Community Submissions

A Poker Player’s Guide To Fantasy Football

I funded my first year of college playing Texas hold ’em. Not bigtime, Late Night Poker hold ’em: Online low-stakes tables – up to eight at a time, up six hours a night. Starting off, I thought I was going to be the next Phil Ivey or Daniel Negreanu. Let me tell you, the novelty wears off fast and I burnt out. When it comes down to it, online poker is about grinding out a couple of big blinds/hour. But this isn’t a steady return – some nights you could get “stacked” multiple times, losing a whole buy-in in one hand and ending up down hundreds of big blinds at the end of the night. And of course, some nights you’re the beneficiary and end up stacking someone else. So how does this apply to Fantasy Football? The answer is a a simple but powerful concept: Expected Value.

 

Expected Value Explained

When you are experiencing swings of +/- 50 blinds/hour on a daily basis, yet your actual rate of return is only 1 or 2 blinds/hour, how can you be sure you’re making the right decisions? This kind of swing is known as variance. You’ve just lost your whole buy-in, but you had aces, and he lucked out on the river with a rubbish hand. But should you have known? Maybe you missed a tell…

Or maybe you made the correct decision with the information you had and simply lost against the odds. This is also termed “getting your money in when you’re ahead” or making a decision with positive Expected Value (+ EV)*. If you consistently make decisions with +EV, you will be successful in the long run. For example: Arsenal assets such as Santi Cazorla, Theo Walcott and Mesut Ozil were overwhelmingly backed to do well in the Gunner’s opening match at home to West Ham. Talk of “Walcott (C) fails”, etc. littered the match day comments. The clamouring for Dimitri Payet and kneejerk transfers began not long after.

I’ll use Hazard as an example, but this applies to every player. I gave the armband to Eden Hazard in Gameweek 1, as with the best information available I was making (in my judgement) a +EV decision. Hazard averaged 6.1 points per appearance last season; in simple terms, I’m expecting a return of 12.2 points from this fixture (for the sake of brevity I’m not going into depth on his stats here). Out of my squad, I was expecting the highest returns from Hazard, and therefore captained him.

As it panned out, Hazard blanked. I only got four points out of the 12.2 I expected. My Actual Value was four points, but my Expected Value was 12.2. I’ve attached a chart (http://i.imgur.com/uVXk6qq.png) illustrating three scenarios that could have happened (over several Gameweeks for ease of illustration).

The first is Negative Actual Value – you make decisions that over the long run are expected to return positive results. Imagine that Hazard has a 70% chance of scoring/assisting in any one match. That means that 30% of the time, he blanks. This was one of those occasions. That doesn’t mean that because he blanked, the decision was incorrect. This is a “bad beat”.

The second is Neutral Actual Value – you were expecting about 12.2 points, and Hazard bagged an assist and a BP. A return of 12 points means your Actual Value and Expected Value closely align – the decision was about as profitable as expected.

The third scenario is Positive Actual Value – Hazard goes on a rampage and comes away with a brace, an assist and maximum BP. Fantastic. However – this was not expected. There was maybe a 5% chance of this outcome. You did better than the information indicates you should have – you flopped the “nuts”, and if you’re on a streak of these results, you’re “running hot”.

Over the course of the season, you expect Actual Value and Expected Value to converge – in the long run, if you are making decisions with Positive Expected Value you will end up with a Positive Actual Value. If you are making decisions with Negative Actual Value, then vice versa.

Conclusion

The lesson here is to ignore the bad beats (when they are bad beats, and not the result of bad decisions), and ignore when you’re running hot (just because Leicester’s Riyad Mahrez scored twice doesn’t make captaining him a +EV decision). Points earned or missed out on don’t retrospectively justify a decision after the fact – make consistently correct decisions with the information available, and you’ll be rewarded in the long run. A word of warning though: there’s a saying in poker that “the long run is longer than you think”. Just try to strap down that knee in the mean-time.

*I’m heavily stretching the actual definition of some of these poker terms to suit FPL; poker nerds please be gentle.

85 Comments Post a Comment
  1. Clump
    • 4 Years
    3 years, 9 months ago

    Nice read. Bravo.

  2. Iceball
    • Fantasy Football Scout Member
    • 4 Years
    3 years, 9 months ago

    Ivan worth keeping or ship to make way for Aguero/Sanchez? Özil other option.

  3. John t penguin
    • 4 Years
    3 years, 9 months ago

    I must have forgotten to take the jokers out my pack first

  4. tm245
    • 7 Years
    3 years, 9 months ago

    Nice read. Don't forget how apt "tilt" and "bad beats" analogies are on this site as well.

    Too many of the Hot Topics and posts over the past 48 hours have ignored the basics of what you are saying and have been cataclysmic in their reactions or have been seeking solace in performing below expectations, instead of the more measured expectations that any poker player would bring to the table.

    1. dribbler
      • 9 Years
      3 years, 9 months ago

      "instead of the more measured expectations that any poker player would bring to the table"...you missed out the word "competent"

  5. itsmilks
    • 4 Years
    3 years, 9 months ago

    Good read. Like the analogies - agree 100% with the sentiment. Most people are a Mahrez away from the top. Just frustrating it's the first week

    Bad week? Yes.
    Bad luck? Yes.
    Is your strategy sound? Yes.
    Do I need to change anything? No.

    1. OShaughnessy
      • Fantasy Football Scout Member
      • 6 Years
      3 years, 9 months ago

      Comment so nice I wish I could upvote it twice!

  6. FrankieTheGent
    • 4 Years
    3 years, 9 months ago

    The question therefore is what information can we mine from GW1 to improve the decisions we make?

    So did we learn we should captain Mahrez away to West Ham or Oxford Reece will play 90mins every week for 38 games. No.

    But what did we learn?

    Let me suggest that...
    Silva could be a prime MOTM regular (for those Fantasy Sky Players)

  7. OShaughnessy
    • Fantasy Football Scout Member
    • 6 Years
    3 years, 9 months ago

    @Get_in been trying to explain this idea of +EV in FPL terms for years & you've done an exceptional job! (Kinda jealous that you managed this much better than I ever could tbh ;))

    Make the most +EV decision every time & *know* whatever the result that you've actually come out ahead points wise.

    Example:
    Last year I captained Aguero away to West Ham.
    He blanked but, managed 7 SiB in the process.
    All I could do was smile & shake my head...
    Right decision, just the wrong result.
    Because, anytime someone gives you Aguero's Pts x 2 in a match he got 7 attempts on goal is one you're going expect to come out a winner in the long run.

    1. dribbler
      • 9 Years
      3 years, 9 months ago

      "been trying to explain this idea of +EV in FPL terms for years"

      whether FPL or poker, you have to separate the decisions you make from the results you get...if you make every decision you take a "good decision" you'll end up with good results in the long term...against a "normal" or unknown player if you have AA or KK you go all-in...that is the correct decision...the outcome doesn't matter

      EVERYBODY suffers from tilt at least to some extent...just recognise when it's happening and stop playing until you calm down...in FPL the same thing happens...I played much better when I stopped reading the chat on this site and recognised that my "inferior" team was good because I got it with very few hits

  8. Sciolist
    • Fantasy Football Scout Member
    • 7 Years
    3 years, 9 months ago

    The player who came 7th last season is a poker player. FPL is pretty popular within the UK online poker community, though I don't think many have finished quite that high before..

  9. TuttiFrutti
    • 6 Years
    3 years, 9 months ago

    This is bullshit! There will always be hard to predict matches, but how do you explain that Silva didn't get his goal yesterday? This west-ham kid seems promising too.

    1. Hanked
      • Fantasy Football Scout Member
      • 6 Years
      3 years, 9 months ago

      I think he's point may have been wasted on you. Perhaps have another read and then respond, like those comprehension exams in school.

      1. TuttiFrutti
        • 6 Years
        3 years, 9 months ago

        Sorry, but you just got trolled big time, dude!

        1. Hanked
          • Fantasy Football Scout Member
          • 6 Years
          3 years, 9 months ago

          Nice one. Good trolling. I remember when trolling was called being tedious and laughing out loud was called laughing. You crazy kids!

        2. Klostermann
          • Fantasy Football Scout Member
          • 6 Years
          3 years, 9 months ago

          What a positive contribution to the site you make

  10. John t penguin
    • 4 Years
    3 years, 9 months ago

    Based on this approach would you say that bringing in a keeper for courtois such as Adrian therefore automatically gives me a +EV for that place. Therefore is a no brainier with my transfer considering any other move would have element of speculation

    1. OShaughnessy
      • Fantasy Football Scout Member
      • 6 Years
      3 years, 9 months ago

      No need to over think anything here...
      We've got 3 options w/ our FT:

      - Use it on Courtois
      - Use it on Someone else
      - Save it

      So, which of these 3 choices do you think will earn us the most points in the long term?

      1. John t penguin
        • 4 Years
        3 years, 9 months ago

        Long term isn't really that long if WCing in weeks 4-8

        1. OShaughnessy
          • Fantasy Football Scout Member
          • 6 Years
          3 years, 9 months ago

          I should have typed *over* the long term, I guess?
          Like if we could somehow make this decision 100x what would be the best choice of the 3?
          What's going to earn us the most points / is the most +EV play?

          So, if we believe using one of our 38 FT's on replacing our keeper for his one-week absence then we should go for #1.
          If we think another transfer will earn us more points, take #2.
          Maybe the most +EV play is to stand pat? And if so, #3 is the way to go.

          1. John t penguin
            • 4 Years
            3 years, 9 months ago

            Ok I see now
            But the 38 FT (isn't it just 37) doesn't come into does it, if I WC in week 8 then I only really have 6 FT before that so I would only be using one of those

            1. OShaughnessy
              • Fantasy Football Scout Member
              • 6 Years
              3 years, 9 months ago

              Sorry mate, we're getting caught up in the nuts & bolts rather than this whole EV thing...

              Look at the article & see how OP broke down the expected results of his choice...
              Now, try to apply this thinking to your decision.

              Don't just say, I'm going to replace my keeper.
              Try to actually figure out how many points that's worth.
              After that, look at a different transfer & what you'd expect to get there.
              Finally, see what you expect to get from saving the FT.

              The answer will hopefully become more clear...

            2. dribbler
              • 9 Years
              3 years, 9 months ago

              "But the 38 FT (isn't it just 37)"...special rules for me...I get 80 free transfers

  11. Bobbydhino - ⚒ Saint Slav…
    • 9 Years
    3 years, 9 months ago

    Nice read.

    Cheers,

  12. T Money
    • Fantasy Football Scout Member
    • 3 Years
    3 years, 9 months ago

    Fantastic article

  13. invertebrate
    • Fantasy Football Scout Member
    • 4 Years
    3 years, 9 months ago

    Anyone else think UFPL would be well served to use the poker model of pay outs? If they paid, say, the top 10% they'd get an awful lot more players in my opinion. £25k to the winner sounds amazing but £0 to second place doesn't hold that much appeal for people who don't have delusions of winning the thing.

  14. Hoof It and Hope
    • Fantasy Football Scout Member
    • 5 Years
    3 years, 9 months ago

    I know I am well late on this, but wouldn't a better analogy be with an investment portfolio?

    The market judges the value of your assets (players) and reflects that in their price to own. At the same time they earn profits (points) when they play (you don't really get losses).

    So the trick is to make good judgement calls on where the market undervalues a player given their points potential. That way you get both points and price increases.

    Or you could just run a tracker fund. You won't win but you should end up top 10%

    1. dribbler
      • 9 Years
      3 years, 9 months ago

      depending exactly what you mean in FPL terms, if you run a tracker fund you could win and you'll do way way way better than a garbage result like top 10%

  15. ManofKent
    • Fantasy Football Scout Member
    • 8 Years
    3 years, 9 months ago

    Interesting article - I think one key point is 'Do you have all the information needed?' You can look at the average points per game a player scored the previous season, but there's so much more - who else is in the team, have formations/roles within the teams changed, are the player's scores historically better against certain team styles, home/away variation etc. ?

    1. dribbler
      • 9 Years
      3 years, 9 months ago

      stop being boring...misuse of statistics is the bread and butter of this site

  16. Insert your own Joe Kinnear
    • 3 Years
    3 years, 9 months ago

    Good work. Two points. First - pedant that I am, apologies - EV is not just gambling theory but a cornerstone of probability theory.

    Second your EV is last year's average --- but players & teams both improve or deteriorate. Will Harry Kane get the same average score this year? (technically is last year's average ppg an unbiased estimator of this years points)

    By your reasoning we might naively predict the BPL table will be the same this year as last year... oh hang on http://www.sportingintelligence.com/2015/05/26/the-premier-league-not-as-accurately-predictable-as-you-think-260501/

    1. dribbler
      • 9 Years
      3 years, 9 months ago

      you're (accidentally) misrepresenting what the author said

      "By your reasoning we might naively predict the BPL table will be the same this year as last year"

      he didn't at any stage say that you should use last years points-per-game as the only thing you use to predict this seasons results...he was just using that 6.1ppg as an example to make his point about +EV

  17. Netley Lucas
    • 3 Years
    3 years, 9 months ago

    I think what Get-In is saying is by selecting the better players and capping the best you've got, will give you a better score at the end of the season than not doing so. In other words the manager who capped Mahrez is more likely to continue riskier bets, and lose, than if he played it safe. It's a simple method but is often forgotten in our clamour for success... greed.

  18. dribbler
    • 9 Years
    3 years, 9 months ago

    a very fine (and dangerous) article...is writing an article this good a "+EV decision"?

    1. dribbler
      • 9 Years
      3 years, 9 months ago

      I've now realised that there was a page 1 of comments, I've read them all and I think you got some unfair critisism

  19. Daniel - The Veteran
    • 9 Years
    3 years, 9 months ago

    Article of the month. I vote