Netserk wrote:You cannot put a number on it a priori, absolute or relative (just take the hypothetical scenario where every rider scores exactly the same as the year before). You're average rider will cost 227. If the (average) return for many of the best picks around that price is 400, then the value of a rider costing 0 with a return of 100 will depend on how larger the return (in absolute measures) of a more expensive pick is than that of the average rider. In this case, you only have to find a single rider with a cost of 454 (or less) and a return of more than 700 for the swap to be good. If there's two above average priced riders, they will have to have a combined cost of 681 (or less) and a return of more than 1100 for the swap (of three average riders with the cheap rider and the two more expensive riders) to be favorable.
In a year with many good very low cost riders, the expensive riders will not have to be as good as in other years, and vice versa, just like if the return of an average cost rider is high, more of those will be able to out-perform a more diverse selection.
So in short, both relative and absolute return is important for all picks, and the market decides how those two needs to be combined for a pick to be good.
*Sometime, I will take a closer look on last year's game and the market of the most picked riders (probably top-100) and analyze how good picks they/(some of them) were, or rather how much of a return a rider would have to have in the different price ranges to be a contributing factor for a top team.
Yes, to put it in a more simple way: For any N amount of riders you have (in practice I work with pairs or trios, for simplicity), you need to be sure that there is not another combination of N riders at the same price which together will score a higher (probable) amount of CQ points (in absolute measures). Which is what I wrote last year. Then of course you need to weigh the points ceiling/floor for a rider against his average expected score (if he rode an infinite amount of seasons) and decide how much risk you want to take. I think hakkie2's theory is much closer to this than the notion that every rider needs to double his points.
On another note; you seem to take an interest in this game, so it's a pity you didn't submit a team. I'm sure you would've done well.
Edit: Oh, and about the analysis for last year that you're planning. I remember Skibby once did an analysis on the optimal team for one of the years, and calculated the "penalty" for not including each rider, i.e. how much less points the optimal team without that rider would score. That's kind of what you're thinking about, isn't it?