And the cows have all gone running home to put their coats on

[austin_dern]

Oh, yeah, there's the monthly Price Is Right report. December 2009 had a week of reruns, oddly running from Christmas Day through the 30th of December, so that the 31st could see a special New Year's edition with the ``favorite prizes'' of staff and board participants from the year. As they like it. In the Showcase Showdown, overall, the first spinner won 12 times during the month; the second spinner 8 times; the third spinner, 16 times. For the season so far, this gives the first spinner 42 wins, the second spinner 49 wins, and the third spinner 51 wins. This is consistent with the revised hypothesis about the third spinner having an advantage, with the early horrible performance of the third spinner looking more like a freak event.

The lowest winning spin in the Showdown on record now is 10 cents, achieved the 24th of December, although that was after the first two spinners went over a dollar. The lowest competitive winning spin remains 30 cents, which turned up on the 2nd of December when the second and third spinners tied on that. (The first spinner had a meager 10 cents.) The lowest competitive spin that wasn't won by a spinoff remains 50 cents, which most recently came up on the 2nd of December.

In the showcase, the first-revealed was the winner 12 times, the second-revealed 3 times, and there were three double overbids, the first ones of the season. For the season overall the first-revealed has won 54 times, the second-revealed 14 times, and there have been 3 double overbids.

Incidentally, on Let's Make A Deal on the 30th of December one of the Zonk prizes was ``a family of coatimundis'', which announcer Jonathan Manke thought was particularly hard to say. I'd like to include a link to it, but the supposed video link doesn't actually work. Anyway, it's excerpted in the clip ``Purse or Curtain?'' with air date 12/30/09, at least for the next week or two, at http://www.cbs.com/daytime/lets_make_a_deal/video/. (Right this moment it's on the second page of Clips.) It's something very like publicity!

bunny_hugger had a good question about the coatis, who were inexplicably presented as a zonk. At least in the 70s show contestants wouldn't really get the zonks but would instead receive a modest consolation prize of enough value that they shouldn't press their claim to some novelty bit of nonsense. While I don't know with certainty just how the modern Deal works I can't imagine that the show would be more reckless in giving away animals (exotic or otherwise) than they were forty years ago.

Trivia: In the first quarter of 1824, about 250 private bills were filed in the British parliament to set up companies (many of them insurance companies). Source: The Company: A Short History of A Revolutionary Idea, John Micklethwait, Adrian Wooldridge.

Currently Reading: Empire Of The Air: The Men Who Made Radio, Tom Lewis.

Comments

[gafennec.livejournal.com]

Zonks were $100 in cash. Everybody had to agree to take them in lieu of actual zonks which might have been funny, but some were pretty expensive jokes.

[rcoony.livejournal.com]

What a ripoff. Raccoonery is worth far more than that.

[orv.livejournal.com]

True, but on the other hand, $100 in cash won't casually disassemble/destroy everything in your house. ;)

[austin-dern.livejournal.com]

Well, no, not if you don't apply it right anyway.

[austin-dern.livejournal.com]

And when you think of the coati premium on top of raccoonery, well!

[austin-dern.livejournal.com]

Was it $100? I knew it was something respectable in the consolation-prize category of things --- this was a time when the Big Deal of the Day might be a new Vega, after all --- but the exact was always vague to me. Anyway, I'd assume they've increased it a bit for the new show, if for nothing else to forego clowns who insist they have a perfect right to the rusted-out hulk of a Vega for some reason, but it's still probably that order of actual prize.

[gafennec.livejournal.com]

Well that was the old show. I can't speak for the new show. It probably was increased.

[austin-dern.livejournal.com]

Probably was. I had an idea of the scale of the consolation prize from the old show, but not just where it was set.

[mondhasen.livejournal.com]

I loved this show when Monty Hall hosted it, long time gone. I didn't realize it was running again.

Nice clip (page 3 at the moment). She (or he?) got a whopping $100, and a purse. The family of coatimundis should have been the top prize, not a zonk ;o)

[austin-dern.livejournal.com]

Ah! Wonderful work getting a screen grab, thank you.

The trouble with coatis as prizes is you really, really have to know what you're doing with them, and I'm not comfortable assuming contestants have that background.

[bunny-hugger.livejournal.com]

Ah, do you suppose I know what I'm doing?

[austin-dern.livejournal.com]

I think you do a fantastic job.

[chefmongoose.livejournal.com]

42/49/51 is a smaller split than I'd expect, at that. But there was that first month, and.. well, this is why you're taking a large statistical sample.

[austin-dern.livejournal.com]

It gets a little more lopsided by adding in the partial season I got from 2008-2009. From that stretch the winning distribution was 84-106-124. With an overall distribution of 126-155-175 that's a winning percentage of, ah, 27.6/34.0/38.4, which is not overall very different where where it stood during summer break. That's reassuring, actually, that it's stabilizing like that.

[chefmongoose.livejournal.com]

That seems close to my initial rough guesses (1/4 1/3 2/5, and yes i know that's only adding up to 98.3% , I said it was rough). One of us really must break this down thoroughly at some point.

With two different stretches both giving the same overall results of the better odds for later spinners, I think we can consider this rather empirically confirmed.

[austin-dern.livejournal.com]

I haven't put any serious thought into error bars, so I think the 1/4 and 1/3 and 2/5th breakdown is probably a close enough approximation to the actual values and simple enough to be remembered.

Now the problem is explaining why it should be those numbers, and how they would change with variants on the Showdown, such as the one you outlined where there were only two possible 'spins', or where there were four contestants, or the like.

[chefmongoose.livejournal.com]

Four contestants, now that'd be rather a tricky chain of dependent events to calculate.

It seems someone's done the very solid math at http://ow.ly/WKhz , but it's a locked article. drat.

--Chi

[austin-dern.livejournal.com]

Even three contestants is quite a chain. I'm really sure I'd have it worked out except I went about trying to follow the whole chain by hand and I keep making little errors of calculation.

I ought to look into the literature about the Showcase Showdown, and maybe other games too. I think the most interesting question would be whether it's a good strategy to listen to the crowd's opinion, but that's also something pretty much impossible to actually test.