Hey, yes this video walks through how to do it.
Basically you set the plausible weeks you might want to use each chip in- it then iterates through all the combinations
youtu.be/REoJEED-Qqw?...
There's some chance it could still be rescheduled to fit in that GW, unlikely scenario though
Hey, yes it's a new set of drafts every week
The reason for this is that many drafts will end up being incoherent with the actual moves you played before the deadline- so it resets to your base team each GW
Hey, this can happen for 2 reasons:
1- If the tab is changed while the "loading" screen appears.
2- Some browsers/devices seem to trigger this, so trying a different device or browser might solve it
If you send me your team ID, I can manually get the data for you too
Ludicrous season, well done Stefan!
Just need to refresh the background sample- the NFO-BRE fixture was likely too tight to the next GW deadline, so the data wasn't there when that GW34 background calcs were made.
Will be synced back up following the next GW deadline
Basically all his mins are as a winger at the moment
Going to respond here too (in case any one here stumbles on the same)
Issue relates to an "in conflict" tag and how it interacts with the code.
The 1/99 fix should sort this out for you, until I patch in a fix to get it to stop searching for a fixture removed from the "uncertain" pool
This is a game? I thought we've all been engaging in long running self torment
Hi Mike, it will depend on the timing
For example of you choose GW30 for AM and GW31 + GW32 for WC, there is actually no valid chip plan because the options are in conflict
Two addendums now that I think of it:
- Bear in mind the found chip plans are based on full knowledge of the fixtures. In reality, there will be some value in holding for announcements- so avoid taking too literally
- "Exit Solve" doesn't seem to be linked in, will get to it but not priority atm
I intend to extend the "Chip Plan => Search" capability to regular solves, though I've just completely ran out of steam at this point in the race to be ready for GW30
The next weeks will be aimed at smaller, nagging fixes that are required and also at longer term projects.
Binary solves was a much requested feature over the past month, the concept has been built out from the analytics discord community over the past months- my understanding is that @jonnyc.bsky.social, OG & Joose all contributed to the various elements so huge credit is deserved on their side!
Chip Plan > Search: You can now choose various GWs for as an option per chip. This is a two-stage process, using literal settings (decay=1.01, ftval=0 etc.) to select chip plan and then reverts to user settings to solve for the root move
Chip Plan > Fixed: Default use, from main planner page
Binary Fixture Solves: Every possible combination of fixture sets is generated (and will update on user input).
For each fixture set, EV is recalculated for each binary schedule and solved individually.
(% weights in tables based on schedule likelihoods)
One more big addition to the fplreview.com toolset...
- "Binary Fixture Solves" integrated into the linear optimiser
- Enabled "Chip Plan => Search" with binary solves
This allows users to solve for all possible fixture permutations and gain insight for transfer & chip planning
Details below,
It was changed as the feedback found the numerical input a bit unintuitive, but I can see where you coming from it is nice to have that extra granularity
I think the best solution for all will be an annotated slider, to give full control and intuition for input
Just updated that now btw, should be settable up to 1.05
Glad to have it across in time, just about 😅
Ah, thanks for letting me know that- I'll update that in the morning to let the setting go to 1.01/1.02
Thanks mate!
I would have liked to make these additions earlier, AM was a bit of a bomb going off alongside running into extra tasks with the day-job & some non-work stuff too which made the past months a bit relentless
Fortunately we're back to normality now, and there's plenty of ideas and upgrades to come!
Downloadable Results
Downloadable results are a small addition, but beats opening up the snipping tool- I think it would be nice if it printed the solver settings used too to add that bit more context so that can be something for the future
(applies to both solvers)
Live Output + Exit
This addition really helps with running a few solve lines, as it was a bit frustrating waiting for longer solves with no info. The same goes for sensitivity analysis, as the lack of updates with the non-linear Transfer Solver can leave you wondering if something has gone wrong
Solver Setting Interface
This upgrade should make working with the settings a bit more usable as they were a bit cramped as it was.
Equally, the revamp has also made it easier to add settings in future and I suspect they'll be categorised into groups in the future.
(applies to both solvers)
Single Solve Chip Plans
Previously solves were segmented by chip plan, so for example WC32 & FH34 would be
Solve 1: GW29-31
Solve 2 (WC): GW32-38 (exc. GW34)
Solve 3 (FH): GW34
Now rather than having 3 solves, as in the Transfer Solver, there is just 1 solve required that will cover it all.
Sensitivity Analysis
Effectively this is creating a number of simulated alternative worlds with some random noise injected. The noise can be controlled for each xMins, player & team ratings and solver settings.
This solves the top line for each world first, before moving onto the next etc.
A few upgrades to the sites linear optimisation tool in advance of GW29
- Sensitivity Analysis added!
- Chip plans now solved at once, rather than sequentially
- Solver setting interface updated
- Live output + ability to exit solves early added
- Results now downloadable as a .jpg
Details below
Have it added in now, with a refresh of likelihoods based on latest market data
Yes, the fixture announcement will be reflected on-site in ~40 mins
