Answers to fantasy baseball questions (and much more) since 1996
The New York TImes has an informative story today about the rise of the daily fantasy baseball game, which has been embraced by MLBAM and managed to avoid being classified as a game of chance.
The writer quotes a guy who says he works 35 hours a week playing the game and made $50,000 last year, which seems very possible. It also mentions a recent surge of casual players who are increasing the pools. It might be time to start playing.
The Wall Street Journal ran a profile last month of a super sophisticated graduate student at Notre Dame who has supposedly made $200,000 in the last year playing daily fantasy games.
Cory Albertson has, according to the story, put together an algorithm that helps him put together many daily fantasy lineups on any given day, which allows him to enter many fantasy contests and overall make him money. Not every day, the story says a couple of times, but overall. He expects to make $1,000,000 this year, he says.
He makes so much money, according to the story, that he went out and test drove a Tesla!
Yes, that’s my snark. There are a few red flags in this story that challenge the writer’s competence or veracity. For one thing, Albertson didn’t buy an expensive car, he test drove one. And he broke the speed limit!
For another, Brad Regan, the writer, blithely reports that Albertson got into the game because a friend started a Daily Sports Fantasy Game last year and waived the fees so that Albertson would help populate the board, making it look like his contests were more popular than they were.
The biggest impediment to winning any gambling game, from daily fantasy basketball to trading stocks as a day trader to online poker to horse racing, is the takeout. That is, you and the other players may put in $100 in money into the pot, but the service rakes some of that for itself, before they pay out some lesser percentage to the winners. That’s how they make money. It should be needless to say, that if your bets are not raked you have a much better chance of winning than if they are.
Regan doesn’t pursue the question of whether Albertson is now subject to the house fees. He doesn’t discuss how much the house usually takes out of a daily fantasy pot when Albertson isn’t playing. He leaves out the single most important piece of the way the business works, while pitching us that Albertson is some new breed of non-gambler who uses data to drive his decisions.
I say Non-gambler because Albertson tries to make the case that betting on Daily Fantasy Sports isn’t a gamble, because he uses his algorithm, which apparently takes all the subjectivity out of it. You can be the judge of that judgment, intellectual and ethical. I say, no wonder Albertson’s religious parents remain concerned about him.
The other interesting bit comes when Andrew Wiggins, who started DraftDay, a daily fantasy game service, talks about the need to get casual players to play. The idea is that the small fish, who might put up a $10 or $20 bid on a daily fantasy team, will drive the growth of the game. That’s what Wiggins wants, because he gets a cut out of every bet made.
But Regan quotes Albertson deftly crushing Wiggins’ dream: The smart guys, Albertson says, will feast on the casual player. This, as Wiggins surely knows (he and I played in a fantasy baseball league that was populated with some professional poker players who feasted on online poker guppies, back in the day before that business collapsed for legal reasons, as well as this inconvenient fact), is surely what will happen.
That imbalance, when heavy advertising is drawing in fresh blood, is one reason that a shark like Albertson and his algorithm might legitimately be doing well. Seasoned players with good data will crush the haphazard random player, the way a card sharp will crush a county fair card game if you give him enough time (but watch out for the rake, it’s going to charity).
Right now daily fantasy sports are a young and growing business. They have appeal to skilled and less-skilled players because of the short results horizon. I can see the appeal, but any smart story about the game should really talk about the way it works, and not pitch some fairy tale get rich quick by working hard line that reads more like a slow pitch PR piece for the industry than the human interest story is is dressed up as.
I don’t know that Cory Albertson didn’t make $200,000 playing daily fantasy sports, and if he did I don’t know for sure that he won because he didn’t have to pay the fees that normal fantasy players would usually have to pay. I do know that this story and its failure to accurately describe the way the games work casts doubt on every part of the story that can’t be fact checked.
Over the winter I came up with the idea of posting a Tout Wars Leaderboard at toutwars.com. The idea would be to assign an entry fee for each team, and then award prizes at the end of the season reflecting the payout, based on the number of teams in the league (since we have 12 in AL, 13 in NL, and have had anywhere from 12 to 17 in the Mixed league).
One sensitive issue was the prevailing ethos when the first 12 years of Tout Wars were played, which said that, “Second place is first loser.” I know that there were times when, dealt a tough hand, I made the high-risk high-reward (if only) trade rather than grind into battle for fourth place, the way I would have if there was money at stake. I’m certain others played this way, too, so these retroactive results are imposed from above. They don’t reflect how behavior might have been changed if this way of measuring was known when the games were played. In spite of this limitation, I forged forward.
Choosing a $100 entry fee was easy. It’s round, easy to average, like an index. Done.
Deciding how to pay out fairly was more complicated. I’m not going to go into all the round and around I did, but I decided to pay out to the top 33 percent of the teams in each league. In 12 and 13 team leagues this means the top four teams are paid. In 14-16 team leagues five teams are paid. In 17-19 team leagues 6 teams are paid. This seemed in keeping with the original roto rules of Top 4 being paid in a 12 team league.
But the standard roto payout is the somewhat awkward 50-25-15-10, which not only isn’t linear, but it doesn’t scale to the larger-sized league payouts. Since the idea was to compare across leagues, the payout reflecting the difficulty of prevailing, it seemed to me to be a good practice to have the percentages be fairly consistent as they scale up. So I made a simple equation, assigning X to the last money spot and then doubling the value of X for each higher spot. For a 12 team league this looks like: 1x+2x+4x+ 8x. In this case x = 1200/15 = 80. The payouts go:
Fourth=$80
Third=$160
Second=$320
First=$640
Sweet. The same process was used in the larger leagues, all of which–I think–does a fairly good job of representing the value we take away from our standard roto leagues, which were the models for Tout Wars. But this isn’t the only way to do this.
I’ve played (and play) in leagues with a more graduated payout. That is, last place may get nothing, and each place up the standings earns a little something more than the one below it. This is more like the way real baseball works. It isn’t all or nothing between fourth and fifth, but a graduated scale reflecting success in winning and managing costs. It would certainly be instructive if we could see the profil/loss statements for the actual teams in addition to the won/lost records. A graduated payout changes the way teams play, since there isn’t the same desperation to get into the money, but there is incentive to win more money.
I’ve played in leagues with no salary cap, where you could spend as much or as little as you wanted (and deemed prudent), in order to win a percentage of the pool. This, I found, turned out to be less interesting than (I think, because we never got that game together) )if we’d played for fixed amounts for first, second, third and fourth, with no salary cap, and any extra money that was spent was paid to those who finished fifth, sixth, seventh and so on, down to next to last.
I’ve also played in winner take all games, and others that paid bonuses for rare events, like no-hitters and hitting for the cycle, or allowed the roster to carry over into the post season, for a second pool. These payouts weren’t major enough to change regular season play, but if they were incentives might be skewed and teams might play differently. The point is that designing the payout structure of your league (which pertains even if you don’t play for actual money) reflects decisions you’re making about how you value winning, being the runner up, the importance of participation all season long, and a host of other things that make your game fun or more or less so.
There isn’t a right or wrong way in the end, but whatever decisions you make will inevitably reflect the values you bring to the game and the value of winning (or not). That’s pretty darn interesting, if you ask me, and relates to whole lot of things we do outside of our fantasy sports pursuits.
I joined a new high-stakes 5×5 AL only auction league this year. Some of the prize money is put up by a poker education site, cardrunners.com, and some by the participants, who are a mix of fantasy experts, professional poker players, and financial pros. There are only 10 teams, but you can spend your money on all 28 of your rosterable players (you don’t have to, there is a draft when all teams are out of money). This changes the endgame some, as Rotowire’s Chris Liss notes in his post at Rotosynthesis (where he also posted the draft results).
Another wrinkle is that you can buy NL players. I spent some time trying to figure out what Adrian Gonzalez would be worth, and considered throwing him out early, but someone (I don’t remember who) beat me to it. My back of the envelope calculation was that a 50/50 chance for half a season of Gonzalez was worth a blank $8, though that calculation would change as the auction progressed. As teams recognize their strengths and weaknesses it might make sense to bid more for the high risk play. The gambit of coming out early could mean a bargain. In fact, I bumped a $3 bid to $4 and Daniel Dobish, Dave Gonos’ partner, muttering “I’m not letting him go to someone for free,” bid $7 and won him. Not a huge risk, but a nick in his budget he’ll feel if Gonzalez doesn’t come over.
There was a similar calculation in my most uncharacteristic moment in the auction. After adjusting my prices for the smaller league I was pleased in nearly every case but one (there was also a blip in the late early part of the auction where the price of outfielders who steal, namely Ichiro and Denard, went for scandalously low prices) that they were accurately describing the action. The difference came with the catchers, where huge draft inflation persisted all night. The action players, at least the guys who won the high-priced catchers through most of the auction, were the non-experts. They took Mauer to $40 and Victor Martinez to $35, and Napoli and Suzuki to $18. Even at the low end, guys I had listed for $2 were going for $5. Matt Wieters name was called fairly late, but there was still plenty of money around. His price surged past my $16 bid limit, but I had money to spend and when the bidding slowed at $20 I bid $21 and won the sophomore backstop. The move effectively changed my team from Nolan Reimhold and two scrub catchers to Wieters, Jose Guillen and a scrub who turned out to be Brayan Pena.
I don’t remember who had the penultimate bid on Wieters, but if it was one of the Cardrunners boys my brash reach means I wrecked the purity of a position-scarcity experiment, with the so-called experts buying cheap catchers and the so-called amateurs buying the pricey ones. All of them, as noted before, were inflated.
This morning I ran the projected stats of all the teams using the CHONE projections, mostly because I have Chone Figgins on my team. The key is to avoid testing your team using your own projections, since they naturally flavor the players you pick up. I don’t want to give up any competitive edge this exercise offers in its details, but I’m delighted to share for posterity the final standings, which surely won’t look anything like this next October.
TEAM PTS
Phipps 62
Carty 56
Rotoman 53
Hastings 51
Chad 50
Gonos 49
Wiggins 49
Eric 46
Liss 40
Erickson/
Sheehan 38
Since these include active rosters and reserves, and NLers Gonzalez and Ricky Nolasco, and Chone’s projections are generous with the playing time, upping the value here of guys who may not even play, they should be taken with a grain of salt. But they’re a start while we wait for games that matter.
I wrote about ProTrade a couple of months ago, or you would probably better say that I linked to it. I was impressed by the software and the user experience, but to tell the truth I haven’t been back.
Michael Lewis’s story in the premiere issue of the new Conde Nast business magazine about ProTrade suffers from some boosterism, but you can also call that enthusiasm and conclude that he’s right. The future in sports projecting will be culling increasingly sophisticated information from the crowds. ProTrade didn’t invent this business, by a longshot, but they’re very nicely positioned to take advantage of it.
I haven’t been back to ProTrade after my initial foray. But Lewis’s description of all the funny money, the emotional backing, that a sports exchange will draw into its market, is provocative.
The only problem is that a whole lot of die hard financial folks also happen to be sports fans, too. So what happens when Merril Lynch starts a Football Trading division?
I’ve only spent 15 minutes on the site, where I learned that my immediate “friend” Derrek Lee lists all the members of the Chicago Cubs as his “friend”s (where am I?), but I get the distinct feeling that someone has gotten the virtual stock market right. How it works out will depend on how free they let the shorts roam, and how real money plays a part (I’m not clear about this).
The “analysis” offered that I’ve seen mimics broader cable TV analysis, but it seems at Protrade everyone gets their own blog, so in time reliable reporters and analysts might prevail (if there’s a reason for them to persist, it doesn’t look like there is a meritocratic payment system for analysts yet).
But most importantly the scoring system is clear, so if they can convert their virtual financial system into a real one there is a gambling game here that might warrant some real attention. Stay tuned.