## Comparing CoinRoster Prices to Market Odds

This post is about trying to beat the odds on CoinRoster daily fantasy golf, a game where the fantasy points are determined based on score to par. The worst score to par becomes zero and each fantasy point is awarded based on the difference from the worst score. For example, if the last placed player has a real world score to par of +4 and the first placed player has a score to par of -10, then the first placed player will have 14 fantasy points. All other players are assigned a relative point value.  Six players are drafted to each roster, so the total score of each roster is the sum of the individual scores of all players on each roster.

My goal is to find some drafting strategies that take advantage of the difference between the price of golfers available to draft on CoinRoster versus the odds of betting on golfers to win the tournament.

The price of each player in the CoinRoster draft is relative to their world golf ranking. But I wondered how to determine whether this is a good price or a bad one? And which factors should I consider while attempting to make this determination?

I have compared the prices of the players available to draft with their odds of winning the tournament. I know from previous research that the odds of each player is related to their expected relative score to par.

I made a list of all players available to draft using excel with their associated price on CoinRoster, then I calculated the percentage chance implied by the prices of each players in the draft by taking their drafting price divided by the sum of all player prices. So if a player costs \$3000 to draft and the sum of all players prices is \$80,000, then this implies 3.75% (\$3,000 / \$80,000).

By finding this drafting price ratio, it gives me an “implied” percentage chance of each player having the most fantasy points based on CoinRoster’s drafting price. Then for curiosity’s sake, I converted those implied percentage numbers into decimal odds. The next thing I did was add a column of each player’s odds of winning the tournament and then I converted that into a percentage number also. Now I can compare apples to apples using their percentage chances.

With this information, I can eyeball and compare the odds numbers and the percentage numbers (which are the same value, but just displayed in different formats). I can also rank and sort the data in various ways using excel, so I subtracted the implied market percentage chance from the implied price percentage chance to uncover the difference between the two numbers.

By subtracting the two prices (the percentage chance implied by drafting price with the percentage chance implied in the player’s betting odds) I can notice if the price for a player available to draft is relatively lower on CoinRoster than is implied by their market odds. If the player is cheap and represents good value to draft their price on CoinRoster will be lower than implied by their betting odds. This means the player is cheap to draft, but if a player’s market odds are lower than their relative drafting price on CoinRoster, the player is expensive.

The table below shows the numbers for the BMW Championship. The table is ranked best to worst value.  Dustin Johnson has the best value for this tournament according to this formula. His odds were 9.36 (which implies a more than 10% chance he will win the tournament, and thereby have the best score to par) but his CoinRoster price was relatively cheap (even though he is the second highest priced player). Justin Rose, Rory McIlroy, and Hideki Matsuyama are also some of the most relatively best priced players.

Tiger Woods also has the third highest relative value to draft. This makes sense since Tiger’s world golf ranking (which the CoinRoster drafting price is based on) is relatively low compared to his chances of winning the tournament. So intuitively, this affirms my belief that I’m on the right track.

Some of the most expensive players on CoinRoster are Tyrrell Hatton, Alex Noren, Brian Harman, Kevin Kisner, and Peter Uihlein.

 CoinRoster Price Implied % Implied Odds Market Odds Market % Diff % Cost Dustin Johnson USA \$2,839 0.0364 27 9.36 0.1068 -0.0705 \$9,410 Justin Rose ENG \$2,366 0.0303 33 15.70 0.0637 -0.0334 \$8,368 Tiger Woods USA \$979 0.0125 80 24.00 0.0417 -0.0291 \$8,060 Rory McIlroy NIR \$1,854 0.0238 42 19.00 0.0526 -0.0289 \$9,769 Hideki Matsuyama JPN \$1,372 0.0176 57 22.00 0.0455 -0.0279 \$10,915 Bryson DeChambeau USA \$1,797 0.0230 43 20.00 0.0500 -0.0270 \$10,995 Jordan Spieth USA \$2,058 0.0264 38 19.00 0.0526 -0.0263 \$10,794 Brooks Koepka USA \$2,688 0.0344 29 18.10 0.0552 -0.0208 \$9,548 Justin Thomas USA \$3,000 0.0384 26 17.60 0.0568 -0.0184 \$7,905 Tony Finau USA \$1,452 0.0186 54 32.00 0.0313 -0.0126 \$5,840 Jason Day AUS \$1,596 0.0204 49 32.00 0.0313 -0.0108 \$5,068 Adam Scott AUS \$812 0.0104 96 48.00 0.0208 -0.0104 \$5,377 Patrick Cantlay USA \$1,045 0.0134 75 48.00 0.0208 -0.0074 \$5,254 Cameron Smith AUS \$935 0.0120 83 57.00 0.0175 -0.0056 \$5,010 Brandt Snedeker USA \$680 0.0087 115 76.00 0.0132 -0.0044 \$6,305 Rickie Fowler USA \$1,905 0.0244 41 36.00 0.0278 -0.0034 \$7,128 Emiliano Grillo ARG \$689 0.0088 113 86.00 0.0116 -0.0028 \$6,426 C.T. Pan TPE \$801 0.0103 97 86.00 0.0116 -0.0014 \$6,418 Jon Rahm ESP \$2,230 0.0286 35 34.00 0.0294 -0.0008 \$6,776 Bubba Watson USA \$1,503 0.0193 52 52.00 0.0192 0.0000 \$5,441 Phil Mickelson USA \$1,203 0.0154 65 67.00 0.0149 0.0005 \$5,889 Billy Horschel USA \$681 0.0087 115 124.00 0.0081 0.0007 \$5,417 Henrik Stenson SWE \$1,159 0.0148 67 71.00 0.0141 0.0008 \$4,963 Gary Woodland USA \$895 0.0115 87 95.00 0.0105 0.0009 \$4,210 Tommy Fleetwood ENG \$1,951 0.0250 40 42.00 0.0238 0.0012 \$3,880 Aaron Wise USA \$731 0.0094 107 124.00 0.0081 0.0013 \$2,691 Scott Piercy USA \$227 0.0029 344 722.00 0.0014 0.0015 \$2,810 Chris Kirk USA \$406 0.0052 192 276.00 0.0036 0.0016 \$3,978 Beau Hossler USA \$565 0.0072 138 181.00 0.0055 0.0017 \$4,269 Zach Johnson USA \$762 0.0098 102 133.00 0.0075 0.0022 \$5,207 Louis Oosthuizen RSA \$850 0.0109 92 124.00 0.0081 0.0028 \$4,835 Marc Leishman AUS \$1,395 0.0179 56 67.00 0.0149 0.0029 \$4,387 Abraham Ancer MEX \$697 0.0089 112 171.00 0.0058 0.0031 \$3,448 Webb Simpson USA \$1,503 0.0193 52 62.00 0.0161 0.0031 \$3,436 Ryan Palmer USA \$390 0.0050 200 637.00 0.0016 0.0034 \$2,399 J.J. Spaun USA \$402 0.0051 194 589.00 0.0017 0.0035 \$2,496 Jason Kokrak USA \$456 0.0058 171 428.00 0.0023 0.0035 \$3,531 Keegan Bradley USA \$685 0.0088 114 200.00 0.0050 0.0038 \$3,544 Brian Gay USA \$466 0.0060 168 532.00 0.0019 0.0041 \$4,034 Austin Cook USA \$487 0.0062 160 475.00 0.0021 0.0041 \$4,078 Paul Casey ENG \$1,437 0.0184 54 71.00 0.0141 0.0043 \$4,322 Ryan Armour USA \$469 0.0060 166 646.00 0.0015 0.0045 \$3,725 Rafa Cabrera Bello ESP \$1,175 0.0151 66 95.00 0.0105 0.0045 \$3,863 Andrew Putnam USA \$510 0.0065 153 504.00 0.0020 0.0045 \$3,466 Adam Hadwin CAN \$731 0.0094 107 209.00 0.0048 0.0046 \$3,715 Byeong Hun An KOR \$840 0.0108 93 162.00 0.0062 0.0046 \$5,258 Charles Howell III USA \$607 0.0078 129 323.00 0.0031 0.0047 \$4,986 Luke List USA \$778 0.0100 100 190.00 0.0053 0.0047 \$4,898 Si Woo Kim KOR \$759 0.0097 103 209.00 0.0048 0.0049 \$4,671 Francesco Molinari ITA \$2,274 0.0291 34 42.00 0.0238 0.0053 \$4,537 Andrew Landry USA \$568 0.0073 137 551.00 0.0018 0.0055 \$2,823 Jr. Potter USA \$519 0.0066 150 893.00 0.0011 0.0055 \$3,463 Brendan Steele USA \$551 0.0071 142 675.00 0.0015 0.0056 \$3,593 Keith Mitchell USA \$625 0.0080 125 428.00 0.0023 0.0057 \$3,691 Patton Kizzire USA \$560 0.0072 139 808.00 0.0012 0.0059 \$4,175 Kyle Stanley USA \$1,208 0.0155 65 105.00 0.0095 0.0060 \$4,514 Chesson Hadley USA \$649 0.0083 120 428.00 0.0023 0.0060 \$4,047 Brice Garnett USA \$649 0.0083 120 561.00 0.0018 0.0065 \$5,113 Ian Poulter ENG \$1,109 0.0142 70 133.00 0.0075 0.0067 \$5,284 Kevin Na USA \$899 0.0115 87 228.00 0.0044 0.0071 \$5,017 Chez Reavie USA \$741 0.0095 105 475.00 0.0021 0.0074 \$5,532 Patrick Reed USA \$1,715 0.0220 46 71.00 0.0141 0.0079 \$5,928 Pat Perez USA \$820 0.0105 95 428.00 0.0023 0.0082 \$5,436 Daniel Berger USA \$842 0.0108 93 428.00 0.0023 0.0084 \$5,661 Xander Schauffele USA \$1,414 0.0181 55 114.00 0.0088 0.0093 \$6,399 Peter Uihlein USA \$1,137 0.0146 69 219.00 0.0046 0.0100 \$7,100 Kevin Kisner USA \$1,223 0.0157 64 181.00 0.0055 0.0101 \$5,963 Brian Harman USA \$1,045 0.0134 75 361.00 0.0028 0.0106 \$82,801 Alex Noren SWE \$1,580 0.0202 49 114.00 0.0088 0.0115 \$81,756 Tyrrell Hatton ENG \$2,115 0.0271 37 67.00 0.0149 0.0122 \$80,176 \$78,061 1.0000

## Opening a FanDuel Account

Today I’m going to open a FanDuel account. My plan is to make a deposit and play PGA and NHL fantasy.

The first step to to enter your e-mail address, username, and password. Once I enter these credentials, FanDuel takes me to a page where I’m asked to choose my favorite sport. In my case, this is golf (and it seems like FanDuel assumes I only have 1 favorite sport).

The first thing I’m noticing right away is its been a few minutes after opening my account and FanDuel seems intent on prompting me to enter a free contest (which I’m not interested in, I’d rather just play PGA for stakes). So I quickly choose a seemingly random lineup of players and enter the free signup contest.

After hastily entering a free contest, I’m now ready to make a deposit. I choose the \$100 amount and enter my credit card credentials. FanDuel users can also deposit using PayPal. But after trying a few different times, I was unable to make a deposit using either of these methods. I assume my credit card (visa) that could not be processed (as I tried paying with visa by way of my PayPal account also), I guess my card issuer might not process payment for FanDuel, I’m just guessing. But after I disabled the popup blocker from my browser and tried making a PayPal deposit debiting my bank account directly instead of my credit card, I was able to make a \$100 deposit and receive a \$20 deposit bonus.

## Gearing up for DFS hockey season

Starting with the 2018/19 NHL season, I’ll be posting my experience playing daily fantasy sports (DFS) on this blog. I’m planning on playing both NHL fantasy and PGA fantasy and I’ll be playing on multiple sites in order to compare my experience. I’ve only ever played fantasy sports casually as a fan and sometimes with friends, but this season my goal is to try to find an edge and look for ways to win. I’m not sure if this is even possible, especially considering the rake and competition, but I’m going to try.

I’m not sure yet how big my stakes will be and which sites I’ll be playing on specifically, but I’m sure I’ll try the biggest sites such as Draftkings & FanDuel, and I’m also hoping to learn more the smaller sites such as StarsDraft & FantasyDraft. I’m also going to compare and combine DFS with sports betting, and I’m also planning to investigate the legal side of DFS and sports betting.

## Free Play at CoinRoster Fantasy Sports

My friends are promoting their fantasy sports site, based on bitcoin, called CoinRoster.

As a special promotion, they are giving away 0.02 bitcoins worth of free play to any new user that signs up using a link from my page!

The bitcoins can be used to play on CoinRoster, but they can’t be withdrawn until they have been played 20 times, which in this case is 0.40 bitcoins worth of play.

For those of you who are unfamiliar with bitcoins, at the time of writing these 0.02 bitcoins are worth about \$40 Canadian dollars or \$25 US dollars.

I’d say the best games they currently offer are about Golf. They have a weekly golf fantasy game and also have a daily hockey game. In the summer they tell me they’ll have a baseball game going and in the fall they will do NFL.

## US Fantasy is Making Progress

Vic Salerno has been making steady progress with his pari-mutuel fantasy sports platform. He is reporting steady growth thru the NFL season and is also expanding the service to offer a wider range of bets. USF is also applying the pari-mutuel model to teams, not just players. Pick a group of teams to score an amount of points. If you’re watching college football or NFL all day, you can plarlay your bets using a pari-mutuel fantasy pool with USF.  There are some drawbacks for players to this type of wagering, but I expect USF to continue marching forward.

The pari-mutuel fantasy format is has another unique advantage of fitting in well with regulatory frameworks of many US states. If USF can gain acceptance with more US states, it would represent a huge growth opportunity. The first mover benefits for USF will end up being a huge advantage. The network effects caused by the most liquid pool is pronounced for par-mutuel betting.

USF’s model can be easily copied online, and the market is currently wide open for a bitcoin fantasy sports site to offer this product.

#### Businessman sees growth in pari-mutuel based fantasy sports

Vic Salerno’s bid to fill the gap created by the absence of daily fantasy sports play in Nevada has grown exponentially during the football season and is making even greater strides outside the state. Salerno, president of USFantasy and a 38-year veteran of the gaming industry, said handle has grown roughly 10 percent a week through the National Football League season.

## Dynamic Pari-Mutuel Rake

Recently, Bob Dancer posted a blog that briefly reviews US Fantasy, which is a pari-mutuel based fantasy sports product currently being offered in Nevada. Bob Dancer touches on two major areas where US Fantasy should improve. First, there is no app and no ability for bettors to place wagers online. This is a major drawback for players. Much of the growth of sports book handle over the past few years is due to the increased access that players have by using mobile apps. But this is also a fairly straight forward downside to overcome. It just takes some commitment of time and resources by US Fantasy.

The second area where Bob Dancer thinks US Fantasy can improve is in the basic structure of the games. He notes because US Fantasy uses a pari-mutuel format, there is no reason to bet early. It is in the player’s best interest to wait as long as possible to place bets. Waiting longer gives the player two major bits of info. By betting late, the player gets to see more of the pool structure. The better can see the latest wagering odds. By betting late, the player also gets to use the latest information on the event.  This could come in the form of scratches, lineup changes, weather, etc, that can impact the outcome of the event.

I think Bob Dancer makes two good points and his comments got me thinking more about how the pari-mutuel DFS model and pool wagering in general can be improved, especially with new technologies that are available.

Essentially, a pari-mutuel pool works by awarding a pool of wagers to the bettors who bet on the wining outcome. Odds are determined by the amount bet on each outcome and payoffs are determined the same way. Typically, operators charge a fixed rake for running the pool. I describe pari-mutuel pools in more detail in a previous post. For example, consider a football game.  There are two teams, team x and team y. \$4 is bet on team x, and \$6 is bet on team y. The implied odds of team x is 4/10 = 40% or 2.50 decimal odds. The implied odds of team y is 6/10 = 60% or 1.66 decimal odds. If team x wins, then the \$10 pool will get distributed proportionately to everyone who bet on team x and visa versa for team y.

Pari-mutuel pools are typically used for horse racing. This betting format takes away all the risk from the pool operator since a fixed rake is charged on the overall pool and the odds are set by the pool itself. It works well for obscure events where the outcome is difficult to handicap. The pari-mutuel betting format is not so great for bettors who have access to fixed odds alternatives. Pari-mutuel pool bettors don’t know the payoff/odds they receive until after the pool has closed. This makes it a little riskier for them and discourages them from betting early. The format can be gamed by sharps who can handicap the pool odds or who have access to an alternative pool or fixed price book.

For the reasons outlined above, pari-mutuel pools have not caught on with many sports bettors. Most sports bettors prefer to use bookies where they can receive fixed price odds. And since there are so many alternative bookies available today, especially online, a liquid secondary market exists on so many sporting events, bettors can know for sure whether they are getting a good price or not.

So how can we use new technologies to improve pari-mutuel pools?  Specifically, is there a way to level the playing field between sharps and squares to generate overall greater liquidity while also taking advantage of the beneficial aspects of betting pools?

According to Bob Dancer, there is a benefit to betting as late as possible in a pari-mutuel pool. Sharp bettors will wait till the last possible second to place their bets. They will scan alternative markets to determine the best odds for a particular outcome and calculate what the impact on the pool will be at a certain amount of wagering. Sharp bettors will find opportunities where a certain outcome is cheap in the pool and will place enough money in the pool for that outcome to gain an advantage. Square bettors are disadvantaged because they might not be able to spend the same time handicapping events and placing their bets at specific times.

I propose a system that will level the playing field between sharps and squares that will also grow the overall pool.  I call the system a “dynamic rake”. A dynamic rake will provide inducements to both sharps and squares which will level the playing field between both types of bettors while at the same time increasing the total amount wagered.

A dynamically raked pool will still have a fixed rake like any other pari-mutuel pool. What changes is the way players are rewarded for their bets and the information provided to participants. A dynamically raked pool will reward early bettors with more player’s club/cashback/credits. The earlier a bet is made, the more points it will receive. If a pool has a 10% rake, players who bet early might get a 20% bonus on their bets. This bonus will induce bettors to bet earlier. The points that are awarded can be determined by an algorithm and tweaked to suit the particular market. The pool operator’s experience will help them tweak the formula to an optimal amount so that the advantage between early betting and late betting is decreased or eliminated. The points could be awarded based on time, but also based on the amount that has been bet in the pool. The operator could determine the amount of points to award based on the amount the pool has received in wagers to certain limits. Once thresholds are met, the operator would be indifferent to inducing any more early bets as the minimum amounts will have been met.

In addition to awarding points based on time or amount in the pool, the pari-mutuel pool can also give pool participants more information concurrently. This will further induce sharp bettors. The pool operator could make it known the amounts bet on each outcome. This will help bettors determine how much certain wagers will impact the pool odds. The operators could also give betters an API that they could use to automatically place wagers. This will help sharps bet at the very last minute with the most up to date information.

What I believe a dynamic rake pool will accomplish by inducing early betting and helping sharp bettors, is increase the pool handle and increase the efficiency of the pool. This will enable the operator to make a higher overall rake and also make the odds closer to their optimal level.

Why aren’t dynamic rake methods used more commonly? I believe there are three reasons. First, regulations prevent pool operators from getting too creative. Most jurisdictions around the world put too many restrictions on wagering, and so stifle competition and innovation. Second, the growth of fixed price and betting exchanges has turned the industry’s attention away from pari-mutuel wagering. There is lot’s more growth left in legal sports betting, exchange betting, and DFS formats. Pari-mutuel pools are seen by the industry as an old horse racing format. Third many sports books, DFS operators, and betting exchanges have not become very sophisticated in their player rating systems. There is lot’s of room to grow in the way players are rated and rewarded but because the overall pie is still growing, more resources is being put into marketing and new player acquisition than into player retention.