Comparing Cats: A Discussion of Similarities & Differences

[GuateGojira]

About methods and averages:

Resently there is a confusion about the methods used in the averages, specially by people that is "new" on the business. I will try to share a light on this point.

Before to start, please check this information, in post No. 361: https://wildfact.com/forum/topic-modern-...rs?page=25
This is about the problems in using old data and the reliability of the figures uses on averages.

Also this link in the post 390: https://wildfact.com/forum/topic-modern-...rs?page=26
This is about the exceptional specimens, and this most be applied to lions too, specially with those two obscure figures of 750 and 800 lb from two lions quoted in a newspaper that is more unreliable than the old figures from Amur tigers of Baikov.

Ok, lets fo to the main point:

The method that all of us is using since the old days of AVA since year 2003 is the "simple average" of all the figures available. Is the easiest and faster of all the methods. We only need to take all the figures available and make an average, simple as it is. Now, the issue is when we don't have the single figures but only averages and this is an issue that we found specially with some lion and tiger populations. So, what we have done is just to make the simple average of all the average figures and get a result, we know that this is not perfect but has been used and provided reliable results since those days. Now, I see that two posters (one of them banned twice) insist in put doubt about the reliability of the method and suggest changes, but how reliable are those changes? Are they necesary? Will they produce different results?

The first person that tried to provide other method was, as far I remember, @tigerluver in the AVA forum and he proposed the "weighted average", which means to include the "weight" of each value in the population. This means to take in count the size of the sample used, but that method was actually never used in any of the calculations until this day. @Shadow explained in this moment his opinion about the importance of its aplication, but the issue here is that we are normally using small samples and that is a fact that we can't avoid and some people had used this as an excuse to label the samples as "unreliable" just because do not fit they "agendas". Even using all those values we had never reached samples that actually represent the entire populations. Interesting while this is an "issue" for the fans at forums, while it is something that do not disturb the real scientists to much, as they know how dificult is to collect the morphological information of wild animals. So how reliable is to use just 3 tigers from a population of at least 40, at the end the weighed of the values by the sample size will also provide results that are not 100% true.

Now, the other method that I tried to use is the "weighted average" using the value of the size of the populations like @Ashutosh mentioned. In fact, I tried to weight the average figures with the representative size that each group means for the entire population. For me that will be more "reliable", but the issue is that we do not have historic information about the population size of each region, so we can use the last information provided by the census made in modern years, but that will represent the tiger population at 2010 or 2014, not the population at 1900 for example. Other problem is how relaible are the population figures? Check how many times the figures of Sundarbans had been inflated from 100 to over 400 animals! I am aware of this because Dr Karanth had mentioned this many times. On the other side check the values from Nepal and Russia, which are more reliable and based only in adults specimens. So, also using population figures will create problems.

As we can see, using any of the methods proposed at the end there are problems that will cause bias in the results, even when we try to make it the most accuarate possible and this is NOT something new. Those that actually followed these debates that started in AVA knows that the results are allways debated because there is an external group of crazy and immature people that still want to continue with the "Lion-vs-Tiger" debate, specially with a group of 3-4 lion fanatics that at this day still insist in presenting results that are biased and completelly unreliable. Sadly, these new posters do not know the history behind all this efforts in presenting information and reliable/unbiased results and only critize without knowing that there is a big history behing all this.

So, knowing this, it is reliable to use the "weighted average"? Well, the funny thing is that I made it with the Bengal tigers and the average of all the populations (n=166, including Sundarbans) is of 200.3 kg, just 3 kg more than the simple method. But remember that it is now stablished that the samples includes several subadults tigers from 150 - 180 kg that were included by hunters in they records, so that is why I estimated a figures of c.200 kg for Bengal tigers overall, including Sundarbans and also the subadults (if some one insist in using my old paper from from 2010 or 2015, I don't even remember, know than that paper os OLD and do not have all the new specimens that I have now, so don't insist in using old information to battle with the new reults). When I made this to the lions, the figure of the lions from Botswana changed from 184 to 190 kg, but the figures from Namibia also changed from 197 to 190 kg, so it standarized all the values for the lions in Southern Africa at c.190 kg, but the figures of the lions from East Africa droped fom 174 to 173 kg. So, if there is a person there tryting to diminish the weight of the tigers using this method, actually is the one of the lions that will result in even more diminishing.

Now, how necesary is the use of these methods to stablish the averages? Is this even this important? Well it seems that for people like Dr Yamaguchi it is not, and we can see this in the book from 2010 (Chapter 4) "Tigers of the Woeld" when he says that in order to calculate the average at species level he used only the simple average of all the subspecies or populations, he did not made any "weighed average" or a "population weighing", he just put the values and made a simple average of averages, and incredible my values match those from him.

Other thing is that other experts only used ranges and not averages, which is a lot easier and we can see this with Mazák, Karanth and Jhala, which do not botter to make all these calculations, they only stablished ranges and that was all.

Finaly, is actually the "average" the most important measurement? I think not entirelly, because also the Mode is a very imporatant value that is completelly forgoten in all these studies and debates. I noted this situation when I investigated the size of the orca and I found that the mode was used in several of the studies, showing how important it is. Check this image, is self explanatory:

*This image is copyright of its original author

In this case the mode is fundamental, as it will show to us wich is the most common weight in a population. Now the problem here is that we can't use it in all the populations because in some cases, like Brander or Smuts, we only know the averages, ranges and sample sizes.

So, what I am going to do? Been honest, NOTHING. In this mometn I don't have the time to enter in this type of futile discutions, so I will leave this to you, the "experts" that want to make the difference, so you can discuss, fight, insult, and walk aaaaalllllllllllll the way to here that the old posters had already done, and when you get to a conclution with VALID arguments, call me again and I will recalculate the figures with this proposed new method, ok? But take in count that we are working in some cases with only average figures and samples and in other cases (the good ones) we have the all single values.

If some one here gets offended with my words, please know that it is not my intention, but understand that in this moment I am really tired of this back and forth with the same thing again and again and again, and I will want to finish this thing of the sizes ones for all (that was the objective of the comparative images after all).

So call me when you are ready, then we will talk about this again. 

Good night. Reply