Worst Graph Ever

This may be the worst graph I’ve seen in my professional career.

bad graph

  • From a distance it looks like all the lines are heading down. I saw this on a dim screen. I initially thought the takeaway was that all the scores were going down.
  • It is 3-dimensional for no reason. There is no extra information, only graphics for graphics sake.
  • There is no reason to have a legend, since there is only one kind of thing being graphed, but one is present. And the labeling doesn’t match the title of the slide.
  • Differentiation of horizontal scale is essentially even – they all have the same score. That fact is not emphasized. Instead, the horizontal scale is manipulated to emphasize the difference (doesn’t start at zero), but the stretching isn’t enough to show a difference. The manipulation didn’t even help. This leaves the labeling in non-intuitive numbers, 5, 7, and 9.
  • There are horizontal axes guidelines for no reason. And worse, the lines go right through the labeling of the individual values, making the graph cluttered and hard to read.
  • There are axes guidelines on the vertical axes, when the values are just the classes. Just horrible.
  • The values are in no order. They are not sorted by results, they are not alphabetical. There should always be some logic to the way values are stacked.

Mainly, the graph serves no purpose. It gives no actual information. I have no idea what I, as a consumer of this information, am supposed to get. Perhaps it was “All classes are high value to participants”. That information could have been given by writing that exact statement, perhaps with “(all were in the 8-9 out 10 range as judged by participants)”. Since the biggest difference was 0.15 out of 10 points (a piddling 1.5%), perhaps the point was “Our survey showed that no matter what we teach we get the same scores.”

Or perhaps the key takeaway is, “Human Resources should never try to present analytic information.”

Poker Update: Poker Stats

I’ve been playing a lot of Rush Poker at Full Tilt Poker. It records statistics for you. It doesn’t actually log the hands in detail, but gives you some gross metrics. Here is a sample from a recent session:

poker stats

The problem is, I don’t know what to do with this information. What is good, what is bad? I can only see two patterns:

1) I play a lot of hands. 15 out of 89 is a lot, just over every one in six hands. This is true, I’ve been playing a more aggressive style lately.
2) The percentage of hands won goes up as the hand goes on. This seems like a good thing. You have much more money invested at the later stages of a hand, those are the ones you want to win. I don’t mind losing several small pots pre-flop if I can get one big payoff on the river.

Readers – what do you think? What other patterns or indicators should I be keeping an eye on?

Links o’ Interest

Another TSA outrage

Your brain on meth. But seriously, Meth can mess you up good.

Joke book for math people

The good old days

Marching band win

How did I ever miss this? Jerry Seinfeld rips an ignorant Larry King

This is TV

Stealth fighter

We’re not in Kansas anymore…

If Ikea made instructions for other things

Now we finally know, who was on first

The laughing makes me laugh too.

Now that’s acting

Inception, done in real-time

Perspective from Harry Potter

The perfect obituary

It’s a trap!


Child stars: Then and now


Camping invite

The dead bodies of Everest

The Kids Have Another Classic

I came out of the shower. I was drying myself. The eight-year old says, “Daddy, you should always dry off your penis last.”
“What? Why?”
“You should always do it last so the towel doesn’t get the rest of you.”
“You mean because it’s dirty?”
“Even right after a shower?”
“You mean, because the penis is somehow always dirty and filthy?”

There is a long pause. Then the five-year old chips in.
“Not if it’s made out of candy!”