I don't see a 150-odd point gain on the guy in first in 4/5 weeks as being vaguely possible under all but very very extreme circumstances tbh, but let's not argue that any further.
Ruth_NZ wrote:
The application of this is to have a more concrete means to assess relative player performance on the basis of value rather than purely on the basis of FPL points. If you know - even if you only have a good working estimate - what you need to achieve from £0.5m of your "free" budget then it makes player comparisons easier to do. As an example - Ozil is currently 9.0, Sanchez is 11.5. That means Sanchez needs to score around 2.2 points per GW more than Ozil (or 55 points over the remainder of the season) to be equivalent value according to my estimate. Now you can ask yourself whether you think he will do so. Being able to quantify what you need from each £0.5m - even if it's only a good estimate and not an absolute figure - provides a neutral measurement to help you make decisions like that.
OK, so let me take a step back then and say why I think it's important that the estimates you base this value calculation on are realistic expectations rather than hopeful targets.
I think all of us here are very interested in any calculation that gives a tangible way of working out how much bang each of our 0.1m bucks should be getting us - I'm fully for articles of the type you kindly made for FFS. But, I would say this. I think almost all of us are interested in the value calculation during the season for two reasons:
(1) So we can compare players of two different prices more effectively - who is the 'value' pick?
(2) So we can work out if, say, at which points of the season a 4ph is worth it to save 0.1/0.2m (a common dilemma).
(Perhaps another far more complicated application might be that we want to assess more broadly how aggressive a hit policy we should be implementing in general - do we go the full Finisher1 and aim for TV as a huge priority, or do we play a bit more cautiously? That's kindof all part of (2) though.)
I don't think it's particularly useful to use the calculation to "define a ballpark figure for what a manager needs to be aiming to achieve from 0.5m if they want to finish with a very high OR.", which is what you describe the main motivation of your article as. I feel that would just be saying -
to be a 1k-tiger we must simply put on a 1k-tiger suit. But wearing a tiger suit doesn't make you much better at being a tiger. Knowing that I might need something like 65 points a week to win fpl doesn't particularly help me get 65 points per week - I'm already trying my hardest to get the most points possible. I think your calculation is a very useful type of calculation, don't get me wrong, but I don't think it has good application as a yardsticks for final ranking - we have a good one for that already, current rank. As you yourself do in the quote above, we really imo want it for calculations like (1) (or (2)).
Anyway, ignoring the weird tiger metaphor, my point about what estimates you base this value calculation on is the following. If we want to do say calculation (1) effectively for our teams, we should use in our calculation the best possible estimate we have of where we will finish the season for an important reason. If we put in last year's winners numbers into the calculation (Simon achieved a much higher points per 0.1m of his budget that we can reasonably hope to this year) then this will skew the calculation and make it much harder than is really true for a more expensive player to be worth it over a cheaper one. The points difference given by your value calculation tells us what we need to be aiming for in a given upgrade to be worth pro rata the same amount of points as someone that is near winning the comp might achieve with similar money - but by basing the calculation on such lofty standards we might often conclude that the upgrade is not worth it (because it doesn't meet this high standard) when in reality for our teams it very well might be worth it (as we will almost certainly finish the season with a far lower PP0.5m per week than the winner).
The same thing with (2). If we are working by your method we want the calculation to be based on as close an approximation as possible to our final position because then it makes the calculation accurate for our teams. Maybe a points hit for our team to save 0.1m would be worth it until about week 15 when we do the calc. But by basing the numbers on last years winner this calculation might give us a false answer (with respect to our team) of about GW20, or 25 etc. The more accurate our estimate for our season's points total/finishing position is, the more accurate our value estimation is.
In reality, the difference in PP0.5m per week calculated using an accurate estimate of our finishing positions might not be too different to the one using a much higher than expected finishing position - the error might be small enough for it not to matter. But, without demonstrating that it indeed doesn't matter the choice of unrealistic high (imo) estimated finishing positions to investigate (1) and (2) seems unreasonable to me. I suspect it does only make a small but noticeable difference (as Mav and Sutter point out they expected a slightly lower number so this would agree with them).
On a side note (as this is a separate point to argue) I don't think to attack (1) it is particularly sensible to do the value calculation over all players - it would seem more sensible to do a separate calc for each position as I'm sure the rate at which points potential increases per 0.1m is slightly different for each position. In fact, my biggest gripe is that I don't think it would necessarily be the best route to use season points totals for these calcs anyway - as we've been through, it's hard to estimate and end of season points total because points total vs position changes every season - perhaps the expected points totals for individual picks in a given position of a certain price would be better. For (2) I guess this doesn't really matter as it's a global calculation about how much the cash (not an individual player of an individual position) is worth so estimating realistically how many points we'll finish with probably is worthwhile.