MLB FORUM

Comments

  • JonBales

    RotoAcademy Lead Instructor

    • 595

      RG Overall Ranking

    • 2019 DraftKings FFWC Finalist

    • x3

      2015 DraftKings FFWC Finalist

    It’s that time of the year again – the time when no one is thinking about baseball but I am because I’ll be writing another daily fantasy baseball book.

    I have a bit of time before I actually begin writing, so I wanted to reach out to the community and see if there’s any specific topics you want me to cover or data you want me to collect. Obviously I’ll be doing all of the normal stuff, updated for 2016 (stacking data, cap allocation, park factors, etc), but feel free to throw out any specific ideas you’d like to see me research and write about.

  • rsigler65

    1) Predicting grand slams
    2) Keeping track of losses in the RG bankroll tracker
    3) The effects of global warming on weather forecast reliability and how it affects ownership % in East vs. West coast start times.
    4) Reverse stacking – how to use just one player from as many teams as possible to reduce or increase variance
    5) Excel techniques for identifying players that exceed value in small slates
    6) Advanced scripting: Multi-entering single entry GPP’s and strategy for single lineup entry in multiple multi-entry GPP’s

    These are just a few ideas.

  • Maxkim

    I like the idea of showing how to use excel for dfs.

  • JonBales

    RotoAcademy Lead Instructor

    • 611

      RG Overall Ranking

    • x3

      2015 DraftKings FFWC Finalist

    • 2019 DraftKings FFWC Finalist

    @mlemming15 said...

    Do you play even more GPPs in the beginning of the year because there is a very limited sample to start the year, thus making it a bit easier to exploit initial hot hitters/pitchers at the start of the season before sites like FD and DK can catch up?

    Great point, and yes. I’ve had far more success during the initial few weeks of MLB than any other period in any other sport, and it 100% has to do with the lack of data. When you’re a contrarian tournament player, a lack of predictability is actually you’re best friend. The more randomness the better, as long as others are underestimating that randomness and it’s apparent in ownership.

  • JonBales

    RotoAcademy Lead Instructor

    • 611

      RG Overall Ranking

    • x3

      2015 DraftKings FFWC Finalist

    • 2019 DraftKings FFWC Finalist

    @NeRo9k said...

    How about the Hot-hand fallacy?!

    Yeah, I do plan to talk about hot streaks and whether or not they’re predictive, as well as instances when we can leverage recent play into an edge.

  • JonBales

    RotoAcademy Lead Instructor

    • 611

      RG Overall Ranking

    • x3

      2015 DraftKings FFWC Finalist

    • 2019 DraftKings FFWC Finalist

    @rsigler65 said...

    3) The effects of global warming on weather forecast reliability and how it affects ownership % in East vs. West coast start times.

    I think this might be a little too broad.

  • Priptonite

    • Blogger of the Month

    @Jon Bales said...

    Yeah, I do plan to talk about hot streaks and whether or not they’re predictive, as well as instances when we can leverage recent play into an edge.

    Why? The Book has done this for you already

    Re: whether or not hot streaks are predictive. can definitely leverage other people believing in hot streaks to your advantage

  • JonBales

    RotoAcademy Lead Instructor

    • 611

      RG Overall Ranking

    • x3

      2015 DraftKings FFWC Finalist

    • 2019 DraftKings FFWC Finalist

    Right…even if not predictive, there’s the GPP game theory angle I’m thinking about more.

  • Maxkim

    https://www.washingtonpost.com/sports/nationals/research-supports-the-notion-of-the-hot-hand-baseball-players-always-believed-in-it/2014/07/16/5a70653e-0cf9-11e4-b8e5-d0de80767fc2_story.html

    Does anyone believe in these studies? They suggest there might actually be something to hot streaks

  • smallANDflaccid

    I keep finding myself only looking for studies on confirmation bias

  • zachlowry

    Id be interested to see which players preform better at the beginning of the season (colder temps) vs. the end of the season (hotter temps). Not trying to be racist here, but it seems like the Hispanics do better as the temps rise…

  • smallANDflaccid

    @zachlowry said...

    Id be interested to see which players preform better at the beginning of the season (colder temps) vs. the end of the season (hotter temps). Not trying to be racist here, but it seems like the Hispanics do better as the temps rise…

    If purely looking for race/birth origin vs temperature, shouldn’t you just look at how they do at various temperatures, vs where in the season? If they do better at start/end, it could be unrelated to temp.

  • Jeets232

    Big fan of your previous books and looking forward to the new one since I’m focusing on MLB to get my feet wet in DFS. I know you aren’t a big projections guy but maybe including insight to your or other pros’ models would be great. I having a lot of trouble trying to put together my own model! Also having trouble grasping your theories on “anti-fragility”.

  • Jeets232

    Also including what specific factors go into sites player salaries if its not proprietor

  • kantiger77

    I’d love some information on the best daily routine. Almost like a check list.

    I’d also love some advice on how to proceed as a low stakes guy playing a few bucks a day. Should it be H2H, where a low score can win if your competitor has a low score? Should it be 50/50, where a decent score does okay? Or GPPs, where you have to be in the top 20%, but where sometimes the cash line isn’t all that higher from the cash line in a 50/50? Or should we go for smaller leagues, 10-100 man leagues?

  • Putz

    @smallANDflaccid said...

    It would involve getting access to the right data feeds (most of which are not free and not easy to pull free), historically as well as live (drinking from the firehose of Twitter is expensive and a very large amount of data).

    From there you need to build the right natural language processing tools to analyze the sentiment type you want to look for – you could likely leverage existing Python libraries for it.

    From there, historically map the names/phrases and their sentiment to their events – and decide if you map actual day to actual day, front lag them, or post lag them – or some combination.

    Then run that through machine learning processing – at the bare minimum for correlations, but something more complex (random forest, XGB, Vowpal Wabbit, or nn) to then get any predictive elements out of it… and then score them to see how accurate they are vs random noise.

    I know for a fact there are hedge funds doing the equivalent for trading stocks (look up what happens to Berkshire Hathaway trade volume when positive news about Anne Hathaway movies hits – as for a way to not do it – I have’t looked recently, but I assume whatever fund was f’ing that up fixed it or went survivor bias into the night).

    From experience, I would say this likely isn’t as helpful as you might think.

    The wisdom of the crowd is one thing, but there are also a lot of noisy idiots, so you need to figure out how to mask the noise.
    Much harder than you may think.

    I never said its easy from my experience. There are plenty of NLP tools out there. Some tools that could help for scraping….

    http://fantasyfootballanalytics.net/2014/06/scraping-fantasy-football-projections.html

  • smallANDflaccid

    @Putz said...

    I never said its easy from my experience. There are plenty of NLP tools out there. Some tools that could help for scraping….

    http://fantasyfootballanalytics.net/2014/06/scraping-fantasy-football-projections.html

    Cool, good luck, sounds like you’ve got it sorted out.

  • Putz

    @rsigler65 said...

    1) Predicting grand slams
    2) Keeping track of losses in the RG bankroll tracker
    3) The effects of global warming on weather forecast reliability and how it affects ownership % in East vs. West coast start times.
    4) Reverse stacking – how to use just one player from as many teams as possible to reduce or increase variance
    5) Excel techniques for identifying players that exceed value in small slates
    6) Advanced scripting: Multi-entering single entry GPP’s and strategy for single lineup entry in multiple multi-entry GPP’s

    These are just a few ideas.

    One swing is all you need…

    https://rotogrinders.com/threads/one-swing-is-all-you-need-interesting-read-674580?page=1

  • britdevine

    • 2014 StarStreet MLB Playboy Mansion Finalist

    I’d love to see some advanced Pitch FX data vs batter

    Group pitchers into 4-5 groups based on fastball speed and movement, then look at how hitters do vs pitchers that fall into each group etc.

  • Putz

    @smallANDflaccid said...

    Cool, good luck, sounds like you’ve got it sorted out.

    Wow, you sound like someone who is easily defeated. Here is some good reading for you based on my experience and the papers can be found in Google Scholar. I simply asked for this type of analysis applied to sports and disagreed it was an art, as Jon stated, and have evidence it is a science. Never thought it was easy, but would love to see it applied to DFS.

    https://doc.research-and-analytics.csfb.com/docView?language=ENG&format=PDF&source_id=em&document_id=1053681521&serialid=gRAGx5o9KjpeAGBLPq7bpyJRa6r6fj06KjHB6PGBbGU%3d

    http://thegovlab.org/the-good-judgment-project-harnessing-the-wisdom-of-the-crowd-to-forecast-world-events/

    http://www.economist.com/news/21589145-how-sort-best-rest-whos-good-forecasts

    https://edge.org/conversation/philip_tetlock-how-to-win-at-forecasting

  • dpadilla

    I’d also like to see if there’s any split advantage between power hitters, power pitchers, slap hitters, and soft tossers.

  • jakz101

    • $1M Prize Winner

    • x5

      2015 DraftKings FFWC Finalist

    I’d love to see a section that highlights strategies for entering multiple lineups in a single contest and the differences between doing it in basketball and in baseball.

  • fandjinbkalltheway

    It’d be awesome to see a section that was devoted to bounce back games in relations to hot streaks and ownage percentages. So like what % guys are owned after going below their average Pt/g for 3 games straight vs above their average. This relates to the hot hand topic; are the “hot” hitters averaging more points over the next 2 nights or is it a good spot to be contrarian and go with the “cold hitter”(probably the answer in GPPs)

  • JonBales

    RotoAcademy Lead Instructor

    • 611

      RG Overall Ranking

    • x3

      2015 DraftKings FFWC Finalist

    • 2019 DraftKings FFWC Finalist

    @Jeets232 said...

    I know you aren’t a big projections guy but maybe including insight to your or other pros’ models would be great.

    Sure…you can create a model without projections, of course. I think projections are more valuable for more predictable sports, but I can discuss this in the book.

  • JonBales

    RotoAcademy Lead Instructor

    • 611

      RG Overall Ranking

    • x3

      2015 DraftKings FFWC Finalist

    • 2019 DraftKings FFWC Finalist

    @britdevine said...

    I’d love to see some advanced Pitch FX data vs batter

    Group pitchers into 4-5 groups based on fastball speed and movement, then look at how hitters do vs pitchers that fall into each group etc.

    I believe the guys here do something similar with Inside Edge. I will talk about this because I think a revised version of BvP in theory can be extremely valuable if we can overcome small samples in a smart way.

  • dp47

    • 2012 RotoGrinders TPOY Champion

    • 2012 FanDuel WFFC Finalist

    • 2013 FanDuel WFBC Finalist

    • 2013 FanDuel NFL Playboy Mansion Finalist

    Graph wants:

    How much more likely is a 4.5 run Team Total to get to 12 runs as opposed to a 3.5 run Team Total?

    Not sure that 12 runs should be the goal, I’m good with 10-15.

    But I’d love to see a breakdown of how often a 3, 3.5, 4, 4.5, etc got to X runs.

  • X Unread Thread
  • X Thread with New Replies*
  • *Jumps to your first unread reply

Subforum Index

RotoGrinders.com is the home of the daily fantasy sports community. Our content, rankings, member blogs, promotions and forum discussion all cater to the players that like to create a new fantasy team every day of the week.

If you or someone you know has a gambling problem, crisis counseling and referral services can be accessed by calling 1-800-GAMBLER (1-800-426-2537) (IL). Gambling problem? Call 1-800-GAMBLER (NJ/WV/PA/MI), 1-800-9-WITH-IT (IN), 1-800-522-4700 (CO), 1-800-BETS OFF (IA), 1-888-532-3500 (VA) or call/text TN REDLINE 1-800-889-9789 (TN).