File talk:DJIA historical graph to jan09 (log).svg

Don't you think this logarithmic graph is a little misleading for the main graph?--UhOhFeeling (talk) 21:01, 29 January 2009 (UTC)

What's misleading about it?DavidRF (talk) 03:48, 1 February 2009 (UTC)

A 50% drop in price from the peak would be a tiny drop on the log scale (indeed the dow dropped to about 6k from 11k but it isn't even noticeable). A log base of the value of the dow is meaingless. THe only reason to show it log based is to try to smooth out recent ups and down while promoting historical values. —Preceding unsigned comment added by 58.111.143.218 (talk) 12:29, 8 February 2009 (UTC)

On the contrary. Using a log scale means that a 50% drop will look the same whether or not it happened in 1929 or 2008. Increases/decreases in finance are proportional to their values. That's what "rate of return" means. Plotting things that increase/decrease as a proportion of their current value is exactly what a log scale is for!!! On a log scale, a "constant rate of return" ends up looking like a straight line instead of an exponential curve. If you want to highlight recent events, then you simply don't go back more than 10 years... and when you do that, log-scale and linear-scale plots look much different. DavidRF (talk) 17:06, 8 February 2009 (UTC)

Also, the ~50% drop (from 14k to 8k) is extremely noticable on this very plot. The problem with the plot is that the drop was so sharp that its hard to distinguish it from the vertical line on the right. As more points are added in 2009, the drop will be easier to see. DavidRF (talk) 17:09, 8 February 2009 (UTC)

I've added a new plot which zooms in on the 2000s so that recent behavior is more visible. File:DJIA 2000s graph (log).svg DavidRF (talk) 21:03, 8 February 2009 (UTC)

by the way Thanks david this issue of the 2000s bear markets not being able to be seen was resolved. thanks for resolving it very well i can see the Dow jones corrections of the 2000s very well if possible you could remove the bottom hashmarks of 10 to 20 points braket and replace it on the upper portion with to brakets from 10,000 to 12500 then to 15000 that way the chart can be showing the 2000 to 2020 AD period better. make it a hybred of Log chart like i and you are describing the job is unfinished User:UhOhFeeling has a great idea its not a stupid idea. the problem with a Pure log scale chart is that the log brakets grow exponetially has the index grows forinstance if 100 years from now the dow jones was at 50,000 points the 10,000 number brakets would be so tiny that we could not see them at al so a hybrid number chart is needed after 10,000 points. 76.244.154.251 (talk) 00:31, 5 April 2012 (UTC)

Of course it's misleading, and don't over intellectualize it, if you look at the same graph at face value instead of "log scale" then you see a that it's not until the mid to late 1980's that we start to see a steep climb in the graph. So for over 60 (20's-80's)years, growth was almost a steady state on a very mild slope. Then we get to the last 20 years of balloon economics. Not exactly sure at this point what happened or the logic used to make it happen in the 80's. But no doubt that it happened during Reagan’s watch.-Chris —Preceding unsigned comment added by 128.194.176.179 (talk) 23:21, 18 February 2009 (UTC)

You said: "for over 60 (20's-80's)years, growth was almost a steady state on a very mild slope". What you are telling me is that when plot against a linear scale, you completely miss the Great Depression! You miss the boom of the 50s and early 60s, you miss the bear market of the 70s... all of these things are obvious when looking at a log scale. The stock market crash in 1929 was two consecutive trading days where the DJIA dropped 30.57 and 38.33 points, respectively. You think two days like that would make news today? Do you think a point of DJIA mean the same thing today that it means in 1896? A log scale makes all percentage drops appear equivalent and you *can* see increases in slope in the 20s, 80s and 90s booms. DavidRF (talk) 03:21, 19 February 2009 (UTC)

The fact that the recent market drop isn't very noticeable on a graph like this should be reason alone to not use this as the main graph. Don't over intellectualize. I mean the market is cut in half and it is barely noticeable. Most people, myself included, will just be somewhat confused by this graph. It should not be the first graph of the article.--UhOhFeeling (talk) 21:56, 21 February 2009 (UTC)

Its no longer the first graph of the article. A graph highlighting more recent behavior of the index is now the first graph. This graph is labeled as a "historical" graph.

The scale was never the "barely noticeable" problem. That problem is due to the fact that three months is a very narrow slice of 110 years and also the "fill" under the points. The steep drop is actually merging with the vertical drop on the right hand side of the data. As time goes by, this drop will look at least as bad on this plot as other 50% drops in history.DavidRF (talk) 22:18, 21 February 2009 (UTC)

Uh...OK--UhOhFeeling (talk) 06:31, 7 May 2009 (UTC)

i read in World almanac of facts that the DJIA reached a 100 point mile stone in 1906 and i dont see this on here and it should be visible on a log scale.76.244.154.251 (talk) 15:00, 4 April 2012 (UTC)

Interesting. I took the data that was already in the file from the previous editor (Lalala666). The data comes from this website (not that exact page, but a csv file on the same site with similar data). There is definitely a panic in 1907 but a drop from 60 to 40 instead of from over 100 to 53 as described at other websites: . I don't know the cause of the descrepancy.DavidRF (talk) 15:32, 4 April 2012 (UTC)

thanks for telling me that. just out of curiusity when did the Dow jones actually cross the 100 Point mark for the first time? try to give a exact day or week so i can see if this trully is a descreptancy or even a clerical error on user lalala666's part. 76.244.154.251 (talk) 00:39, 5 April 2012 (UTC)

I don't know. I am finding conflicting data. The Closing_milestones_of_the_Dow_Jones_Industrial_Average gets its information from and says it crossed 100 in 1906. This plot gets its data from which says that it didn't cross 100 until September 1916. I don't know the reason for the descrepancy. I found one other website for each. If someone can provide the accurate daily data then I'll change the plot to match that.DavidRF (talk) 02:16, 5 April 2012 (UTC)