Pick Six Strategy, Groups: The Chronicles of ottoneu
After picking six for over 20 days straight, I’ve won one day. One stinking Bobby Valentine. Perhaps I’m not the best person to turn to for strategy in this particular game. I’ll offer it up anyway.
Mimicking my general fantasy strategy, I’ve been pushing money away from pitching. There are so many good matchups every day, why not find a good, cheap pitcher against a terrible offense? Last night, I chose a $17.75 Shaun Marcum (at home against New York), but that’s about the most expensive I’ll go for a starter. I was severely tempted by a $10.75 Daniel Hudson in Pittsburgh, and even Jonathan Sanchez at home against the Nationals made sense at $6.25 (or his opponent, the $3.50 Jordan Zimmermann). Don’t spend on starting.
If you are not going to spend on starting, there’s really no reason to spend on relieving. If a starter is no guarantee to perform on any given night, at least he’s guaranteed to show up and, you know, play. The ottoneu Dark Lord Niv Shah was kind enough to give me the points per position in the Pick Six game so far. Clearly, relievers are the red-headed step-children of the game.
| Position | Average Points |
|---|---|
| C | 4.248 |
| CI | 6.271 |
| MI | 4.596 |
| OF | 5.909 |
| RP | 3.193 |
| SP | 7.091 |
After all that blabbing about not paying for pitching, it might be surprising to see that starting pitchers have the best number in this table. But the point wasn’t that pitchers are not worth spending money on, the point was that there were plenty of bargains. Niv is nothing if not eminently bug-able – he graciously gave me the points per dollar spent for each position.
| Position | Average Points Per Dollar |
|---|---|
| C | 3.096 |
| CI | 5.461 |
| MI | 4.383 |
| OF | 4.185 |
| RP | 1.873 |
| SP | 2.057 |
Ah-hah. So, even just by groping my way through the dark with intuition instead of the numbers in front of me, it looks like I stumbled onto a viable strategy. Spend where you get the most return for your dollar: non-catcher offense.
Oh, and while you are at it, join a group! Now Pick Six has groups so that you attract like-minded players easily. Oh, and with groups you might be able to put a little friendly gummy-bear wager on the thing. If you like gummy bears or whatever. I’m in #PraiseBeltran and I started RotoGraphRegulars. Pimp your group here if you want others to join in!
With a phone full of pictures of pitchers' fingers, strange beers, and his two toddler sons, Eno Sarris can be found at the ballpark or a brewery most days. Read him here, writing about the A's or Giants at The Athletic, or about beer at October. Follow him on Twitter @enosarris if you can handle the sandwiches and inanity.
I really think we need some sort of replacement level analysis here. Is the minimum 0 points and 0 dollars? Probably not.
Maybe a simple scatter plot (or one for each position) showing points against cost for all players?
Or maybe you’d want to plot “positional entries” instead of “players”. So you might plot Matt Joyce at $2 25 times, but someone else no times.
With ~200 entries per day, the average player would expect to win once in about 200 tries. So once in 20 is a good sign.
I dig Pick Six because it’s free and quick, but IMO the pricing system for players is simply not good enough. It doesn’t react nearly quickly enough to new information. And by going off the dynasty league prices, it’s got way too much injury-proneness value built in. You got guys who are dirt cheap in dynasty league I guess because they are old/injury prone, but for tonight’s game they are projected for a .380+ wOBA. Do I pay $30+ for Youk or Ryan Howard, or get the same/better production from Big Papi for $3? Tons of examples of guys like this: Berkman, Matt Joyce, Carlos Quentin, etc. all bargain basement prices.
This is why your pts/$ is highest at CI. It’s not because that $30+ investment in Ryan Howard is better than that $30+ investment in Cole Hamels. It’s because you can get Ryan Howard’s production for $3-$5 in Lance Berkman or David Ortiz.
IMO the pricing system needs to have built into it feedback from pts/$. Everytime someone outproduces their value, their price goes up, and vice versa. Just expecting the market to fix it is busted because with fixed budgets I’m forced to spend my money and after buying one or two dynamite <$3 guys and a cheap RP, I've got $115 to spread around to get three of the best players at their positions even if the pts/$ for them is less attractive than the cheap guys. So the market still has me paying top dollar for the big guns even though they aren't remotely "worth it" compared to the cheap guys. If I could just cash in my extra dollars at, say, 0.1 pts/$ then I'd probably be spending <$20 on every roster and buying tons of points. (BTW, your points per dollar numbers make no sense to me. I'm really averaging 90 points from my $30 catcher?)
Generally agree, although I’d ok with prices dictated by popularity instead of actual production if the system reacts a lot more quickly. Crazy day to day fluctuations are fine, IMO. Keeps things fresh and it’d be fun to find the “deal-of-the-day” instead of simply deciding if I want Matt Joyce at $2 or not.
Agree completely that too much fluctuation in prices makes a better game than not enough when the underlying pricing model is a little naive.
Could change prices every day at 10pm ET instead of just Sundays at 10pm. Very trivial change to make.
For the record, all the “value picks” you’re seeing… That is happening because people are just not picking those players. Keep hitting those values, but don’t think that the model is not responsive. Bautista since the game started has gone up $30, so the model definitely takes into account how people act – maybe the sample size isn’t big enough (only 350 or so repeat users), but maybe people just don’t realize they should be picking Joyce every day.
I’d guess that a lot of people don’t scroll too far down — don’t want to wade through all the scrubs to find a guy like Joyce. (For a while, I, too, assumed that good players would rise above those prices.) Not sure what you do about that.
It’s a free game, so a lot of people aren’t going to play to maximize their chance of winning… they’re just going to pick players they like regardless of value. Bautista is the nexus – a guy who was both a value guy (and will continue to be until he’s the most expensive player in the game) and right now the most fun guy in baseball.
As a guy who is hardwired to play any game no matter how casual with an eye to winning, however, I’m compelled to find the cheap guys. The game would be more fun for me if I’m not basically forced (again, I’m hardwired) to pick from a pool of five super-value guys every time for my CI/OF slot and put Bautista in the other one.
BTW, Niv, great job with the game. It IS fun… takes me just a few minutes to set my lineup, and despite some of the wonky pricing, matchups do matter so I scan the list of players and get a good feel for all of that days games before work. Also, I kinda like trying to predict which relievers are going to play. I’ve gotten better at that.
My suggestions: Update the pricing more frequently. Add to the updating function a factor that takes recent performance into account (if not a ton of coding obv). IMO it’s fine if “hot” players tend to get a bit overpriced and “cold” guys get underpriced. Ortiz has been killing it this season and he should be up there with the big boys even if no one picks him. With a guy like him, actually, I would bet he gets picked MORE at $30 than he does at $3 because of the lost with the middle phenomenon Sky mentioned. Huge name, huge team, having a great season… you’d think people would want to play him.
I think Sky might be right about most users are treating this more casually. And I understand the concern about having 3-4 “must picks” every day in terms of value. Maybe updating pricing 2 times a week instead of just once will help. I’ll look into that.
Glad you are fundamentally having fun with it though! I know these conversations wouldn’t even be happening if you didn’t like the game to begin with.
Niv,
Would it be possible to automatically post a “dummy” team of the best value plays of each day? Not sure how easy this would be to automate for you, but it would be nice to see everyone’s scores vs. the best possible lineup that could have been built that given day under the $115 cap. One day that may include Albert Pujols and a bunch of scrubs, or one day it could be very balanced and more similar to the lineups that other teams are choosing.
Hope this make sense, but it would be great to see what the most productive team could be on a given day, especially if you find yourself at the top of the leaderboard and realize you could have scored more by spending a little less (or a little more).
I don’t have a great place to announce this stuff, but ideal lineup has been in the works for some time, just figuring out the best way to integrate it with the site! Thanks for the suggestion Trey, definitely something I’m interested in as well.
I think today is day 4 for me… One Bobby Valentine and a lot of fun so far. I am on the verge of becoming an achievments whore.
Can anyone tell me where i can find these pick 6 games?
http://ottoneu.fangraphs.com/picksix