Basic Settings

Evaluation Score (Eval)

  • What it is:
    • This is how the model rates a plan based on the points predictions (EV), time decay setting, risk setting, estimated AutoSubs etc.
  • Why use it:
    • Effectively, this provides information on how good a plan is beyond just pure EV data
  • How to interpret:
    • Quite simply higher scores are better, it is not the same as raw EV

FT Value

  • What it is:
    • This is the default value we assign to a free transfer. An otherwise burnt FT in the upcoming GW is always valued at 0 points, regardless of this setting.
  • Why use it:
    • It reflects the added value by rolling a transfer and provides a generalised value of assumed future moves. If set to 0.0, solvers will happily piece together intricate moves for marginal EV gains
  • How to use:
    • Past studies suggest a value of approximately 1.5 is recommended, however you may want to adjust based on situational information
    • Lower FT values make the solver greedier/intricate, higher promotes more towards skeleton plans where only key moves are highlighted

Solver Speed

  • What it is:
    • This is a setting that allows us to accelerate the solver, or alternatively carry out heavier solves. Heavier solves may uncover some extra EV gain (though this cannot be guaranteed).
  • Why use it:
    • Fast Solves:
      • We may be close to the deadline and some significant news occurs- a fast solve can work nicely here
      • We may be using a device such as a phone which might struggle with heavier solves
      • We might be cycling through GWs and re-solving
    • Heavy Solves:
      • This is valuable for problems such as WCs
      • If you are in absolutely no rush
  • How to use:
    • The fastest setting should carry out an ~6GW solve depth in the scale of ~5-10 seconds
    • The medium setting should carry out an ~6GW solve depth in the scale of ~30-90 seconds
    • The heaviest setting should carry out an ~6GW solve depth in the scale of ~3-10 minutes (go make some coffee)
    • It is suggested to leave the solver tab open in the browser, switching tabs during a solve will greatly slow down computation

Time Decay

  • What it is:
    • It is a compound multiplier that is applied to the evaluation of each future GW
      • A setting of 0.85 values GW1 at 100%, GW2 at 85%, GW3 at 72%, GW4 at 52% etc. 
  • Why use it:
    • This setting reflects that increasing levels of uncertainty exist in projections and plans the further into the future we look
  • How to use:
    • Past studies suggest a value of approximately 0.85 (0.80-0.90)  is recommended, however you may want to adjust based on situational information.
    • The lower the setting the greater the priority on short-term plans and vice-versa

Solve Depth

  • What it is:
    • The number of GWs forward the solver will consider making transfers in
  • Why use it:
    • If we want to revert to a single period solve (can be a good idea for very WC robust plans)- ie. set to 1GW
    • If we would rather not make “dead-end” moves in the last couple of GWs- set it a few GWs short of the prediction horizon
    • Alternatively, we may not care about the above and should it to it’s maximum
    • If we want to reduce solve time, this setting can be of great help combined with Solve Speed
  • How to use:
    • Simply input the number of GWs you want the solver to consider 

Advanced Settings

Burn Value

  • What it is:
    • This is the default value we assign to a future “burnable” transfer. In the upcoming GW a burnable transfer is always valued at 0.
  • Why use it:
    • As per FT Value, this setting can be used to create a robust skeleton plan or a more intricate literal plan
  • How to use:
    • In general, it is advised to set as per FT Value, some users prefer to set this to 0.
    • A low setting makes the solver output greedier/more intricate, higher promotes more towards skeleton plans where only key moves are highlighted

ITB Value

  • What it is:
    • This is a multiplier which is used to reward remaining bank in future GWs (not applied for the upcoming GW)
  • Why use it:
    • It reflects that the bank opened up in future GWs can be utilised in future transfers, making enabling moves more appealing in the live GW
  • How to use:
    • The scale of this setting depends on the value of each marginal million and the expected utilisation rate of any remaining bank. Values in the range of 0.10-0.20 would generally be recommended

Max GK £

  • What it is:
    • The maximum investment price you are willing to make on a GK
  • Why use it:
    • Often we want to avoid using up slots from premium teams on GKs, equally it’s fun to spend money on attackers
  • How to use:
    • Simply input the maximum price point you are willing to invest 

Risk Position

  • What it is:
    • Currently, risk in the planner tool is proxied as, Exposure= Ownership x EV. From this, your team’s overall exposure to risk is then calculated
    • As ownership is even more changeable than EV this setting decays even more quickly over time
  • Why use it:
    • This setting can reflect your personal preference, sometimes we want to play safer or entertain increased levels of risk. Generally, we don’t need to take a risk position unless we want to (it costs some level of EV to buy a risk position)
  • How to use:
    • 0.00 is the default setting and works as a solver that gives no consideration to risk. Typically, pure model approaches lead to reasonably high-risk plans
    • I would say that 0.00-0.30 is the general range for normal risk considerations if we wish to play the game more safely
    • Setting negative values increases risk. I would suggest being more cautious with this setting, given the baseline already experiences risk (unless you are in a game-theory driven situation which requires major risk)

Max Team GK+DF

  • What it is:
    • This maximum number of GKs & DFs from the same team you are willing to pick 
  • Why use it:
    • Often, managers want to diversify defensive picks
  • How to use:
    • I’d suggest usign a setting of 2 or 3, based on personal preference

Sub Weight

  • What it is:
    • This is a multiplier applied to estimated AutoSub contributions
  • Why use it:
    • Some managers may want to disincentive bench use
  • How to use:
    • It is recommended to leave this set to 1.00, which uses the probabilities determined by the algorithm