Saturday, June 29, 2019

The Ultimate Price is Right Strategy Guide: Balance Game

Balance Game

Rules
A prize is shown. Next to the contestant is a table with 4 bags. One of the bags has a 3 digit price on it; those are the last three digits of the price of the prize. The other three bags have values that are multiples of $1,000. The contestant must choose which two of those bags, when added together with the 3 digits from the 4th bag, correctly match the price of the prize.

Random fact
This is the second iteration of Balance Game; the first iteration ran from 1984-1985 and was nothing like this one except the fact that scales are involved in the prop. However, that game involved Barker dollars to balance a scale, and thus the Barker dollars (and now Drew dollars) that are supposedly in the bags of the current version are an homage to the 1984 version of the game. Here's a video of the old version:



Win-Loss record (seasons 34-46)
(Note the current iteration of Balance Game was introduced in season 34.)
  • Actual: 80-109 (42.33%)
  • What it would be by random chance: 1/3 (33.33%)
Number of times when, in order to win, the contestant had to pick...
  • The two largest valued bags: 63 (33.33% of playings)
  • The largest valued bag and the smallest valued bag: 69 (36.51%)
  • The two smallest valued bags: 57 (30.16%) 
Strategy
Know the price. Those numbers above aren't statistically far enough away from pure randomness to suggest a strategy.

Friday, June 28, 2019

The Ultimate Price is Right Strategy Guide: Any Number

Any Number

Rules
Two prizes are shown: a car and a prize with three digits in its price. A board is shown with three rows: one for the car, one for the 3 digit prize, and one for the piggy bank. The first number of the car is revealed. In the remaining 10 squares, each number 0-9 inclusive appears exactly once. The contestant picks digits one by one; after each selection, the location of that digit is revealed. The contestant wins the prize whose digits they fill in first.

Random fact
In addition to being the first game in alphabetical order, this is the first game that was ever played on the current iteration of the show. You can see it, and the entire first episode of the show, here:

(Note the still picture is of Bonus Game, not Any Number.)

Key Stats
  • Win-loss record (seasons 29-46): 166-285 (36.81%)
  • Chance of winning if you pick numbers completely randomly: 9/35 (25.71%)
    (Note I calculated that by enumerating all orderings of choosing the numbers 0-9. Later, I found this mathematical analysis of the game. Thankfully, we both ended up with the same result.)
  • Which digit appears where? (Seasons 40-46)

                     % in 3 digit  % in piggy
    Digit  % in car      prize         bank  
      0      16.07       47.62        36.31
      1      45.24        8.93        45.83
      2      44.64       16.67        38.69
      3      58.92       12.50        28.57
      4      39.88       23.81        36.31
      5      29.76       46.43        23.81
      6      50.00       33.33        16.67
      7      35.71       32.14        32.14
      8      40.48       35.71        23.81
      9      39.29       42.86        17.86


    Let me explain how to read that table with an example. The first entry, 16.07, means that the number 0 was in the car 16.07% of the time, which is less than 1/6. By random chance, you would expect it to be in the car 40% of the time. (Actually, you could argue that the first digit of the 3 digit prize and the first digit of the piggy bank are never 0, so it should be in the price of the car 50% of the time based on random chance. But I digress.)
  • The ordering of the piggy bank digits (seasons 40-46):

    Order    %  

    $3.21  38.69
    $3.12  48.81
    $2.31   4.76
    $2.13   7.14
    $1.32   0.00
    $1.23   0.60


    This table is referring to the ordering of the digits in the piggy bank. "$3.21" means the largest digit that appears in the piggy bank is first, the middle digit is second, and the smallest digit is last; a piggy bank price of $9.42 would fit that example. Similarly, $2.31 means the largest digit is in the middle of the piggy bank, the second largest digit is in the first spot, and the smallest digit is in the last spot; a piggy bank price of $4.92 would fit that example.
Strategy
The strategy comes out of those tables above, along with the paint and fabric protection rule. Cars in Any Number almost always have paint and fabric protection in order that the last number won't be 0. In fact, in those 168 playings of Any Number, the last digit of the car was a 0 eight times. That's less than 1 in every 20 playings. The 5 was the last digit only 11 times. So here is the strategy to this game:
  1. Start by trying to find the second digit of the car, as you should be able to get that in just a couple of tries. If the first digit is a 2, go for a 1, 2, or 3 as the second digit if you're not sure. And this is where you can try a 0 if the first digit is a 2. Of the 27 times a 0 was in the price in the car from seasons 40-46, 18 times it was the second digit. 
  2. Once you've found that, pick the 3 and the 6 if you haven't already. You have at least a coin flip chance of those being in the car.
  3. At this point, you've probably found at least one number in the 3 digit prize and/or the piggy bank. Note the largest digit in the piggy bank is the first digit 87.5% of the time, so if you see _ 6 _ in the piggy bank, pick numbers less than 6. But if you have 6_ _ in the piggy bank, guess numbers greater than 6. Update: jhc2010 pointed out at golden-road.net that the first digit of the piggy bank has been its largest digit in every playing since October 29, 2015. So you should really follow this piece of advice.
  4. If at any point the above tips don't help, and you can't figure out which numbers would appear in the 3 digit prize, then pick the remaining numbers in the order they appear most frequently in the car. That order is 3,6,1,2,8,4,9,7,5,0. But don't just pick them haphazardly in that order--after each digit, look to see if you can apply the piggy bank concept in rule 3 or if you got any clues to what the 3 digit price prize might be.
  5. Speaking of the 3 digit prize, thanks to pannoni1 at golden-road.net, I realized there hasn't been a prize under $500 offered since season 42. So feel free to try the lower numbers a bit if you only have the first number to go in the second prize.
  6. If you can't remember that order, just remember this: don't pick the 5 or the 0. Those are by far the least frequent numbers in the car, which is not surprising, since those are the numbers contestants pick the most frequently thinking they're the last number in the car. But they almost never are.

Thursday, June 27, 2019

The Ultimate Price is Right Strategy Guide: The Showcases

The showcases

Rules
Two contestants participate. A showcase with multiple prizes is shown to both contestants. The contestant who has won more up to that point in the show decides if they want to bid on that showcase or pass it to the other contestant. The second showcase is then presented and whoever didn't bid on the first showcase bids on the second one. Whichever contestant is closer to the actual retail price of the showcase they bid on without going over wins their showcase. If the winner was less than or equal to $250 away from the price of their showcase, they win both showcases.

Random facts
  • Do NOT call this the showcase showdown!! The showcase showdown refers to the wheel, not the showcase round at the end of the show. There's probably no easier way to make a normally casual, easy-going Price is Right fan suddenly go ballistic than by calling this segment the showcase showdown.
  • The closest difference between the two contestants was $1:
  • In 2008, Terry Kneiss bid perfectly on his showcase. There's a whole documentary about how that happened which I highly recommend.
Who won? (Seasons 29-46)
  • The top winner won their showcase 43.09% of the time.
  • The runner up won their showcase 49.98% of the time.
  • A double overbid happened 6.93% of the time.
Pricing patterns
  • Showcases are never below $20,000. (The last time a showcase cost less than $20,000 was in season 42, 5 years ago).
  • The last three digits of the price of each showcase are always between -251 and -999, inclusive. (I believe the last time the last three digits of a showcase's value were between -000 and -250, inclusive, was season 40, 7 years ago. I haven't been able to check this exhaustively, so I welcome correction. But I do know it's been a long time.)
Strategy
While nothing beats knowing the prices of the prizes, you should use that last bullet point above to your advantage. The reason they don't use prices ending between -000 and -250 is believed to be that the producers don't want a contestant to win both showcases by bidding a value that's divisible by $1,000. For example, they don't want bids like $23,000 to win both showcases, so they'll add an excursion to a trip or an option to a car to raise the price of a showcase from $23,123 to $23,452. Thus, the bid of $23,000 doesn't win both showcases. As a result, the last three digits of your bid should never be between -000 and -250, inclusive. So instead of bidding $23,000, bid $23,251. And thus your minimum showcase bid should be $20,251 when you combine the two bullet points in the pricing patterns section above. So to sum up:
  • Your bid should be at least $20,251.
  • The last three digits of your bid should be between 251 and 999 inclusive.

Wednesday, June 26, 2019

The Ultimate Price is Right Strategy Guide: Showcase Showdown (a.k.a. The Wheel)

Showcase Showdown (a.k.a. The Wheel)

Rules
There is a wheel with 20 values on it; those values are multiples of 5 cents from 5 cents to $1 (so 5 cents, 10 cents, 15 cents, ..., $1.) Each contestant must spin at least once and can spin twice if they choose to. A spin must go all the way around or it doesn't count. A contestant's score is the value of their one spin or sum of their two spins. The contestant who gets closest to $1 without going over goes to the showcase. If two or more contestants tie, they each get exactly one more spin; the contestant who spun the higher or highest value during the spin-off goes to the showcase. And there's a bonus and bonus spins available for getting exactly $1, but that's irrelevant to the strategy.

Random fact
The original wheel looked nothing like today's wheel; instead it looked like a carnival wheel with various colors. It's often referred to as the Rainbow Wheel. You can see the first playing of it here:



Who actually won? (seasons 29-46)
  • The first spinner won 30.17% of the time.
  • The second spinner won 33.81% of the time.
  • The third spinner won 36.05% of the time, so a small, but not huge, advantage.
Probabilities
(The following probabilities were calculated by enumerating all possible combinations of wheel spins and applying the strategy below.)
  • Who wins:
    • First spinner: 30.82%
    • Second spinner: 32.96%
    • Third spinner: 36.22%
  • Getting $1 in one spin or a combination of two spins:
    • At least one spinner: 23.03%
    • At least two spinners: 2.35%
    • All three spinners: 0.08%
      (Note the above three cases exclude the possibility of getting $1 in a spin-off, which would slightly increase the numbers.)
  • Having a spin-off with...:
    • Exactly two players: 11.32%
    • All three players: 0.48%
Strategy
The only decision a contestant makes during this segment of the show is whether to take their second spin or not. How hard can that be? It turns out it's hard enough that people have written papers on it. So rather than attempt to derive the numbers here, I'll just spit out the results:
  • Spinner 1 strategy:
    • Stay if you spun 0.70 or more.
    • Spin again if you spun 0.65 or less.
  • Spinner 2 strategy:
    • If you spun less than spinner 1, spin again. (Duh.)
    • If you spun more than spinner 1, or if spinner 1 went over, stay if and only if you spun 0.55 or more. If you are in the lead after your first spin but spun 0.50 or less, spin again.
    • If you tied spinner 1, stay if and only if you spun 0.70 or more. If you tied spinner 1 but spun 0.65 or less, spin again.
  • Spinner 3 strategy:
    • If you spun less than the total of whoever is leading after the first two contestants have spun, spin again. (Duh.)
    • If you spun more than the total of whoever is leading after the first two contestants have spun, stay. (Again, duh. In fact, Drew probably won't even let you spin again in this case.)
    • If you have exactly one person to beat and you tied them, stay and take the spin-off if and only if you spun 0.55 or more. If the value you tied is 0.50 or less, spin again.
    • If the first two spinners tied and you tied both of them, stay and take the spin-off if and only if you spun 0.70 or more. If you tied two people at 0.65 or less, spin again.

Tuesday, June 25, 2019

The Ultimate Price is Right Strategy Guide: Items up for Bid

Items up for bids

Rules
A prize is shown. Each contestant bids on the prize; whoever bids closest to the actual retail price (ARP) without going over wins the prize and goes on stage to play a pricing game. If all four contestants give a bid that is over the ARP, a buzzer sounds and the contestants get another chance to bid on the prize. If a contestant bids the exact price, they get a $500 bonus.

Random fact
If you want to impress your friends, you can call these "one bids," as that's the term used by the show to refer to the prizes the contestants bid on. This term goes all the way back to the original The Price is Right hosted by Bill Cullen in the 1950s and 1960s. You can even use that fact as a pick-up line: "Hey, if they used you as a one bid on The Price is Right, it would be impossible to for anyone to overbid because you're priceless." If you do use that pick-up line, feel free to *not* tell the person you heard it from me. 

Who won? (Seasons 29-46)
  • Bidder #1 won 18.19% of the time.
  • Bidder #2 won 19.94% of the time.
  • Bidder #3 won 22.05% of the time.
  • Bidder #4 won 39.82% of the time.
Pricing patterns
  • One bid prizes are never less than $500. There has not been a one bid prize worth less than $500 since season 39 (8 years ago). There have been a number of one bids that were exactly $500 since then, but none under.
  • I've rarely seen trips used as one bids cost less than $2,000. I don't have an easy way to check this with any exact stats, though, so use with care.
Strategy based on your position in the bidding
Your strategy in contestants' row depends on your position. But your bid should never be less than $500.

Contestant #1: Bid what you think the price is, with a minimum bid of $500.

Contestants #2 and #3: Bid what you think the price is, except:
  • Your minimum bid should be $500. This supersedes all points below.
  • If you're going to bid less than someone, give yourself at least $100 of wiggle room. For example, if bidder #1 bid $1150 and you think the prize is $1100, bid $1050. (Note: I admit the $100 is more gut feeling and less something stats can back up, but only having a $50 range of prices you can win on just isn't enough.)
  • DO NOT overbid someone by $1 in position 2 or 3. All that does is tempt the person after you to do the same to you. I would advise that if you want to bid more than someone else, your bid should be at least $50 over their bid. (Again, that's arbitrary, but a $50 range at least gives contestant #4 something to think about when deciding whom they should bid $1 more than.)
  • Similarly, DO NOT bid $1 in position 2 or 3. Not only are you in violation of the $500 minimum bid rule, that just tempts someone to bid $2.
Contestant #4: The fourth contestant should bid $1 more than someone else. Yes, the person you 1-up will be mad at you, but that's the show. Also, it is almost always wrong to bid $1 or similar. There are two reasons:
  1. The minimum price of the one bid is $500. You might as well bid $500 and try to hit it right on the nose. But you shouldn't do that either because...
  2. More importantly, your chances of winning when you bid $1 (or $500) drop from 33.33% to 25%. Let me explain with an example:
  • Anne bids $600
  • Bob bids $750
  • Charlie bids $900
  • David bids $751
What is David's chance of winning? Unless Bob got it exactly right, David has shut him out, which means only 3 contestants (Anne, Charlie, or David) can win the prize. Thus, David's chance is 1/3, or 33.3333...%. Even if everyone has bid too much, everyone just bids again, and David's chance of winning is still 1/3 if he overbids someone by $1. But if:
  • Anne bids $600
  • Bob bids $750
  • Charlie bids $900
  • David bids $1 (or $500)
Now, David's chance of winning is 1/4 (25%), as no contestant is shut out. "But Brian," you may be thinking, "those percentages assume that you're picking the winner randomly. Contestants aren't bidding random prices so this doesn't apply!" The stats suggest the opposite. From seasons 41-46, less than 27% of $1 bids from the fourth bidder ended up as winning bids--that's pretty close to the 25% you'd get from random chance. That means that over 73% of the time, $1 bids that were made were wrong--that's almost 3 out of 4. Compare that to the fact that when bidder 4 doesn't bid $1, they won 44.1% of the time--that's more than 50% more often than when they bid $1.

To be fair, I said bidding $1 or $500 is "almost" always wrong. I can think of two rare exceptions:
  1. If you are somehow absolutely positive that everyone else has overbid, then go ahead and bid $500. But you'd better be really positive. Prices are almost always higher than you think they are.
  2. If one of the first three bidders has bid $1 (or something similarly low), feel free to 1-up them, even if that means bidding $2 or $70. They deserve to lose for having that horrible of a strategy. Technically, you should still bid at least $500 in that case, but as a viewer, I take sadistic pleasure in it when the fourth bidder bids $2 after someone else has bid $1 😈.
Addendum thanks to a poster at golden-road.net: Who should bidder #4 1-up? If bidder 4 isn't sure of the price, they should not 1-up bidder #1. The reason is this: if they don't win now, they want a chance to be bidder #4 again next round. If they 1-up bidder #1, bidder #1 cannot win (unless they got the price exactly right) and thus you've made it so you're guaranteed not to be bidder #4 next round. Thanks to jhc2010 for that insight!

Monday, June 24, 2019

The Ultimate Price is Right Strategy Guide: Introduction

Hi! What is this, you ask? It's a project that's been ratting around in brain since my appearance on The Price is Right. I used strategy to win Master Key, so what other games have strategy? As a completionist, there's no way I could do this for just a couple of games. Thus, my goal is to create the most comprehensive Price is Right strategy guide anywhere on the Internet. I will include stats and tips about every single game in the current rotation (as of June 2019, there are 77 pricing games). Granted, for some games it will be "there is no strategy, just know the price." But you may be surprised about some of the strategies for some of the games. For example, did you know there's a strategy in Magic # that has guaranteed a win in every single playing of that game since season 43 without knowing anything about the actual prices? Or that contestants play Cover Up so horribly that even though they usually know the first two digits of the car's price, they barely win more than they would if they just picked everything completely randomly? Stay tuned!

My plan is to post one of these every day, six days a week. (Sunday will be a day of rest.) I'll start with three posts for things that happen on every show: items up for bids, the showcase showdowns (a.k.a the wheel), and the showcases. Then I'll go through all 77 games in alphabetical order, for a total of 80 posts after this one. My goal is to be done before season 48 starts in September, though they haven't announced the start date of the new season yet, so we'll see how it goes.

Resources: I would be remiss if I didn't post three key resources I'll be using to come up with my strategies:
  • http://www.golden-road.net
    THE best site for TPiR information on the internet, bar none. They have very active forums there and people there have come up with many of the patterns I'll be sharing throughout my posts. They also have incredible information about the show's past, insider photos, and so on. I highly recommend it. (My username there is LiteBulb88 should you wish to see my posts.)
  • http://www.tpirstats.com
    This is a fan of the show who has posted TPiR stats since season 29*. Most of the raw data I use to calculate my stats comes from this site.
  • MATLAB
    I'm using MATLAB to do lots of number crunching on the data from the TPiR stats web page. Full disclosure: I worked at the company that makes MATLAB for over 13 years, and I still consult for them. I am a big believer in their products! I don't plan to post the actual code so as not to overwhelm the posts, but I'm happy to share the scripts if anyone asks.
Why am I not using season 47 data very much? While most of season 47 has aired, there are a couple of preempted episodes and a July 4 episode yet to air. According to the Price is Right calendar, the last first-run season 47 episode doesn't air until July 17. Thus, to avoid using data from an incomplete season that could quickly go out of date after I post, I'll mostly be using data through season 46 only.
Some general strategies:

Trip rule: The farther a trip is from Los Angeles, the more expensive the price. So if you have to decide which trip is more expensive, New York or Salt Lake City, simply remember the trip rule and you'll choose NYC and be right every time.

The rule of pairs: If you have two of an identical prize (such as two identical motorcycles, two identical surfboards, etc.), then the total price is almost certainly going to end in an even number. The reason is that 2 times any integer is guaranteed to be even! The only reason it might be odd is if the original price is something like $2999.45. The show adds and then rounds to the nearest dollar, not the other way around, so in this case, the total price would be $5999, not $5998, as $2999.45 + $2999.45 = $5998.90, which gets rounded to $5999. But that kind of price is rather rare, so if you have a choice of an even or odd number for the price of 2 of an item and you're not sure, go for the even price.

Paint and fabric protection car option rule: If you're playing for a car, listen to the options that George describes. If you hear paint and fabric protection as one of the options, the price will NOT end in a 0 or 5. If you don't hear paint and fabric protection, there's a chance the price will end in a 0 or 5, though it's not guaranteed.

Pick the end points: If you ever see an arrangement of items, pick the one on the far left or far right! (Yup, that was my Master Key strategy.) Most people find it more comfortable to pick the items in the middle and the producers know this. I'll talk about how this applies in specific games as I go along.

The following two were pointed out at golden-road.net after I posted this article:

Consecutive car digits rarely repeat: With the exception of the first two digits of the price, consecutive digits in a car's price rarely repeat. So $22,136 is a price that could happen, but $21,336 or $24,419 are not likely due to this rule.

Three or more small prizes rarely have the same correct decision: If you have three or four small prizes that you need to guess higher/lower or true/false on, it's extremely rare that all three or four have the same answer. It's almost always a 2/1 split in the case of three small prizes and usually a 3/1 or less commonly a 2/2 split in the case of four small prizes. I'll point out specific stats in those games as I go along.

And finally...

Get inside your own head: The producers are inside your head before you ever make it on to the stage. They know you're more likely to pick numbers 2-5 in Pass the Buck, choose 5's in Lucky $even, or feel the pressure of everyone staring at you while you're pushing the lever in Magic # and thus stop too soon because you feel it just can't be right to take a long time to set the value.  If you want to maximize your chances of winning, you need to figure out what you'd be afraid to do while playing a game and go counter to that fear. Magic # is an excellent example--people are afraid of taking too long to push the lever. As a result, the game is never lost by putting the price too high; it's always lost by putting the price too low. (In fact, out of the 67 losses in that game from seasons 32-46, ONE of those losses was by setting the number too high.) If you know this in advance, you'll recognize that fear when you get on stage, be willing to set that price high, and have a much better chance of winning. And the time to figure that out is before you ever go to the show; once you're on that stage, things move so fast you probably won't have time to unpack everything you're thinking.

Good luck out there! And if you do use these strategies to win something on the show, drop me a line--I'd love to hear about it!

*A season of the Price is Right typically runs from September of one year to August of the next. There are some exceptions where a season has spilled into September of the next year, but that's not terribly relevant for the information I'll be posting. Here are some specific seasons and the approximate time frames they refer to:

Season 29: Fall 2000-Summer 2001
Season 30: Fall 2001-Summer 2002
Season 31: Fall 2002-Summer 2003
Season 32: Fall 2003-Summer 2004
Season 33: Fall 2004-Summer 2005
Season 34: Fall 2005-Summer 2006
Season 35: Fall 2006-Summer 2007
Season 36: Fall 2007-Summer 2008
Season 37: Fall 2008-Summer 2009
Season 38: Fall 2009-Summer 2010
Season 39: Fall 2010-Summer 2011
Season 40: Fall 2011-Summer 2012
Season 41: Fall 2012-Summer 2013
Season 42: Fall 2013-Summer 2014
Season 43: Fall 2014-Summer 2015
Season 44: Fall 2015-Summer 2016
Season 45: Fall 2016-Summer 2017
Season 46: Fall 2017-Summer 2018
Season 47: Fall 2018-Summer 2019