Testing Scientific Hypotheses Using Specified Complexity


Take, for example, a deck of 59 distinguishable playing cards, randomly shuffled and placed on a table in a row.

@Agauger Doesn’t a deck of cards have 52 cards?

1 Like

@Patrick, You are correct. I reported that.

Every ID argument regarding probabilities always uses the simplified method of multiplying independent probabilities for random events together to get the overall probability. They do it for 52 card deck draws, they do it for coin flips, they do if for 150AA proteins. What the IDers never do is to provide probability for conditional events, iterative processes which use feedback to determine each step of the causal chain. It’s the difference between playing poker with keeping the 5 cards you were dealt or using feedback to keep good cards and redraw.

Evolution uses exactly such feedback where natural selection is in effect determining which cards to keep for the next redraw. Until the IDers begin modeling actual evolutionary processes using feedback all their “specified complexity” calculations are so much garbage.

I very much want to play poker with whoever wrote this. Easy money

They still haven’t learned how to do probabilities correctly. For example:

The probability of any 6 specific people winning the lottery is just as improbable as the same person winning 6 lotteries. When you look at the results after they happen, as ID supporters do, you can make any outcome look improbable. For example, let’s say that the chances of winning the lottery is 1 in 100 million. The last 6 winners are John, Susan, Frank, Jeff, Rusty, and Sam. The chances of those 6 specific people winning the lottery are 100 million to the 6th power, or 1 x 10^48. The chances of those specific people winning is so low that it shouldn’t have happened, right? Wrong. The arrow of time guarantees that extremely improbable events are guaranteed. By simply selling tickets and having a lottery drawing you are guaranteeing that extremely improbable results will occur.

The same for cards. If you shuffle the deck and lay out 52 cards, the chances of getting that order of cards is 1 in 52!, or 1 in 8 x 10^67. According to ID supporters, we shouldn’t be able to shuffle a deck of cards and deal 52 of them.

1 Like

I think this where they will bring in the subjective “specification”. Imagine if all cards were laid out after being shuffled and they were all in order and in the same suit. But of course I can make any arrangements meaningful to myself. To me and my sister a jack of spades,jack of hearts, two of diamonds, 10 of clubs and 3 of hearts is just as meaningful as a royal flush. That’s why CSI is incapable of really making a design inference

1 Like

To be fair, the article makes the exact same point in the subsequent paragraph.

True. Nevertheless, if the same person did win 6 drawings in a row, I suspect it is correct that an investigation would be triggered. Just as I suspect one would be if your hypothetical John, Susan, Frank, Jeff, Rusty and Sam were named in advance and then went on to win.

Maybe there are better ways of describing p values and rejecting the null hypothesis for a lay reader (feel free to share examples), but ultimately I don’t think with the introduction that they are trying to say anything out of the ordinary about those topics.

1 Like

The key issue is “named in advance”. For data such as DNA sequences, they are looking at the results afterwards.

Coming at it from a different angle, we could use human genetic variation in their model. If we took a simplistic view we would predict even distribution of mutations within the patterns expected from known biochemical pathways. However, what we find is some areas where there are fewer changes. This pattern becomes even more apparent as you compare human genomes with genomes from other species.

Of course, we already know what causes this: natural selection. At least on the surface, it would appear that natural selection produces specified complexity in their model.