competition. If you play several S&Gs at a time at lower stakes, you do not have as much time to make decisions, and you lose your feel and reads for the table lowering your win rate. So the amount of leverage you can apply in online poker is greatly limited. You will eventually find high stakes that you can’t beat, or you will find that playing an additional S&G, actually lowers your hourly win rate. In sports betting your leverage is limited as well. There are maximums you can bet on a game, and they will start moving the line against you if you start betting a ton on one side. By contrast, in fantasy sports the leverage you can apply can be nearly infinite. Once you draft your team for a daily league you are done. Joining another league does not reduce your chances in the first one. If you can’t beat the highest levels on the site, you can focus on the lower levels and leverage your advantage there to a great extent. If you can draft your team in 2 minutes, you can enter 30 leagues an hour. For one hour of work you can 30x your leverage. What you win on average over the 30 leagues is your hourly win rate. As long as spending 2 minutes to draft another team, increases your hourly win rate, it is worth your time. So as long as you have a fast draft interface (like we do) you are inviting players to leverage their skill advantage. A live draft that took 30+ minutes would solve this issue, but that is not really what our customers want. So we have a trade off between customers wanting a quick and easy way to play fantasy sports, and skillful players exploiting a quick draft to leverage their skill advantage. We have been aware of this issue from almost the beginning in June 2007. There are several things that we already do to address this issue.

What do we currently do to discourage / prevent this type of behavior? discourages this type of behavior, in several ways based on our rules and how we operate. When this is not enough, we contact the player and ask them to voluntarily change what they are doing. For the players that were “Blanketing the Contests”, they have always complied with our requests. For the other type, we have also had good luck in the past, but still receive some complaints. The current user being discussed in the forum really only enters about 10 contests a day out of about 60 basketball leagues that we run, and they tend to be at the higher stakes. This is enough compliance in our opinion, though some users to not think it is enough. It is in our volume user’s best interests to comply with our requests, because in the long-term, anything that helps out FSL helps out our volume users. They will get more people to play against, and higher stakes leagues as well. This is why we currently have stuck with this approach. The following additional measures are already in place on the site.

Drafting is not instant at FSL

We do not allow you to enter a second contest with the same team without redrafting your team. Since redrafting a team takes time this discourages volume users from focusing on the lower entry fee leagues where less money can be made. Other competitive sites let you enter additional leagues instantly with a single click which encourages contest hog behavior.

We offer varied salary cap structures and amounts

We offer a wide variety of salary cap types and amounts on our site. Each one requires a thought process on how to draft the most effective team. After figuring out a team for one structure, a contest hog will need to spend the time to figure out the other structures before hogging those leagues as well. Other competitive sites only have a single structure, making it much easier to hog every league on the site.

We offer higher payouts at the higher dollar entry fee amounts

This is a big magnet for the volume user (contest hog). Volume users are painfully aware of how much we payout because it is huge to their bottom line. If we payout 91% at one level, and 93% at another level, they will simply move up to the better value. Not only can more be made at the higher entry fees, but more of the total fees are paid out in prizes as well. This is how we try to keep out best users out of the lower dollar contests. In general, it works pretty well. Our competition does not offer these types of incentives to stay out of the lower dollar leagues.

We pay out multiple places in multiplayer leagues

Our 6 player leagues pay out the top 2, our 10 player leagues pay out the top 3. Even if a good player is “hogging” the multiplayer leagues, you do not have to beat them to get paid. At lot of competitive sites are “winner take all”, meaning you will have to always beat the contest hog to get any prize.

We do not conceal the competition

This is how users may always avoid contest hogs at FSL. If you are the first one to enter a contest, you are basically saying “I am willing to challenge anyone who is currently viewing the site”. When you are the last one to fill a contest you are saying “I am willing to challenge the other users already registered for this league”. It is obviously much riskier to be the first in versus the last in. If you avoid entering contests first, and only fill contests as the last entry, you can completely avoid any contest hogs. As I said above, the hog in question really only enters about one out of six leagues on average. Plenty of users are successfully avoiding him. Many of our competitors conceal the identity of registered entries in some way or another.

We offer $1 100% payout 10 player leagues

For the benefit of our new users we offer 10 player leagues that payout at least 100% of the entry fees, and also pay the top 3. Even is a contest hog were to enter one of these, they are still a better overall value than a league with a standard payout. Also, with the top three paid, you don’t even need to beat the contest hog. This will always be a safe place for new users to compete.

We don’t allow running leagues to be viewed by people who are not entered

This probably hurts us a bit when it comes down to signing up new users, but it serves a purpose. The idea is that you need to pay to see your opponent’s fantasy team. If you want to know the types of teams a top player is selecting, you will need to move up to their level. When a top player moves down to the lowest levels, they are inexpensively giving their opponents information about the types of fantasy teams that they draft. This helps the lower level competition improve there skills to where they may be able to beat the better player when they move up. This is intended to discourage top players from playing at the lowest levels. They are better off to let the newer players move up to their level before revealing how they play. Otherwise there will be less money to make when new players move up. It does not make a bunch of sense to crush the $3.40 leagues, if it will hurt your chances in the $107 leagues.

Other potential future measures

We came up with a concept of a software fix for this over a year ago. Most “contest hogs” since then have voluntarily complied, and we have not had to roll it out. Any software fix (rule change) would need to be applied equally to everyone. Any restrictions on “contest hogs” would also restrict all users in the same way. We are very reluctant to do this, but it is in our back pocket just in case. The other way this issue goes away is to simply grow the traffic on the website. Let’s say we started running 300 basketball leagues a day (should be there by next season). Even if a hog were to now play 20 leagues a day, they will be entering a much smaller percentage of the leagues. Bankroll issues also come into play. It already takes a pretty large bankroll to try to hog the contests today. With increased traffic the bankroll requirement keeps increasing. Also with increased traffic you are more exposed to better players than yourself. If you are putting ½ of your bankroll at risk every day, and a better player than yourself shows up, you will lose everything in 2 days. So currently rather than spending money to put in a software fix that will reduce the number of leagues we run, we choose to spend that money advertising the site to grow the number of leagues that we run. If we can grow the site fast enough this issue should go away pretty soon.

Now that you know in detail what the issue is and what we are doing to address it, are there any comments or suggestions on our current approach to the issue?

Labels: ,

Thursday, February 19, 2009

The Streak

My Mookie cash streak ended at two last night. It was also an end to a streak in 2009 that had me cashing in every S&G and MTT I have entered so far this year. Well 3 Mookies is all for the MTTs so far, but I am up into the 30s of beginner S&Gs at Bodog and just can't find a way not to cash. Now that I have bragged about it here, I am sure it is over. Poker Gods are you still reading this garbage?


The Blinders Housing Fix

Let me first say that the housing market is not broken, so nothing really needs fixing. House prices are falling which basically shows that the market is working. There is a boatload of supply out there, and demand is constrained by the economy and tighter credit standards. If house prices were going up right now you could claim that the market was broken. So house prices need to drop to where supply and demand can balance out. The only thing that can really be done is to try to prevent the values from overshooting to the downside, which could hurt the economy even further. So a fix to the housing problem will not prevent housing prices from falling, but may soften the landing a bit.

So this is what I propose, since we love to throw billions around like it is nothing. Have the U.S. loan me 10 Billion at a super low interest rate (say 4%). It does not have to be me, it could be anyone (just private sector). What I will do with the money, is go out and buy the most distressed properties in the areas with the highest level of foreclosures. I will hire a ton of people to rehabilitate them, and then I will rent them out for 5 years minimum until the housing market improves. At that point I can choose to continue to rent them out, or to sell.

So how does this help?

It takes the distressed low end completely off of the market. This will drive down inventory a ton and stabilize prices. Figuring you could go in and buy at 10% below current market prices with a low fixed interest rate, you would be profitable at pretty low rents compared to the current mortgage payments. You will be hiring people to find, buy, rehabilitate, and manage the rentals which will stimulate the economy. The government just has to provide the loans, and will earn interest. Unless the housing market is down for 10 more years, the loans should all get paid back. You could even take distressed mortgages off the market, by buying the property from the bank at 10% below market, and renting it back out to the original borrower at a much lower monthly amount.

So with the above plan, we can help stabilize housing prices, and create a bunch of jobs without costing a dime to the government if all goes well. Based on that, it will obviously never happen.


Thursday, February 12, 2009

Fold, Hope, and Mookie

I have been trying to think of a better name for my tournament strategy besides “the Black and White Strategy”.  Last night during the Mookie, I came up with “Fold and Hope”.  It seems kind of perfect in a way.  Long term stock investment used to be the norm.  You would buy some stock, and hold it indefinitely.  You were not sure exactly when it was going to go up in value, you just know that given time eventually the right factors will stack up in the economy for your stock and you will get your gains.  This strategy was ridiculed during the dot com market collapse in 2000.  The term “Hold and Hope” was used to make fun of long term stock investors.  It was more about timing the market now.  Tactically jumping in and out at the right moments, and always looking for any edge you can find and leveraging that edge to the max.  So think about “Fold and Hope”.  What you are dong is like long term stock investing.  You are holding on to your capital for some unknown time in the future when you will get your gains.  You have no idea when they are coming, you just know that you have picked some good stocks (your over all poker strategy), and that holding them will work given enough time.  So in an MTT, you do a lot of folding early and in the middle, and you hope that you will get some goods cards and good situations at some point in the future to make your patience pay off.  The early aggressive players are more like day traders. They don’t trust the market long term. They are active and jumping in and out of pots.  They might make a big score, or they might go bust early on.  They are more about timing the market for a big score now then trusting the market over time.  The Fold and Hoper trusts his stocks (game), and the market (mathematics of MTTs) and lets the chips fall where they may. 

My run through the Mookie this time was a bit tougher.  I did not get an early double-up to go into cruise control.  I stayed around +/-800 of my starting stack for the first 90 minutes or so.  By that point I was way short of the average stack and near the bottom of the leader board.  I was folding and hoping for better times in the future.

I finally started catching a few cards and built my stack up to 8-9k and then was gifted the chip lead in the hand below.  I had really not been stealing at all up to this point, and made a standard 3x "steal" raise from the button with QQ.  The table chip leader shoved from the SB, and I thought briefly before calling.  No way I can put him on AA or KK there, yet I may actually have something.  I had to dodge a flopped gutshot, but did for the chip lead with about 20k and 15 people left.    

Fast forward to the Final Table and we are 5 handed, and I pick up AA UTG.

Full Tilt Poker Game #10580755891: The Mookie (78916851), Table 2 - 800/1600 Ante 200 - No Limit Hold'em - 1:11:38 ET - 2009/02/12
Seat 4: lightning36 (42,743)
Seat 5: BuddyDank (9,531)
Seat 6: NYRambler (43,344)
Seat 8: JoeSpeaker (22,758)
Seat 9: Blinders (25,624)
NYRambler posts the small blind of 800
JoeSpeaker posts the big blind of 1,600
The button is in seat #5
*** HOLE CARDS ***
Dealt to Blinders [Ah Ac]
Blinders calls 1,600

Don't try this one at home.  5 handed is a bit dangerous for this move, but whenever I pick up AA late in an MTT I am looking for a full double-up if that is what's needed at the time.  I am willing to roll the dice a bit to get there if it will improve my chances enough.  Here as tight as I had been playing the UTG limp seems OK.  I figure worst case both blinds call.  Best case someone raises.

lightning36 folds
BuddyDank folds
NYRambler calls 800
JoeSpeaker checks

worst case senario

*** FLOP *** [Kc 4d 5d]
NYRambler checks
JoeSpeaker checks
Blinders bets 3,200

I can't screw around any more here

NYRambler calls 3,200
JoeSpeaker has 15 seconds left to act
JoeSpeaker raises to 20,958, and is all in

his was a bit scary.  I figure I have Rambler beat, but Speaker would need something strong to check raise shove there over a bet and a call.  I figured it was just a strong K or a draw.
Blinders calls 17,758
NYRambler folds
JoeSpeaker shows [Js Kh]
Blinders shows [Ah Ac]
*** TURN *** [Kc 4d 5d] [9h]
*** RIVER *** [Kc 4d 5d 9h] [5s]
JoeSpeaker shows two pair, Kings and Fives
Blinders shows two pair, Aces and Fives
Blinders wins the pot (50,916) with two pair, Aces and Fives

After that I knocked out Buddy, and eventually got heads up with NYRambler.  I had a nice chip lead to start heads-up, and was playing hyper-aggressive which was probably a shock to anyone who had watched me early on.  I lost the chip lead briefly at one point, but rallied right back to the lead before looking down at QQ.  We would get it all in preflop, and if I could fade an Ace the Mookie was mine.

That hurt, but I was still in decent shape.  A few hands later I would get AJs, and again we would get it all in preflop.

This time I would survive the flop, but get out turned on a 119k pot.  IGHN.  GG NYRambler.


Monday, February 02, 2009

Cha Ching!

Thank you Zona!  I made a decent killing on a the Superbowl.  I would have made a ton more, but I am stuck with retarded Bodog as my only sports book.  I almost headed out to Vegas at the last minute to get some better lines to chose from.  Bodog opened the game at Pittsburgh -7.  By game time you needed to lay -130 to get Arizona +7.  I would have loved to take Arizona -6.5 instead, but it was not possible with Bodog.  So I was a bit limited by cash on hand in my Bodog account, and being stuck with their crappy line and moneyline odds.  Ultimately, I would get my entire Bodog balance in action, and came close to doubling it up.  I did not get the moneyline win, but I had Arizona and the over parlayed which offset that to some extent.

The game went pretty much as I had predicted, with Pittsburgh's offense not being able to put up enough points to cover a 7 point spread.  The key play being the INT returned for a TD at the end of the first half.  That was a 10 or a 14 point swing to the spread, and a play like that can absolutely crush your chances.  Arizona was game though, and shook it off to take a late lead that locked down their side against the spread.  I also thought that Arizona's decision to defer the opening kickoff, was a huge mistake.  Not sure what they were thinking there.  Put your strength on the field first, and grab some early momentum.  Deciding to kick-off first, gave Pittsburgh the early momentum for the entire 1st quarter, and made it a much tougher game to win.  The zebras seemed to be favoring Pitt as usual.  Three bad calls were overturned in Zona's favor.  They missed an excessive celebration call on Pitts final TD that was obvious and would have been huge for Zona's field position.  They also did not even review the fumble on Warner's last play that was questionable at best.

Overall, I say Arizona covers the spread in that game 8 or 9 times out of 10 tries.  Looking at the boxscore, Arizona showed up with the better defense than Pitt's so called #1 D, and were huge on the run defense as expected.  Props to Big Ben for stepping up and getting it done late. Also huge props to Warner, who should be in the hall of fame now.  He has the best three Superbowl performances in history by a QB.  Its not really his fault that he has had to do it all himself without much help.  He made one mistake the entire game, (a huge one), but still brought his team back for a chance at a win.

Hope everyone enjoys their second farm!