Who should I play today?

I got an email Sunday from a fellow asking me who I recommend to play tomorrow.

I felt the urge to write something snarky, like there's dozens of sites that out there that will do that sort of thing, mostly poorly.  But I thought better of it.  It's what people have been taught to expect from Daily Fantasy Advice.  Why not?  It's more fun to talk about whether you think Joey Gallo can rip up Carlos Frias or whether Miguel Cabrera will crush Jon Moscot (who the hell is Jon Moscot?).

But it's relatively unimportant in my mind.  Maybe unimportant is the wrong word.  How about meaningless?  No one can accurately predict what will happen in a particular baseball game.  I'm not saying that's bad, it's what makes baseball my favorite sport.  There are no known outcomes, and nearly no "won" games, as in most other sports with time clocks.   And since many sites are so dependent on the Home Run for scoring, I would have to actually be confident in my capability to predict who will hit a Home Run on a particular day.  I can't do that, and anyone who tells you they have a handle on it are lying (take solace, they're not just lying to you, but also to themselves).  If someone could do it regularly, they would just gamble their way to prosperity.  It's something I noticed while working as a stockbroker: the good ones used the job to facilitate trading their own accounts, and the really good ones retired early.  I don't want to be the sort of advice giver that is all mouth and no money.  And DFS is all about the money.

Additionally, because of the special relationship DFS has between production and value, I find it impossible to field my personal lineup (for good or bad) more than about an hour before the games start.  I need to know the daily lineups to make a good decision about my roster.  Baseball lineups are far more fluid than other sports.  And a small change in one lineup can send ripples through your DFS lineup.   For example, let's say your looking at a relatively weak position like Shortstop, and you decide that the only SS you feel any certainty will have a positive night at the plate is Brandon Crawford, but for some crazy reason he costs $4,600.  But then lineups come out at Detroit is giving Andrew Romine the start batting ninth versus the incredibly weak lefty John Danks for only $2500.  Switching from Crawford to Romine at SS allows you to go from Kendrys Morales to Paul Goldschmidt at 1B.  And please note: this is an oversimplification of the process, a savings of $2000 might impact 3 or 4 different roster decisions that day. All decisions in DFS impact other decisions via the Salary Cost mechanism.  That's the beauty of the puzzle and the beauty of the game.

 I understand why sites do this: it's more interesting that hearing about these convoluted scenarios.  Real daily fantasy advice would be like hearing about a two team, eight player trade in season long fantasy baseball. You would have to know all the particulars like knowing that the guy has Mike Moustakas on his bench making the losing of Mat Carpenter not as big of a deal.  It would be boring as hell.  If you are looking for advice, I understand that CBS Sports and ESPN run a nightly DFS baseball programs somewhere in the 6PM EST area. At least it covers the basic problem: it is chronologically done at the proper time.

Anyway, I might consider specific advice somewhere down the line, primarily for other sports where there are far fewer game time decisions.  Or possibly for Starting Pitchers, since that is usually known well ahead of time and so very important, once I figure it out better (I'm not there yet, but I feel I'm getting closer every day).

My preference will always be to talk about the sport conceptually, rather than dwelling on the mercurial particulars.  The hope is you might learn from my experience and mistakes, and when I'm dead wrong, you would give me your thoughtful criticism to help me onto the right path.  Most Daily Advice will tell you how to field a roster, my goal is to help you understand the game.

All this being said, I do appreciate the emails and the positive feedback on this little excursion of mine.

A Monday with CSURam88 (Peter Jennings)

If you play on Draft Kings, you might have noticed the screen name CSURam88.  It's a gentleman by the name of Peter Jennings (No, not that Peter Jennings), who is listed on the site as one of the DK Pros.  He won $1 million bucks in the site's Fantasy Baseball Championship last year, so he's a guy we all can learn something from.

He usually plays at a higher dollar value, a number too high to simply throw away for knowledge, but last night he put out 31 challenges for $1 with a single lineup.  I've played him before, even won one I think, but I thought it might be interesting to analyze his Monday Night's Lineup for everyone (in which he beat me by almost 30 points).

I believe his process had 4 components:

1. He drafted the clear cut top pitcher - There's not of lot of Question the choice of Chris Sale in last night's matchup.  I took him too.  With pitchers there's Clayton Kershaw clearly at the top of the heap, and a small handful of pitchers in the second tier to which Sale clearly belongs alongside Felix Hernandez, Corey Kluber, and possibly one or two others.  Plus, he was
   
   facing the Strikeout prone Houston Astros Monday Night, so he was head-and-shoulders above the other pitching possibilities.
2. He was not afraid of a small reach in his number 2 pitching slot - He played Mike Leake, a guy I didn't even consider for my roster (I went with A J Burnett).  Leake was priced at a very affordable $6800.  Mike Leake had a hell of a preseason, and a good couple of games to start the year, but hasn't been good of late.  The two considerations I believe pushed Mr. Jennings in his direction was the price (which allowed him to spend considerably more than me on hitters), and the fact that he was facing the Philadelphia Phillies.  The Phillies are pretty good at making average Pitchers look good.  And with Cole Hamels starting for them, the Vegas line over/under was the lowest for the MLB slate Monday Night.
3. He loaded up with batters who played at Coors Field - Draft Kings allows significant stacking possibilities, but requires your hitters to come from at least 3 different teams.  7 of his 8 hitters took the field in Denver on Monday.  He went with 3 Cardinals and 4 Rockies: The expensive guys came from the Cardinals side of the ledger in Jhonny Peralta and Matt Carpenter.  The rest were value plays and one of the value plays, Ben Paulson of the Rockies was the deciding factor (we both played DJ LaMahieu, Nick Hundley, & Carlos Gonzalez.)
4. He chose a "hot" hitter to round out his lineup -  Chris Colabello was fresh off a significant hitting streak and a game winning hit the night before.  He had been one of the best deals on Draft Kings for the past three weeks, but on Monday, the powers-that-be jumped his salary from $2900 to an even $4000.  He was facing Brad Hand in Toronto's homer friendly Rogers Centre.  Considering the price tag, I believe he was targeted, rather than someone Mr. Jennings settled on in the end.

CSURam88 is the only guy of the hardcore players, I've ever seen routinely target what I would
consider substandard pitching.  Others will take lesser hurlers at times, but normally on more backing that simply a weak opponent.  There is little doubt it works for him, and may be a bit of a holdover from his tournament strategy success.

I attached a screenshot at the left of an earlier game a week or so ago also versus CSURam88, when the Rockies were on the road, and you can see the mild reach at Pitcher #2 with Yordano Ventura.  It also demonstrates how his team changes when the Rockies saddle up in Coors Field.  It shows how important a knowledge of what is going on at Coors Field is to your success at DFS. My quick take on his other rosters is that he's a complete sellout for Pitcher Strikeouts and Batter Home Runs.  The only non-power guy in the lineup is Jose Altuve (In this game, I lost by 12 points).

Also, the Coors Field focus is one of the first strategies I picked up on.  Games there are high scoring with a greater frequency than any where else, and there are many nights where, like Monday, a bit player like Paulsen or Hundley give tremendous results for their money.  Draft Kings has addressed this with higher pricing in these matchups, but CSURam88 certainly thinks they are still worth the added cost.

Max DaLury Revisited

I heard Max DaLury today on the Rotogrinders Daily Fantasy Fix Podcast while I was at the gym, and I think I vastly underestimated one aspect of his strategy (you can listen to this podcast on Stitcher here, it is 28 minutes long).

I didn't even consider the "stacking" part of the plan.

Max Dalury's Twitter Photo

Max said on the podcast that he normally goes with six players from a team he chooses to stack with.  He was talking about stacking Blue Jays in a night where he won 51% of 888 rosters laid out in tournaments.  When he was pressed as to how he chose the six, beyond the top 2 guys (Josh Donaldson & Joe Bautista), he said he "didn't want to go into that part of the strategy."  This deferral means that stacking strategy is a part of the equation that he feels is too valuable to divulge.  Which clearly tells me that my overlooking this concept is a significant error on my part.

The only other thing he refrained from divulging was how he actually finds the time to enter and track 888 rosters on a given night.

I will venture a guess as to what this means for his "Stacking Strategy," and there are two ideas that may explain it:

• The Cornerstone Concept

I've heard other players who talk about rostering multiple lineups talk about choosing two or three players as "anchors" for their team, while mixing up the other 5 or 6 hitters around them.  Because he was forthcoming with the idea that he did focus on Bautista and Donaldson, it may be that he expands the anchoring concept into the stacking concept when focusing on one team. Pick one or two batters to build around, and make players 3, 4, and 5 change within the team.  If you're stacking the same team in multiple lineups, you probably shouldn't be using the same stack in all of them.  Because the nature of this game is highly random, you need exposure and luck, a lot more than expert prediction skills.  It would make sense that if he had a good read on how to pick secondary and tertiary players with a single MLB team, he'd want to keep that to himself.

• The Volume Concept

I would be remiss if I didn't consider the possibility that there was absolutely no strategy at all.  When you play 500+ lineups, you can cover the vast majority of possible combinations, possibly all of the reasonably likely high quality options. What skill there was may not have been in choosing the players from Toronto to focus on, but rather in choosing Toronto and the weak arm of John Danks to exploit in the first place. If this were the case, he would refrain from revealing this information because it would be embarrassing to say that there's not really any skill to it.

I've noticed that the Max DaLury handle pops up in virtually every Satellite, Step contest, and Qualifier.  This means that it's part of his strategy to get as many free games as he can to make the extremely high volume multiple entry scheme more affordable.

He did say one thing that bears repeating when discussing how he started in Daily Fantasy (This is a paraphrase not a direct quote, but the concept itself is important).

:

The reason you should be playing is because it's fun.  It's more fun to win money, but it needs to be fun for you regardless.

An Evening with Max DaLury...

(Note: after writing this entry, I heard Max DaLury on a podcast, and it made me believe I missed something pretty crucial to his strategy, after reading this it's in your best interest to read my follow up entry here.)

Lately, I've been trying to avoid the big hitters on the Draft Kings site, making attempts to play with less experienced players like myself.  Early on, I sought out the big guys to see what they were doing, I figured it was worth a couple of bucks to see what the seasoned pros did.  I won a few, lost a bunch, and learned a few things.

Tonight, in a $5 10-team league, I was shocked to find Max DaLury pop up in the game.  His name is ubiquitous on the Draft Kings site (if you want to see who is hosting the $10K H2H game, it'll be this guy 9 out of 10 times.  As it's been a few weeks, I thought I might take another look at his line up construction.

Now, I came in second with a 111.95 points and came away with some cash.  And unfortunately, you can't glean too much from the pitchers he chose because it was a two game format where basically everybody took Hendricks & Salazar (as opposed to John Danks and/or Odrisamer Despaigne).  But the hitters show a discernible pattern.  There are three basic groups he picked from, with a little overlap and one outlier.

1. He picked on the obvious worst pitcher of the night.  John Danks was clearly the worst of the four possible pitchers and was the most likely to lose this night.  He's been pretty bad all year long with a 4.66 ERA & a 1..37 WHIP.  A pitcher who is bad can make fairly pedestrian players on the opposing team seem very good.  And this strategy works more often than not.  This reasoning accounts for Carlos Santana, Jason Kipnis, Mike Aviles, Ryan Raburn, and Nick Swisher.(or 62.5% of his roster).

2. He chose players who were hot in the very short term past.  Jason Kipnis has been on fire in the past week. Kris Bryant has been at or very near the top of the Third Base pile since his second game in the big leagues.  Dexter Fowler homered the night before, and Mike Aviles was fresh off a two hit night the day before (both good for more than 10 Draftkings points).  This part of his strategy accounts for 50% of his roster.

3. He chose players who had very good splits versus lefties.  Aviles, Raburn, and Swisher are three players on Cleveland known to beat up on the southpaw pitchers, and have been very good at this during the 2015 season.   So it wasn't enough to target players who were facing Danks, but it was important to get the players Cleveland uses to beat up on pitchers like him..  This accounts for 37.5% of his roster.

We must, of course, note that four of the players appeared in more than one category.  Which means that if a player has one good reason for choosing him that's alright, but more than one makes the case for that particular player all the stronger.

That leaves one guy remaining, the outlier, Chicago White Sox First Basemen Jose Abreu.  It would be easy (and wrong) to build a fourth category around something like Superstar Players (Besides Abreu, Bryant, Kipnis, & Santana might qualify at their respective positions).  But I don't believe this is the case at all.  I believe that the first 7 players were targeted and he took the best available remaining player at the last remaining position.  It is impossible for me to know what exactly went on in his head, but there was no obvious reason for choosing Abreu.  He's not been too spectacular this year and he was facing Salazar who was projected to be the toughest of the four pitchers to face on the slate.  Now, Abreu does have nice Batter vs Pitcher results versus Salazar, but it's only been 6 at-bat's as they both were rookies last year.  This might have been considered secondarily, to kind of give a push for one guy over another, but a sample size that small would never be a principle reason for picking a player.  I ruled out the Superstar factor because I fielded Anthony Rizzo, who is a bigger name so far this year, and they were the exact price tonight at $4900.  Maybe, it was the BvP, Maybe is was the Ball Park factor (US Cellular field is much easier to hit a home run in the Petco), maybe it was a personal preference, Whatever the reason, I believe it was the final player rostered for him this night.

I've played Max DaLury at least a dozen times before, and this sort of strategy is pretty typical based on my previous experiences.  If it works for him, and it appears it does, it should work for you.  Regardless, there are important lessons to be learned from the roster construction of one of the pros.

GPP (sic) Lineups

Generally, there are two types of games available on Daily Fantasy sites.  They are normally referred to as GPP and Cash Games.  It is important to note that the lineup strategy differs in each of these games.

I don't like the term "GPP" used in DFS to refer to tournament style contests where you play a lot of people for higher rewards.  GPP stands for Guaranteed Prize Pool, which covers many of these games, but many of the 50/50 games carry the same designation.  Also, many of the non-guaranteed games, like 10 man leagues or just about any winner take all format, should employ similar tactics. Cash games, on the other hand, cover H2H and 50/50's, and this moniker is also a little problematic, as virtually all games are awarded cash prizes (besides special prize games and steps).  The confusing terminology is probably one of the myriad ways the people in the know about DFS try to obfuscate the game to make it harder on new players.  I also think Triple Ups and 3-5 man leagues are sort of a hybrid between the two, and should be addressed by a blended approach.  The divisions are completely based on odds of winning considering the match a completely random occurrence.

If you remove the skill of lineup selection from the game, your odds of winning are (I've included the phrases I like to use to help understand my writing in other articles):

My Terminology	Type of Game	Approx Odds	Odds as Percent
Head to Head	Cash	1 in 2	50%
Hybrid	"Leagues?"	1 in 3-5	33%
Tournament	GPP	20 in 100	20%

Of course this doesn't take into consideration player skill, and some Double ups & Triple Ups add a few more entrants to make it a a true triple up (not $1 to win $1.80, but $1 to win $2) which decreases the raw odds by a small, but noticeable amount.  And, in tournaments, the percent of winners fluctuates depending upon the game, but 20% is a solid rough approximation.  Your skill is, of course, a very important factor to attempt to skew these odds in your favor.

Anyway, I heard a "pro" DFS player named David Epstein (BW5126-not the World Series MVP), who almost exclusively plays Tournaments on the Rotogrinders Daily Podcast earlier this week.  He recently won $100K in a baseball tournament. I'm not saying I learned everything I need to know from the 20 minute discussion, but my basic interpretation of his process looks like this:

• He starts, as all do, by pitcher selection
• He targets 6 of the 8 batter selections with specific players
• He then selects the remaining 2 based on best available, and can play upward of 20 variations rotating the last two slots on a significant number of scorecards.

I wanted to note this because I'm am going to play a few more tournaments going forward to try to understand this part of the DFS universe.  I'll be showing more frequent losses and, hopefully, higher payouts when I win.  The advice from David Epstein, I believe, is a good starting point.  Thankfully, Draft Kings runs a daily tournament called Quarter Arcade, which costs what you think it does, that makes testing ideas possible without punishing your bankroll.  There are also games at very affordable $1, $2, $3 & $5 price points.