I had an initial instinct to vote against the Omnivore changes, because I didn't like the idea all the negatives... and the posting of the new "actual" scores bears me out.
But now I'm wondering, just common sense suggests it's statistically unlikely that a team will optimally deploy their players at all times. There will frequently be a benched player who would have scored more than someone who was played.
So, given a FUNHL where no trades were made, we would expect all teams to finish with fewer points than they could have optimally made.
My curiosity is: can anyone guess as to how many points a GM is likely to miss out on? What's the Omnivore's likely "0" line?
Richard.
3 comments:
The first thing I'd say is that the Omnivore won't be affected as much by the week to week coaching decisions as it is by horrific trades - exactly for the reasons you hit on. Coaching decisions are felt week to week, but a poor trade involving one of the players you start with can end your Omnivore season once and for all.
The absolute killer of these tend to be goaltender trades, ie; the year I was riding Belfour and Hasek and going into a waiver draft had to decide who to move. I dealt Hasek to Bob for two dressing defensemen (one of whom was Chelios) and a spare forward. Hasek had 4 shutouts in the next four weeks and I ended losing the pool in the last week of the year - to Bob. But I digress.
The coaching decisions could be monitored by a sharp eye each week to get the answer for, but I would hazard the following guess;
Given that a poor coaching decision is most likely to be made when the 4th line player exceeds the 3rd line players points, and that overall the difference between a 3rd and 4th liner over the course of a season is say 10 points for forwards (3 of them), 5 for defenseman (2 for the position) and as much as 30 for the goaltender. Divide these numbers out and you get an average margin of error for the 27 weeks of....(oh dear god my head hurts).... roughly 70 points over 27 weeks which is something like -2 and change for your team per week based on my guesstimates.
First of all I was quite pleased with the recalculated Omnivore as it reflected a fair balance of improved and declining teams (since the bulk of the change should be from trading, and that means one team has to lose while the other benefits in a closed system).
While it was fun to read through Cam's mathematical convolutions, it is interesting to reflect on the fact that if you take the sum total of the Omnivore differences (ie all the positive and negatve scores), the net change for the pool as a whole is PLUS 73.
True, we do have a small infusion of points through the WDs, so it is not a totally closed system, but what the numbers tell us is tht the pool as a whole (through line-ups, etc) garnered 73 MORE points over the season than would the original 12 drafted teams (with prospects) left on autopilot.
While there will be instances of dressing the wrong guy in a given week, there will be a number of offsetting times when the right guy gets played (eg Rick Nash counted on the baseline, but he was hurt for some time and hence benched, and his replacement would have thus increased the Omnivore difference for that team (notwithstanding the trades) so it runs both ways). Similarly, plus-minus cuts both ways but hopefully overall the pool as a whole gets more + points from plus-minus than -.
All in all, a net change of 73 points works out to an average of about 6 points per team - I can certainly live with that...
Post a Comment