Decay channels of the Higgs boson I am currently reading a letter about the coupling of the Higgs field to fermions. From this letter, I have gathered that they have provided evidence for this phenomenon. However, another article in 2018 says that it was only then that the Higgs to bottom anti-bottom pair decay was observed.
So, in simple terms, what exactly did the first letter provide evidence for?
 A: Experiments in high energy physics use statistical definitions in describing their data, and deciding how a fit of data  to a hypothesis predicted can be compared with the theory. The usual standard deviation measure for all statistical quantities. This is usually done using monte carlo methods: generating a large number of artificial events that fit the specific theory , and comparing them with data, to get the standard deviations of agreement.
From the abstract of this 2014 paper:

the study of Higgs decays to bottom quarks15 and τ leptons16. Here, we report the combination of these two channels, which results in strong evidence for the direct coupling of the 125 GeV Higgs boson to down-type fermions, with an observed significance of 3.8 standard deviations, when 4.4 are expected.

In general in high energy experiments one desires a 5 standard deviation  effect to accept that the data fit the theoretical expectations. They have combined two channels to increase the credibility of the hypothesis statistically, but it still is not enough to reach the gauge of 5 sd.
So this is evidence that something might be there, for further data gathering.
Data gathering on the channels four years later, increases the statitistics for one  channel, to 4.9sd,  and

The combination of this result with searches by the CMS experiment for H→b¯b in other production processes yields an observed (expected) significance of 5.6 (5.5) standard deviations and a signal strength of 1.04±0.20.

the desired goal of declaring a solid fit . The first paper gave a hint that may be  the theory would fit the data, given enough statistics, which it did with the data provided to the second paper four years later.
It takes time to do experiments. One studies not so good statistically evidence, less than 5sd, in the hope that with more data one would reach an agreement with theory. It points the way, but sometimes more data smooth out the "evidence" to nothing.
