Is Science in a Constant State?
Modelling – If your assumptions are wrong and your formula is wrong you’re almost guaranteed to be wrong, but sometimes you will be right. Modelling, especially high-speed computer modelling is now the preferred option to experiment by most professionals. You can infer so much more so much faster. With a model for a true value of 0 you can get minus infinity to plus infinity as a result depending on choosing the wrong parameters and assumptions. But one practical experiment is worth a hundred theories, one direct and actual measurement worth a hundred estimates. To mix models, experiments, theories and measurements is a recipe for disaster if you give them equal significance.
Statistics – Choose your fiction and you can find statistics to endorse it. But on rare occasions they will be correct. If you go into finding or working out statistics with desired goal or view in mind you will be able to find statistics to support this and some to reject others goals or views. Statistics are there not to confirm things are a certain way, but to find out what is going on. If you have preconceived ideas about this you will probably implement them in way of finding the right statistics to support it. If a thing is 50/50 in likelihood you can find 60% of the time statistics that support your view. If you look and subsequently find for only 30% of the time it is against your view and 10% of the time it is unclear, you are twice as likely to be proven right as wrong. The true value may be still balanced and equal, but the statistics prove you are much more likely to be right than wrong.
The Future – Choose your future and you can find models and statistics to endorse it. Sometimes it comes true. If you don’t truly and exactly know how exactly a thing works you can’t tell what it will do and any future concerning it. If you take the number of how many predictions there are and how many actually come true, only the correct ones will have the significance of being right and appearing above all the rest. This is called survivorship bias. Business is littered with survivorship bias, the kings of industry and business assumed by others and considered by themselves there because they made the right decisions and were clever. The truth; they are there and successful because they are there, no characteristics putting them there. Similarly, for a lot of professions, especially the media, where you have the self-defining reality stars who are ‘famous because they are famous,’ with no other obvious characteristic.
Science – Science is inexact, takes a long time to work correctly and is full of mistakes. Considered by many to be exact, calculated and self-correcting. The actuality is belief, statistics, modelling and direction with estimated figures and educated guesswork prone to keeping it unchanging. Nothing is exact, nothing is certain, nothing is sure, always open to alternative ways of thinking and working it out, but really a collection of the best guesses. If it isn’t checked by non-dependent means don’t assume something is true. If A is dependent on the results of B and B is checked against the results of A, neither A, B or both are guaranteed to be true. If you add C dependent on those, A, B or C and possibly all may be wrong in a dependency cascade which is called a paradigm shift.
For example: With each new measurement the Hubble constant (it should really be called parameter as it changes over time) changes. Originally around 500 km/s/Mpc in about 1929, and either 50 or 100 km/s/Mpc in the 1960’s, it is now considered between 68 and 72 or 73 km/s/Mpc, all being considered exact values, but more modern measurements tend to have a centred variation of about 5 km/s/Mpc with some being as much as 20 km/s/Mpc. The Webb telescope may eventually change this all and if we ever go anywhere outside the doorstep the reality could be quite different.