Let’s jump right in, since I already previewed this long-awaited statistics analysis. Reminder: I will make the entire dataset public soon.
I’d like to begin with the 5th digit, because it is often the most talked about, and even though it may seem like the hardest to predict, it’s actually not. I was recently arguing with a co-worker about this digit: I thought it’s a good bet to go with 5 or 0, but she thought it’s always some random number, better to guess something like 2 or 7. And my quest to prove her wrong led to countless painstaking hours inputting the prices of 200 cars into a database.
But it was worth it. Cause I was right.
As you can see from this chart, the most common number for the 5th digit is 5, occuring 27% of the time. A random standard distribution would predict each number appears 10% of the time, so anything above 10% is above average. 27%, in my opinion, is quite ridiculous. More than 1 in 4 cars end in 5. Eat it, coworker.
5 is followed by 0, occuring 20.5% of the time, or about twice as much as would be expected by chance. Together, this means that 47.5% of cars on TPIR, almost HALF, end in 0 or 5. You might say, though, that more than half of the cars (52.5%) DON’T end in a 5 or 0, so you should always guess something else, as my coworker said when I showed her these results, but if you said that, you do not understand how statistics work. Based on this analysis, it is ALWAYS better to guess a 5 or 0 as the last digit (unless, in future analyses, we find that certain makes or models trend differently, in which case I will revise this recommendation when playing for said make or model).
The next bit of data surprised me. Many prizes on TPIR end in 9, and I would’ve guessed that it would be the next most probable 5th digit. I was surprised, and intrigued, to learn that 9 only appears in the 5th spot 7.5% of the time (the 5th most common 5th digit), while 4 appears in the 5th spot a whopping 16.5% of the time, making it the 3rd most common number in that spot behind 0 and 5. Too many numbers in one sentence? Let me boil it down: If 5 and 0 aren’t options, and in certain games, even if they are, 4 is a great guess. It is WAY overrepresented in the last digit of TPIR cars. Watch the show, and I bet you’ll be amazed by how often it pops up (and how you never noticed it before). I have highlighted this data point yellow in the chart, since it was the most surprising to me.
The numbers 2, 6, 7, and 9 each appear between 7% and 9% of the time as the final digit, and though I have not yet run significance tests, I am fairly confident that this 7-9 range is due to random variation. Thus, these 4 numbers have a below average chance of occuring, but are still decent guesses after 5, 0, and 4.
1, 3, and 8, on the other hand, turned out to be surprisingly low. 3 and 8 are each at 1.5%, meaning that out of 200 cars, each of those numbers only appeared 3 times as the 5th digit. 1 is just slightly higher at 2.5%; still way below average. Again, I have not run statistical analyses to determine if these results are significant, but my intuition is telling me that these three numbers appear significantly less than should be predicted by chance. I guess I’ll get around to running these tests if you all don’t want to take my word for it.
Or maybe someone out there could run the tests for me?
Nah, I’ll do it. Don’t worry. And stay tuned for more car stats in the future. We’ll take a look at each digit in the same way, and then get into more complex analyses as new issues arrise.