Oh, I see now, it's prediction from finger painting. There will be some orderly development of totally identical subprojects which will give us the absurd linear production decline. There will not be any tendency to develop more now instead of self-rationing and the forecasters have crystal balls that give them exact information about the production capacity of any development in the future.
Hardly finger painting. The way these projections are put together is by taking current production and curve fitting to type curves for that particular play making the assumption that economics are positive going forward. The planned new wells are modeled with an average type curve. This methodology has proven to be fairly accurate when it was applied to conventional North American gas back in the seventies and eighties. Shale is much more predictable than conventional plays, there is little in the way of geologic risk involved and the statistical mean of wells in a given area will approach the average type curve which has been demonstrated by numerous producers. The gradual linear decline of the entirety of the play will not be governed by the decline of an individual well it is governed by the fact fewer wells are being drilled each year. This is pretty basic math. Plot it up for yourself if you don't believe me.
As to there being no tendency to develop more although the well count has dropped drilling continues, it isn't going to suddenly stop especially so in the Marcellus where much of the play has a breakeven cost below $2/Mcf. You confuse the decline in a field where all the wells are drilled and produced at a max rate for a plateau period after which declines set in. No new wells are being drilled so the overall decline will look mostly exponential with a late hyperbolic portion. This is completely different than when wells are continued to be added through time albeit at ever decreasing rate.